Open Collections

UBC Theses and Dissertations

UBC Theses Logo

UBC Theses and Dissertations

Harnessing neighbourhood walkability’s health potential by providing a supportive pedestrian environment… Adhikari, Binay 2021

Your browser doesn't seem to have a PDF viewer, please download the PDF to view this item.

Notice for Google Chrome users:
If you are having trouble viewing or searching the PDF with Google Chrome, please download it here instead.

Item Metadata

Download

Media
24-ubc_2021_may_adhikari_binay.pdf [ 12.02MB ]
Metadata
JSON: 24-1.0396329.json
JSON-LD: 24-1.0396329-ld.json
RDF/XML (Pretty): 24-1.0396329-rdf.xml
RDF/JSON: 24-1.0396329-rdf.json
Turtle: 24-1.0396329-turtle.txt
N-Triples: 24-1.0396329-rdf-ntriples.txt
Original Record: 24-1.0396329-source.json
Full Text
24-1.0396329-fulltext.txt
Citation
24-1.0396329.ris

Full Text

HARNESSING NEIGHBOURHOOD WALKABILITY’S HEALTH POTENTIAL BY PROVIDING A SUPPORTIVE PEDESTRIAN ENVIRONMENT:  A COMPARATIVE ANALYSIS ACROSS  DEMOGRAPHIC AND PSYCHOSOCIAL FACTORS  by BINAY ADHIKARI  B.Arch., Tribhuvan University, 2009 MCRP, The University of Texas, 2013  A THESIS SUBMITTED IN PARTIAL FULFILLMENT OF THE REQUIREMENTS FOR THE DEGREE OF  Doctor of Philosophy in THE FACULTY OF GRADUATE AND POSTDOCTORAL STUDIES (Planning)  THE UNIVERSITY OF BRITISH COLUMBIA (Vancouver)  March 2021  © Binay Adhikari, 2021 ii  The following individuals certify that they have read, and recommend to the Faculty of Graduate and Postdoctoral Studies for acceptance, the dissertation entitled: HARNESSING NEIGHBOURHOOD WALKABILITY’S HEALTH POTENTIAL BY PROVIDING A SUPPORTIVE PEDESTRIAN ENVIRONMENT: A COMPARATIVE ANALYSIS ACROSS DEMOGRAPHIC AND PSYCHOSOCIAL FACTORS  submitted by Binay Adhikari in partial fulfillment of the requirements for the degree of Doctor of Philosophy in Planning  Examining Committee: Dr. Lawrence D. Frank, School of Community and Regional Planning, UBC Supervisor  Dr. Mark Stevens, School of Community and Regional Planning, UBC Supervisory Committee Member  Dr. Guy Faulkner, School of Kinesiology, UBC Supervisory Committee Member Dr. Louise C. Mâsse, School of Population and Public Health, UBC University Examiner AnnaLisa Meyboom, School of Architecture + Landscape Architecture, UBC University Examiner  Additional Supervisory Committee Members: Dr. Ying MacNab, School of Population and Public Health, UBC Supervisory Committee Member   iii  ABSTRACT  Cities and municipalities have invested a significant amount of their time and budgets in creating walkable neighbourhoods to help their citizens be healthier. However, they have not been able to harness the health potential of neighbourhood walkability fully. This study examines whether neighbourhood walkability’s health potential can be harnessed by providing a supportive pedestrian environment. Specifically, it examines whether there is a synergistic effect of neighbourhood walkability and the pedestrian environment on physical activity in children, teens, and older adults. A synergistic effect, or “synergism,” occurs when two or more processes (or factors) interact so that their collective effect is greater than the sum of their separate effects (Myers, 1989, p. 506). This can also be called a “multiplicative effect.” Cities and municipalities are often constrained by budgetary restrictions and need to make decisions that give the highest return. Understanding the synergy between neighbourhood walkability and the pedestrian environment is essential for planners and policymakers to make investment decisions cost-effective and maximize walkable neighbourhoods' health potential. The synergy between neighbourhood walkability and the pedestrian environment is assessed using a series of interaction models. Interaction models help researchers understand the multiplicative effect of two or more factors on an outcome. This study shows various patterns of interactions between neighbourhood walkability and the pedestrian environment across demographic and psychosocial factors.  Planners and policymakers have started to recognize the effects of their decisions on active transport, physical activity, and related health outcomes. However, the century-long land-use and transportation planning approach has created an urban form that requires significant effort to become favourable for walking and physical activity. Making small-scale changes to the iv  pedestrian environment can be an effective approach to fine-tune some of the existing infrastructure created to make neighbourhoods walkable. The evidence provided in this study, though not fully generalizable, is useful for city planning authorities to formulate plans and policies to design better pedestrian environments that can help retrofit the existing urban infrastructure and harness the health potential of neighbourhood walkability.   v  LAY SUMMARY  In this dissertation, I examine whether neighbourhood walkability's health potential can be harnessed by creating a pedestrian-friendly environment. I do so by investigating the Synergy between neighbourhood walkability and the pedestrian environment. A synergistic effect, or “synergism,” occurs when two or more processes (or factors) interact so that their collective effect is greater than the sum of their separate effects. In simpler terms, the synergy between neighbourhood walkability and the pedestrian environment is analogous to adding icing to a cake or adding condiments to food, where icing and the condiments represent the pedestrian environment. These small additions to a cake or food greatly enhance the taste. Similarly, my study results show that providing a supportive pedestrian environment can enhance the health potential of neighbourhood walkability. The results of this study provide evidence for city planners to make investment decisions that yield greater return and maximize the health potential of walkable neighbourhoods.    vi  PREFACE This dissertation used data from three different studies that were conducted in Seattle, Baltimore, and San Diego, USA. Dr. Jim Sallis, Dr. Brian Saelens, Dr. Abby King, and Dr. Lawrence Frank were core collaborators of these studies. I developed three manuscript prospectuses to use the data from the three studies. The study team then reviewed the prospectuses before approving my access to the data. Afterwards, I compiled the three prospectuses to develop a dissertation prospectus, which was reviewed by my dissertation committee. The datasets did not have regional accessibility or weather information; I added this information using publicly available data.  One manuscript using the older adult dataset is currently under development. While I am involved in most of the writing of the manuscript, it will also be reviewed by the co-PIs of the studies for their comments. The other two manuscripts will be developed following a similar procedure. I will be the first author on all three manuscripts.  The Behavioural Research Ethics Board at UBC (H13-03453) approved this research. Additionally, all three studies which provided the data for this research were approved by their respective institutions.   vii  TABLE OF CONTENTS Abstract ......................................................................................................................................... iii Lay Summary .................................................................................................................................v Preface ........................................................................................................................................... vi Table of Contents ........................................................................................................................ vii List of Tables ............................................................................................................................. xvii List of Figures ...............................................................................................................................xx Acknowledgements ....................................................................................................................xxv Chapter 1: Introduction ................................................................................................................1 1.1 Introduction ..................................................................................................................... 1 1.2 Significance of Examining Synergistic Effects .............................................................. 2 1.3 Importance of the Study for Planners and Policymakers ................................................ 3 1.4 The Rationale for Focusing on Children, Teens, and Older Adults ................................ 6 1.5 Policy Implications ......................................................................................................... 8 1.6 Aim ................................................................................................................................. 9 1.7 Interaction Effects to Assess Synergy............................................................................. 9 1.8 Objectives, Hypotheses, and Rationale ......................................................................... 14 1.8.1 Objective 1 ................................................................................................................ 14 1.8.1.1 Hypothesis......................................................................................................... 14 1.8.1.2 Rationale ........................................................................................................... 15 1.8.2 Objective 2 ................................................................................................................ 16 1.8.2.1 Hypothesis......................................................................................................... 16 1.8.2.2 Rationale ........................................................................................................... 16 viii  1.8.3 Objective 3 ................................................................................................................ 17 1.8.3.1 Hypothesis 1...................................................................................................... 18 1.8.3.2 Rationale ........................................................................................................... 19 1.8.3.3 Hypothesis 2...................................................................................................... 19 1.8.3.4 Rationale ........................................................................................................... 20 1.9 Research Context .......................................................................................................... 20 1.10 Organization of the Dissertation ................................................................................... 23 Chapter 2: Theoretical Background ..........................................................................................24 2.1 Overview ....................................................................................................................... 24 2.2 Socio-Ecological Model of Health Behaviour .............................................................. 24 2.3 Synergistic Effects of Multiple Macro-Built Environment Factors on Walking Behaviour .................................................................................................................................. 28 2.4 The Interactions Between the Built Environment and Psychosocial Factors ............... 32 2.5 Place Types Based on Neighbourhood Walkability and the Pedestrian Environment . 33 2.5.1 Place Types Based on Neighbourhood Walkability ................................................. 33 2.5.1.1 Low-Walkability Neighbourhoods ................................................................... 34 2.5.1.2 High-Walkability Neighbourhoods ................................................................... 34 2.5.2 Place Types Based on the Pedestrian Environment .................................................. 35 2.5.2.1 Neighbourhoods with Poor Pedestrian Environment ........................................ 36 2.5.2.2 Neighbourhoods with Good Pedestrian Environment ...................................... 37 2.5.3 Place Types Based on Neighbourhood Walkability and Pedestrian Environment ... 38 2.5.3.1 High Neighbourhood Walkability but Poor Pedestrian Environment .............. 38 2.5.3.2 Low Neighbourhood Walkability but Good Pedestrian Environment .............. 40 ix  2.5.3.3 Low Neighbourhood Walkability and Poor Pedestrian Environment .............. 40 2.5.3.4 High Neighbourhood Walkability and Good Pedestrian Environment ............ 42 2.6 Summary ....................................................................................................................... 43 Chapter 3: Literature Review .....................................................................................................45 3.1 Overview ....................................................................................................................... 45 3.2 The Nexus Between the Built Environment and Walking: Views from Urban Design, Transportation Planning, and Public Health ............................................................................. 45 3.3 Walking as a Form of Physical Activity ....................................................................... 50 3.4 Built Environment Correlates of Walking .................................................................... 52 3.4.1 Metropolitan Scale .................................................................................................... 52 3.4.1.1 The Built Environment at Metropolitan Scale .................................................. 52 3.4.1.2 Built Environment Correlates of Walking at Metropolitan Scale ..................... 54 3.4.2 Neighbourhood Scale ................................................................................................ 55 3.4.2.1 The Built Environment at Neighbourhood Scale .............................................. 55 3.4.2.2 Built Environment Correlates of Walking at Neighbourhood Scale ................ 56 3.4.3 Street Scale................................................................................................................ 60 3.4.3.1 The Built Environment at Street Scale.............................................................. 60 3.4.3.2 Built Environment Correlates of Walking at Street Scale ................................ 61 3.5 Factors Affecting Physical Activity in Children, Teens, and Older Adults.................. 62 3.5.1 Children and Teens ................................................................................................... 64 3.5.1.1 Built Environment Factors ................................................................................ 64 3.5.1.2 Sociodemographic and Psychosocial Factors ................................................... 65 3.5.1.2.1 Sociodemographic Factors .......................................................................... 65 x  3.5.1.2.2 Psychosocial Factors ................................................................................... 68 3.5.2 Older Adults .............................................................................................................. 72 3.5.2.1 Built Environment ............................................................................................. 72 3.5.2.2 Sociodemographic and Psychosocial Factors ................................................... 73 3.5.2.2.1 Sociodemographic Factors .......................................................................... 73 3.5.2.2.2 Psychosocial Factors ................................................................................... 76 3.6 Summary ....................................................................................................................... 80 Chapter 4: Methodology..............................................................................................................82 4.1 Overview ....................................................................................................................... 82 4.2 Data Sources ................................................................................................................. 82 4.2.1 Neighbourhood Impacts on Kids (NIK) ................................................................... 83 4.2.1.1 Neighbourhood Selection.................................................................................. 83 4.2.1.2 Recruitment ....................................................................................................... 84 4.2.1.3 Data Collection Instruments ............................................................................. 85 4.2.1.3.1 Accelerometers to Measure Objective Physical Activity ............................ 85 4.2.1.3.2 Survey Questionnaire .................................................................................. 85 4.2.2 Teen Environment and Neighborhood (TEAN) ....................................................... 87 4.2.2.1 Neighbourhood Selection.................................................................................. 87 4.2.2.2 Recruitment ....................................................................................................... 87 4.2.2.3 Data Collection Instruments ............................................................................. 88 4.2.2.3.1 Accelerometers to Measure Objective Physical Activity ............................ 88 4.2.2.3.2 Survey Questionnaire .................................................................................. 88 4.2.3 Senior Neighborhood Quality of Life Study (SNQLS) ............................................ 89 xi  4.2.3.1 Neighbourhood Selection.................................................................................. 89 4.2.3.2 Recruitment ....................................................................................................... 90 4.2.3.3 Data collection instruments............................................................................... 90 4.2.3.3.1 Accelerometers to Measure Objective Physical Activity ............................ 90 4.2.3.3.2 Survey questionnaire ................................................................................... 91 4.3 Variables used in this study .......................................................................................... 92 4.3.1 Dependent Variables ................................................................................................. 92 4.3.1.1 Self-Reported Physical Activity ....................................................................... 92 4.3.1.1.1 Transport-Related Physical Activity ........................................................... 92 4.3.1.1.2 Leisure and Neighbourhood Physical Activity ........................................... 92 4.3.1.2 Objective Physical Activity .............................................................................. 93 4.3.2 Independent Variable ................................................................................................ 94 4.3.2.1 Neighbourhood Walkability.............................................................................. 94 4.3.2.2 Street-Scale Built Environment Features .......................................................... 95 4.3.2.2.1 Data Collection Method .............................................................................. 96 4.3.2.2.2 The Conceptual Approach to the Scoring System ...................................... 97 4.3.2.2.3 Converting Item Scores to Subscales and Valence scores .......................... 98 4.3.2.2.4 Domain-Specific Scores .............................................................................. 99 4.3.2.3 Psychosocial Variables ................................................................................... 103 4.3.2.3.1 Parents’ Support for Physical Activity (For NIK and TEAN Sample) ..... 103 4.3.2.3.2 Self-Efficacy for Walking (For SNQLS Sample) ..................................... 103 4.3.2.3.3 Social Support for Physical Activity (For SNQLS Sample) ..................... 103 4.3.2.3.4 Parents’ Safety Perceptions (For NIK and TEAN Sample) ...................... 104 xii  4.3.2.3.5 Safety Perception (For SNQLS Sample) ................................................... 104 4.3.2.4 Control Variables ............................................................................................ 105 4.3.2.4.1 Socio-Demography Factors ....................................................................... 105 4.3.2.4.2 Self-Reported Mobility Impairment .......................................................... 105 4.3.2.4.3 Weather Patterns ....................................................................................... 105 4.3.2.4.4 Neighbourhood Self-Selection .................................................................. 105 4.3.2.4.5 Regional Transit Accessibility .................................................................. 106 4.4 Data Analysis Plan ...................................................................................................... 106 4.4.1 Overall Analysis Plan ............................................................................................. 106 4.4.2 Supplementary Analysis Plan ................................................................................. 109 4.4.3 Data Analysis .......................................................................................................... 109 4.4.3.1 Regression Analysis ........................................................................................ 109 4.4.3.2 Model Fit and Robustness Test ....................................................................... 110 4.4.3.3 Visualizing Interaction Effect ......................................................................... 111 4.5 Summary ..................................................................................................................... 111 Chapter 5: Results: Descriptive Statistics................................................................................112 5.1 Overview ..................................................................................................................... 112 5.2 Sample 1: Neighborhood Influences on Kids (NIK) .................................................. 112 5.2.1 Descriptive Statistics ............................................................................................... 112 5.2.2 Bivariate Relationship ............................................................................................. 114 5.3 Sample 2: Teen Environment and Neighborhood (TEAN) ........................................ 117 5.3.1 Descriptive Statistics ............................................................................................... 117 5.3.2 Bivariate Relationship ............................................................................................. 118 xiii  5.4 Sample 3: Senior Neighborhood Quality of Life Study (SNQLS) ............................. 122 5.4.1 Descriptive Statistics ............................................................................................... 122 5.4.2 Bivariate Relationship ............................................................................................. 123 5.5 Summary ..................................................................................................................... 126 Chapter 6: Results: Inferential Statistics .................................................................................128 6.1 Sample 1: Neighbourhood Influences on Kids (NIK) ................................................ 130 6.1.1 Transport Walking or Biking .................................................................................. 130 6.1.2 Neighbourhood Physical Activity ........................................................................... 130 6.1.3 Objective Physical Activity .................................................................................... 130 6.1.3.1 MAPS Grand Score......................................................................................... 130 6.1.3.2 Domain-Specific MAPS Scores ...................................................................... 133 6.1.3.3 Supplementary Analysis Using Valence Scores and Subscales ..................... 136 6.2 Sample 2: Teen Environment and Neighborhood (TEAN) ........................................ 138 6.2.1 Transport Walking or Biking .................................................................................. 138 6.2.2 Neighbourhood Physical Activity ........................................................................... 138 6.2.3 Objective Physical Activity .................................................................................... 138 6.2.3.1 MAPS Grand Score Results ............................................................................ 138 6.2.3.2 Domain-Specific MAPS Scores ...................................................................... 138 6.2.3.3 Supplementary Analysis Using Valence Scores and Subscales ..................... 141 6.3 Sample 3: Senior Quality of Life Study (SNQLS) ..................................................... 141 6.3.1 Transport Walking or Biking .................................................................................. 141 6.3.1.1 MAPS Grand Score......................................................................................... 141 6.3.1.2 Domain-Specific MAPS Score ....................................................................... 145 xiv  6.3.1.3 Supplementary Analysis Using Valence Scores and Subscales ..................... 148 6.3.2 Leisure Walking ...................................................................................................... 150 6.3.2.1 MAPS Grand Score......................................................................................... 150 6.3.2.2 Domain-Specific MAPS Scores ...................................................................... 150 6.3.2.3 Supplementary Analysis Using Valence Scores and Subscales ..................... 150 6.3.3 Objective Physical Activity .................................................................................... 151 6.3.3.1 MAPS Grand Score......................................................................................... 151 6.3.3.2 Domain-Specific MAPS Scores ...................................................................... 154 6.3.3.3 Supplementary Analysis Using Valence Scores and Subscales ..................... 157 6.4 Summary of Results .................................................................................................... 158 Chapter 7: Discussion and Conclusion ....................................................................................160 7.1 Introduction ................................................................................................................. 160 7.2 Discussion of Results .................................................................................................. 160 7.2.1 Sample 1: Neighbourhood Impacts of Kids (NIK) ................................................. 161 7.2.2 Sample 2: Teen Environment and Neighborhood (TEAN) .................................... 163 7.2.3 Sample 3: Senior Neighbourhood Quality of Life Study (SNQLS) ....................... 165 7.3 Limitations .................................................................................................................. 170 7.4 Importance of Studying Interactions for Planning and Policy Formulation ............... 173 7.5 Implications for Planning and Policymaking .............................................................. 176 7.5.1 Implications for Urban Design Guidelines ............................................................. 176 7.5.2 Implications for Form-Based Code ......................................................................... 179 7.5.3 Implications for Complete Street Guidelines .......................................................... 181 7.6 Implications for Canadian City Planners .................................................................... 183 xv  7.7 The Rationale for Investing in the Pedestrian Environment to Optimize Macro-level Walkability .............................................................................................................................. 183 7.8 Future Research .......................................................................................................... 184 7.9 Conclusion .................................................................................................................. 185 REFERENCES ...........................................................................................................................187 APPENDIX .................................................................................................................................226 Appendix A Survey Questionnaires........................................................................................ 226 A.1 Neighborhood Impact on Kids (NIK) Survey ......................................................... 226 A.2 Teen Environment and Neighborhood Study (TEAN) Survey ............................... 232 A.3 Neighborhood Quality of Life Study for Seniors (SNQLS) Survey ....................... 238 Appendix B Full model results ............................................................................................... 249 B.1 Sample 1: Neighborhood Impact on Kids (NIK) .................................................... 249 B.2 Sample 2: Teen Environment and Neighborhood Study (TEAN) .......................... 252 B.3 Sample 3: Neighborhood Quality of Life Study for Seniors (SNQLS) .................. 253 Appendix C Variance Inflation Factor for final models ......................................................... 259 C.1 Sample 1: Neighborhood Impact on Kids (NIK) .................................................... 259 C.2 Sample 2: Teen Environment and Neighborhood Study (TEAN) .......................... 261 C.3 Sample 3: Neighborhood Quality of Life Study for Seniors (SNQLS) .................. 261 Appendix D Robustness Check of Final MAPS Grand Score Models Using Different R- Packages .................................................................................................................................. 267 D.1 Sample 1: Neighborhood Impact on Kids (NIK) .................................................... 267 D.2 Sample 2: Teen Environment and Neighborhood Study (TEAN) .......................... 268 D.3 Sample 3: Neighborhood Quality of Life Study for Seniors (SNQLS) .................. 269 xvi  Appendix E Distribution of objective physical activity (MVPA) for NIK, TEAN and SNQLS datasets. ................................................................................................................................... 271  xvii  LIST OF TABLES Table 4-1 MAPS section descriptions. (Source: Millstein et al., 2013) ....................................... 96 Table 4-2: Destinations definition for MAPS survey. .................................................................. 97 Table 4-3. MAPS valence and subscales. ................................................................................... 102 Table 5-1 NIK: Descriptive summary of demography, psychosocial factors, outcome and other covariates. ................................................................................................................................... 113 Table 5-2. NIK: Bivariate relationship between neighbourhood walkability, MAPS scores, and other key variables based on active transport. ............................................................................ 114 Table 5-3. NIK: Bivariate relationship between walkability, MAPS Grand Score, and other key variables based on neighbourhood physical activity. ................................................................. 115 Table 5-4. NIK: Bivariate relationship between walkability, MAPS Grand Score, and other key variables based on non-school hours of daily moderate to vigorous physical activity (MVPA) measured by accelerometer. ........................................................................................................ 116 Table 5-5. TEAN: Descriptive summary of demography, psychosocial factors, outcome, and other covariates. .......................................................................................................................... 117 Table 5-6. TEAN: Bivariate relationship between neighbourhood walkability, MAPS grand score, and other key variables based on active transport. ........................................................... 119 Table 5-7. TEAN: Bivariate relationship between neighbourhood walkability, MAPS scores, and other key variables based on neighbourhood physical activity. ................................................. 120 Table 5-8. TEAN: Bivariate relationship between neighbourhood walkability, MAPS scores, and other key variables based on non-school hours of daily moderate to vigorous physical activity (MVPA) measured by accelerometer.......................................................................................... 121 xviii  Table 5-9. SNQLS: Descriptive summary of demography, psychosocial factors, outcome, and other covariates. .......................................................................................................................... 122 Table 5-10. SNQLS: Bivariate relationship between neighbourhood walkability, MAPS scores, and other key variables based on active transport. ..................................................................... 124 Table 5-11. SNQLS: Bivariate relationship between neighbourhood walkability, MAPS scores, and other key variables based on leisure walking. ...................................................................... 125 Table 5-12. SNQLS: Bivariate relationship between neighbourhood walkability, MAPS scores, and other key variables based on daily moderate to vigorous physical activity (MVPA) measured by accelerometer. ........................................................................................................................ 126 Table 6-1. Sample table showing the organization of table for reporting the MAPS grand and the domain specific MAPS score results. ......................................................................................... 129 Table 6-2. Sample table showing the organization of table for reporting the MAPS valence score and sub scale results (supplementary analysis). .......................................................................... 129 Table 6-3. NIK: Results of the regression model for objective MVPA ≥ 60 min/day during non-school hour using MAPS grand score. ........................................................................................ 131 Table 6-4: NIK: Results of the regression model for objective MVPA ≥ 60 min/day during non-school hour using MAPS leisure score. ...................................................................................... 134 Table 6-5. Results of the three-way interaction between walkability, parental support, and MAPS valence and subscales. ................................................................................................................ 136 Table 6-6. TEAN: Results of the regression model for objective MVPA ≥ 30 min/day during non-school hour using MAPS leisure score. ............................................................................... 139 Table 6-7. SNQLS: Results of the regression model for transport walking or biking using MAPS grand score. ................................................................................................................................. 143 xix  Table 6-8. SNQLS: Results of the regression model for transport walking or biking using MAPS active transport (MAPS AT) score. ............................................................................................ 146 Table 6-9. SNQLS: Results of the three-way interaction between walkability, gender, and MAPS valence and subscales for transport walking or biking. .............................................................. 148 Table 6-10. SNQLS: Results of the regression model for objective MVPA ≥ 30 min/day MAPS grand score. ................................................................................................................................. 152 Table 6-11. SNQLS: Results of the regression model for objective MVPA ≥ 30 min/day MAPS active transport score. ................................................................................................................. 155 Table 6-12. SNQLS: Results of the three-way interaction between walkability, neighbourhood income, and MAPS valence and subscales for objective physical activity. ............................... 157 Table 6-13. Summary of results. ................................................................................................. 159 xx  LIST OF FIGURES Figure 1-1. Regional scale, neighbourhood scale and pedestrian scale land use and transportation investments made by cities and regional governments. .................................................................. 4 Figure 1-2. Effect of neighbourhood walkability on transport walking with no interaction with the pedestrian environment ........................................................................................................... 10 Figure 1-3. Effect of neighbourhood walkability on transport walking with interaction with the pedestrian environment. ................................................................................................................ 11 Figure 1-4. Disordinal interaction between neighbourhood walkability and the pedestrian environment. ................................................................................................................................. 13 Figure 1-5. Synergistic interaction between neighbourhood walkability and the pedestrian environment. ................................................................................................................................. 13 Figure 1-6. Buffering interaction between neighbourhood walkability and the pedestrian environment. ................................................................................................................................. 14 Figure 1-7. Conceptual model showing the interaction between neighbourhood walkability and pedestrian environment. ................................................................................................................ 15 Figure 1-8. Conceptual model of the variation in interaction between neighbourhood walkability and pedestrian environment based on gender and income............................................................ 17 Figure 1-9. Conceptual model of the variation in the in interaction between neighbourhood walkability and pedestrian environment based on various psychosocial factors. ......................... 18 Figure 1-10. Map showing the locations of the study areas. ........................................................ 21 Figure 2-1. The conceptual framework of the ecological model. (Source: McLeroy, Bibeau, Steckler, & Glanz, 1988) .............................................................................................................. 25 Figure 2-2. Ecological model of physical activity (Source: Sallis et al., 2006). .......................... 27 xxi  Figure 2-3. Aerial view of a low-walkability neighbourhood. ..................................................... 34 Figure 2-4. Aerial view of a high-walkability neighbourhood. .................................................... 35 Figure 2-5. Street with poor pedestrian environment. .................................................................. 36 Figure 2-6. Street with good pedestrian environment. .................................................................. 37 Figure 2-7. Types of neighbourhoods based on walkability and pedestrian environment. .......... 38 Figure 2-8. Example of high neighbourhood walkability and poor pedestrian environment. ...... 39 Figure 2-9. Example of low neighbourhood walkability and good pedestrian environment........ 40 Figure 2-10. Example of low neighbourhood walkability and poor pedestrian environment. ..... 41 Figure 2-11. Example of high neighbourhood walkability and good pedestrian environment. ... 42 Figure 3-1 Conceptual model showing “walking” as a nexus between urban design, transportation planning, and public health. ................................................................................... 46 Figure 4-1. Street network buffer used to compute walkability index (Source: Frank et al., 2013)........................................................................................................................................................ 94 Figure 4-2 MAPS scoring structure and summary of inter-rater reliability: Route section. (Source: Millstein et al., 2013) ................................................................................................... 100 Figure 4-3 MAPS scoring structure and summary of inter-rater reliability: Segments and Crossings sections (multiple surveys per route). (Source: Millstein et al., 2013) ...................... 101 Figure 4-4. Data grouping plan for datasets and outcome of interest. ........................................ 108 Figure 4-5 Steps for testing interactions based on study objectives and hypotheses. ................. 108 Figure 6-1. Schematic diagram showing the organization of results. Result for each dataset is presented separately. ................................................................................................................... 128 xxii  Figure 6-2. Walkability index, MAPS grand score, and parental support for physical activity explaining the likelihood of spending at least 60 minutes doing MVPA daily during non-school hours. High and Low refer to values 1 standard deviation above and below the mean. ............. 132 Figure 6-3. Predicted probabilities of objective physical activity at different levels of walkability and MAPS grand scores for high and low parental support. High and Low refer to values 1 standard deviation above and below the mean. .......................................................................... 133 Figure 6-4. Walkability index, MAPS leisure score, and parental support for physical activity explaining the likelihood of spending at least 60 minutes doing MVPA daily during non-school hours. High and Low refer to values 1 standard deviation above and below the mean. ............. 135 Figure 6-5. Predicted probabilities of objective physical activity at different levels of walkability and MAPS leisure scores for high and low parental support. High and Low refer to values 1 standard deviation above and below the mean. .......................................................................... 135 Figure 6-6. Graphical representation of the interactions between neighborhood walkability, parental support and MAPS valence and sub scores. High and Low refer to values 1 standard deviation above and below the mean. ......................................................................................... 137 Figure 6-7. Neighbourhood walkability and MAPS grand leisure score interaction explaining objective physical activity (MVPA≥  30mins/day during non-school hour). High and Low refer to values 1 standard deviation above and below the mean. ........................................................ 140 Figure 6-8. Predicted probabilities for objective physical activity at different levels of walkability and MAPS grand scores. ............................................................................................................. 140 Figure 6-10. Predicted probabilities for transport walking or biking at different levels of walkability and MAPS grand scores for males and females....................................................... 144 xxiii  Figure 6-9. Neighbourhood walkability, MAPS grand score, and gender interactions explaining transport walking or biking. High and Low refer to values 1 standard deviation above and below the mean. ..................................................................................................................................... 144 Figure 6-11. Neighbourhood walkability, MAPS active transport score, and gender interactions explaining transport walking or biking. High and Low refer to values 1 standard deviation above and below the mean. ................................................................................................................... 147 Figure 6-12. Predicted probabilities for transport walking or biking at different levels of walkability and MAPS active transport scores for male and female. ......................................... 147 Figure 6-13. Graphical representation of the interactions between neighborhood walkability, gender and MAPS valence and sub scores for route. High and Low refer to values 1 standard deviation above and below the mean. ......................................................................................... 149 Figure 6-14. Graphical representation of the interactions between neighborhood walkability, gender and MAPS valence and sub scores for crossing. High and Low refer to values 1 standard deviation above and below the mean. ......................................................................................... 150 Figure 6-15. Interaction between neighbourhood walkability, MAPS grand score, and income for MVPA. High and Low walkability refer to values 1 standard deviation above and below the mean walkbility index. ................................................................................................................ 153 Figure 6-16. Predicted probabilities for MVPA ≥ 30min/day at different levels of walkability and MAPS grand scores for high- and low-income neigbourhoods .................................................. 154 Figure 6-17. Interaction between neighbourhood walkability, MAPS active transport score, and income for MVPA. High and Low walkability refer to values 1 standard deviation above and below the mean walkbility index. ............................................................................................... 156 xxiv  Figure 6-18. Predicted probabilities for MVPA ≥ 30min/day at different levels of walkability and MAPS grand scores for high- and low-income neigbourhoods .................................................. 156 Figure 6-19. Graphical representation of the interactions between neighbourhood walkability, income and MAPS valence and sub scores. High and Low walkability refer to values 1 standard deviation above and below the mean walkability index. ............................................................ 158 Figure 7-1. Relation between neighbourhood walkability, MAPS grand score in males and females for transport walking/biking. A case of no interaction between neighbourhood walkability, MAPS grand score and gender. .............................................................................. 175 Figure 7-2. Interaction between neighbourhood walkability, MAPS grand score and gender for transport walking/biking. A case of interaction between neighbourhood walkability, MAPS grand score and gender. .............................................................................................................. 175 Figure 7-3. Eyes on the street urban design guidelines used by Seattle. (Source: City of Seattle, 2013) ........................................................................................................................................... 177 Figure 7-4. Downtown design guidelines examples used by City of San Diego. (Source: City of San Diego) .................................................................................................................................. 178 Figure 7-5. Difference between how conventional zoning and form-based codes regulate urban form (Source: Form-Based Codes Institute, 2020). .................................................................... 179 Figure 7-6. Example of form based code to regulate public space in commercial side street. (Source: City of Palm Desert, 2017). .......................................................................................... 181 Figure 7-7. Example of complete street design guideline with focus on intersection design. .... 182 xxv  ACKNOWLEDGEMENTS  I would like to express my sincere appreciation to my supervisor, Dr. Lawrence D. Frank, for his constant support, encouragement to push myself, and willingness to provide me with resources to complete this piece of work. Dr. Frank has played a crucial role throughout my Ph.D. journey. Working on various projects at the Health and Community Design Lab with Dr. Frank provided me with a platform to see how academic research is applied in the real world. I would like to thank my dissertation committee members, Dr. Guy Faulkner, Dr. Mark Stevens, and Dr. Ying MacNab, for providing me with insightful comments and feedback as I worked on my research.  I would like to acknowledge Dr. James F. Sallis, Dr. Brian A. Saelens, and Dr. Abby King for allowing me to use the NIK, TEAN, and SNQLS data for my dissertation and for giving me constant feedback throughout my research. I would also like to thank other team members of the NIK, TEAN, and SNQLS studies, with whom I had a chance to interact during the regular calls throughout my research.  My special thanks go to Dr. Andy Hong for mentoring me during his time as a post-doc at the Health and Community Design Lab. I would also like to thank Dr. Jat Sandhu whom I had the opportunity to work with at the Vancouver Coastal Health Decision Support and learn a lot about various health-related data. Maureen Prentice and Heather van der Hoop also deserve a special acknowledgement for helping me polish my writing.  My family has been a great support and my lifeline throughout my journey to this terminal degree. I am grateful to have them around me in both high and low times. This work would not have been possible without their support.1  CHAPTER 1: INTRODUCTION 1.1 Introduction This dissertation's primary aim is to examine whether the health potential of neighbourhood walkability can be harnessed by providing a supportive pedestrian environment. Specifically, it examines whether there is a synergistic effect of neighbourhood walkability and pedestrian environment on physical activity in children, teens, and older adults. A synergistic effect or “synergism” occurs when two or more processes (or factors) interact so that their collective effect is greater than the sum of their separate effects (Myers, 1989, p. 506) . A recent study examined the effects of improving the pedestrian environment by constructing an urban greenway in downtown Vancouver, Canada; it found that people living closer to the greenway (≤ 300 meters) spent more time in daily moderate and vigorous physical activity and less time being sedentary (Frank et al., 2019). The study also found that the effect was attenuated as the distance from the greenway increased. Downtown Vancouver is a highly walkable neighbourhood, which gave the study a unique opportunity to examine the effect of pedestrian environment on physical activity holding neighbourhood level walkability constant. Though the study did not examine the synergistic effects, the distance decay effect shows the neighbourhood walkability’s effect on physical activity is synergistic with the pedestrian environment.  Conceptually, both neighbourhood walkability and the pedestrian environment collectively represent the overall walkability of a neighbourhood. However, they reflect different aspects of the neighbourhood environment. Neighbourhood walkability, though defined and measured in different ways (Shashank & Schuurman, 2019), reflects the conduciveness of a neighbourhood for walking (Cerin et al., 2007) and also reflects the accessibility of various destinations within walking distance (Carr et al., 2011). It can be measured as a function of the 2  mix of different land uses, commercial floor area, street connectivity, and residential density, which all impact pedestrians' travel and activity patterns (Frank, Sallis, et al., 2010). Unlike neighbourhood walkability, the pedestrian environment is composed of street design features that include various aspects, such as sidewalk materials and width, street facade design, crossing characteristics, curb qualities, street lighting, seating, trees, etc. (Maghelal & Capp, 2011). The street design features reflect the overall character of a street environment experienced by pedestrians. Various studies have examined the association of neighbourhood walkability and physical activity separately from the association of pedestrian environment and physical activity. Very few studies have examined the effect of the pedestrian environment on physical activity while controlling for neighborhood walkability, or vice versa. A comprehensive study of the synergistic effect of neighbourhood walkability and the pedestrian environment on physical activity is still lacking. This study fills the gap by examining whether neighbourhood walkability and the pedestrian environment synergistically affect physical activity. The results of the study will inform planners and policymakers about ways to harness neighbourhood walkability’s health potential by providing a supportive pedestrian environment.  1.2 Significance of Examining Synergistic Effects Synergy between two factors can be assessed using interaction approach (more discussion in section 1.7). Interaction effects show a combined effect of two variables on a third variable. As discussed by Van Der Weele & Knol (2014), one of the motivations for studying interactions is related to identifying subgroups that benefit most from an intervention when resources are limited. Cities and municipalities are often constrained by budgetary restrictions (Peck, 2012) and need to make decisions that give the highest return. Understanding the 3  interaction between neighbourhood walkability and the pedestrian environment is essential for planners and policymakers to make investment decisions cost-effective and maximize walkable neighbourhoods' health potential. However, interaction effects are not usually considered an attractive statistical method because of the complexity related to its understanding and interpretation (Ahlbom & Alfredsson, 2005; K. J. Rothman et al., 1974). Furthermore, due to a focus on establishing causal inference (Kärmeniemi et al., 2018; McCormack et al., 2004), interaction effects have not received enough attention in neighbourhood walkability and physical activity research.  1.3 Importance of the Study for Planners and Policymakers Municipal and metropolitan planning agencies invest a significant amount of their time and budget in developing various land-use and transportation infrastructure projects. These infrastructure decisions are a part of a broader long-range vision and growth strategy, which outlines how a city or a metropolitan region will look in 20 or 30 years. Neighbourhood-level characteristics such as residential density, a mix of various land uses, commercial density, and street connectivity are also part of such plans, focusing on creating communities where a resident can access services within walking distance (Figure 1-1). Because these plans are implemented over a long period, they offer the opportunity to be revisited throughout the implementation process. In the present context, we are facing a rapid shift in urban demography, and it is increasingly important for these long-range plans to accommodate the needs of ever-changing urban populations. 4  Figure 1-1. Regional scale, neighbourhood scale and pedestrian scale land use and transportation investments made by cities and regional governments.   The long-range land use and transport plans typically target a general population. The specific needs of various demographic sub-groups, such as children, teens, and older adults, and more vulnerable lower-income groups are unique and have different relationships with environments and place-based design features. The policies and plans formulated to build large-scale urban infrastructure often make it challenging for urban planners and policymakers to meet the needs of various vulnerable urban populations. However, small-scale plans such as streetscape revitalization, traffic-light improvement, and street crossing installations are easier to implement with minimal intervention. These interventions are less resource-intensive and can be 5  completed in a short time. Such smaller-scale interventions can help fine-tune larger-scale neighbourhood infrastructure and maximize the performance of the greater urban infrastructure.  Limited research to date has examined whether street design features enhance or attenuate the association of neighbourhood walkability and physical activity. However, the collective effect of neighbourhood walkability and street design features is implicitly discussed in various studies. For instance, people are more likely to walk a longer distance to use a transit station. Having pedestrian-supportive street design qualities en route to transit stations can also encourage people to walk longer distances (Park, Deakin, et al., 2015). Pedestrian-supportive street design qualities can also help enhance neighbourhood walkability (Adkins et al., 2012). Alternatively, poor street design qualities can diminish the walkability of a neighbourhood, which may otherwise be “walkable” based on neighbourhood-level built environment features (Neckerman et al., 2009). Very little research focused on the synergistic associations between built environment characteristics and their collective effect on physical activity. However, environments are combinations of factors that impact behaviour at a range of scales. Evidence from Cain et al. (2014) shows that pedestrian design features have an additive impact on walking beyond land use and other factors, known as macro-level walkability. This dissertation evaluates the synergistic effects of macro walkability and micro-level pedestrian environment factors on physical activity.  In addition to the built environment factors, various socio-demographic and psychosocial factors also play significant roles in a person’s likelihood of engaging in physical activity. For instance, studies have found women are overall less physically active than men (Y. S. Lee, 2005; Poitras et al., 2016). Women have a longer life expectancy than men, which puts them at higher risk in old age of enduring motor-function problems from sustained physical inactivity.  Various 6  psychosocial factors like social support, self-efficacy, and safety perception are even more important for women.  For example, older adults are more likely to be physically active if they have a walking partner and a supportive built environment (J. A. Carlson et al., 2012). Likewise, active transportation to school is higher in teens who have higher self-efficacy and higher neighbourhood walkability, as well as higher accessibility to neighbourhood parks and recreation facilities (X. Wang et al., 2017).  1.4 The Rationale for Focusing on Children, Teens, and Older Adults Despite the documented evidence suggesting the many health benefits of physical activity, a significant segment of the population does not meet the recommended physical activity levels. Participation in physical activity at an early age plays a vital role in determining lifelong physical activity patterns (Faulkner et al., 2009). However, there is a decreasing trend in physical activity participation among children and teens (Tremblay et al., 2014). Likewise, an increase in the ageing population with longer life expectancy has made it challenging for public health practitioners to ensure the older population is living a healthy and physically active life.  Due to their mobility patterns, children, teens, and older adults are susceptible to the constraints and opportunities provided by various features of neighbourhood built environment. This sensitivity can significantly affect their mobility patterns and, therefore, their physical activity. Research has shown that neighbourhood walkability is positively associated with physical activity in children, teens (Ding et al., 2011; Sallis et al., 2000; Sallis & Glanz, 2006), and older adults (A. C. King et al., 2011). Travel behaviour and physical activity in children, teens, and older adults vary significantly from that of adults because of their unique needs and mobility patterns. Children and teens are more likely to be active during their time in the local 7  neighbourhood (J. A. Carlson et al., 2016; S. Kneeshaw-Price et al., 2013). Their mobility patterns are largely influenced by their parent’s travel behaviour and attitudes. Similarly, older adults are more likely to spend most of their time in their neighbourhood, make few trips to destinations far from their neighbourhood and are more likely to walk for errands (Chudyk et al., 2015). This places children, teens, and older adults at high risk of being physically inactive because of the constraints posed by neighbourhood built environment features (Dagkas & Stathi, 2007; P. Gordon-Larsen et al., 2000; Sun et al., 2013; Tremblay et al., 2014). Therefore, it is crucial to identify what features of the built environment support and what factors constrain physical activity in these age groups.  Being physically active is crucial for children, teens, and older adults for various reasons. Physical activity in children has significant long-term health benefits. Physical activity in teens is particularly vital since teens are in a transitional phase to early adulthood, and any behaviour at this stage can have a lifelong impact (Malina, 2001). The decrease in physical activity is among various health behaviour changes that happen during this transitional phase (Kwan et al., 2012). Therefore, this life stage is crucial for developing a physically active lifestyle. Because of deteriorating physical functioning, older adults are at high risk of losing mobility independence as they age; this loss may create mobility impairment, resulting in low physical activity and poor physical health. Despite these vulnerabilities, there remain significant gaps in the evidence base linking the built environment and physical activity in these age groups.  Most of the research that has examined the relationship between the built environment and physical activity has focused on adults (Cho & Rodríguez, 2015; Handy, 1992; Næss, 2011; Nasri & Zhang, 2014). The findings of these studies have consistently been focused on how neighbourhood walkability relates to physical activity.  Fewer studies have considered how 8  regional accessibility or pedestrian environment features predict physical activity. Though this may be relevant for adults, such a pattern might not be accurate in children, teens, and older adults. Children and teens are not as exposed to the wide range of environments as adults and are likely to spend more of their time in their neighbourhood. Similarly, older adults are also more likely to spend most of their time near home in their neighbourhood because of their shrinking activity space. Therefore, the specific components of the pedestrian environment might play a crucial role in influencing physical activity in children, teens, and older adults.  1.5 Policy Implications This dissertation addresses three critical policy questions related to the built environment and physical activity among children, teens, and older adults.  First, it addresses whether neighbourhood walkability and street design features synergistically affect physical activity. In other words, how do neighbourhood walkability and street design features work collectively to support physical activity? From a fiscal and logistical standpoint, street-scale plans and policies are less demanding than neighbourhood-level walkability and regional plans and strategies; street design features can be implemented at a lower cost. This factor is crucial for formulating plans and policies for creating physical activity-friendly environments in areas where modifying the built environment at a macro level requires significant resources and time. One of the direct urban planning applications of this research is the ability to inform Urban Design Guidelines, including Form-Based Zoning codes used by cities to guide the design, form, and character of the pedestrian environment.  Second, the research addresses how neighbourhood design relates to physical activity and downstream health outcomes for different age groups. Addressing this question is particularly 9  important to find out what works and what does not work in creating physical activity-friendly communities for children, teens, and older adults.  Third, this study assesses how psychosocial factors impact the relationship between neighbourhood built environment and physical activity. This research helps understand how the relationship between neighbourhood built environment and physical activity varies across various individual psychosocial factors.  1.6 Aim The aim of this dissertation is as follows:  To examine the extent to which neighbourhood walkability’s health potential can be “harnessed” by providing a supportive pedestrian environment.  It focuses on three vulnerable age groups: children, teens, and older adults. In addition, this study examines whether the potential of harnessing neighbourhood walkability varies based on demographic and psychosocial factors. This study explores the Synergy between neighbourhood walkability and the pedestrian environment to achieve this aim. As mentioned earlier, Synergy occurs when two or more processes (or factors) interact so that their collective effect is greater than the sum of their separate effects (Myers, 1989, p. 506). Synergy is assessed using a series of regression models that use interaction effects models, which are described in the following section.  1.7 Interaction Effects to Assess Synergy Synergy can be viewed as one of the various patterns of interaction between two variables. When there is an interaction effect, the effect of one variable (X) on an outcome variable (Y) depends on the value of a third variable (Z). In a case of no interaction, the degree of effect of X on Y would not be affected by different levels of Z; the effect would be similar 10  across the different levels of the third variable (Z). For example, if there is an interaction between neighbourhood walkability and the pedestrian environment, the effect of the neighbourhood walkability on walking behaviour will depend on the pedestrian environment. This means that people living in a neighbourhood with high walkability respond to the pedestrian environment differently than people living in a neighbourhood with a similar pedestrian environment but low walkability. If there is no interaction between neighbourhood walkability and the pedestrian environment, the effect of the neighbourhood walkability on physical activity will be increased or decreased by a fixed degree due to the pedestrian environment. This means that the pedestrian environment may increase or decrease the effect of the neighbourhood walkability on physical activity. Still, the increase or decrease in effect would remain the same across all levels of neighbourhood walkability.                                                                                                                                                                                      Figure 1-2. Effect of neighbourhood walkability on transport walking with no interaction with the pedestrian environment 11  Figure 1-2 shows the relationship between pedestrian environment and transport walking at different levels of neighbourhood walkability. The x-axis shows pedestrian environment (high meaning good and low meaning poor), and the y-axis shows transport walking. The blue and orange lines show high and low walkability neighbourhoods, respectively. When there is no interaction, the blue and orange lines are parallel. However, if there is an interaction between neighbourhood walkability and pedestrian environment, the effect of neighbourhood walkability on transport walking will vary by the levels of pedestrian environment. This effect is shown by the two lines (blue and orange) in Figure 1-3. The two lines show that even with the same level of neighbourhood walkability, the level of transport walking is different based on pedestrian environment. The figure shows that neighbourhood walkability positively affects transport walking in neighbourhoods with good and poor pedestrian environments. However, the effect of neighbourhood walkability is higher in a neighbourhood with a good pedestrian environment compared to one with low walkability.                                                                                                                                                                                Figure 1-3. Effect of neighbourhood walkability on transport walking with interaction with the pedestrian environment. 12  Interaction effects can follow various patterns. Cohen, Cohen, West, & Aiken (2013) categorized interaction effects into two categories: i) disordinal interaction and ii) ordinal interaction. In a disordinal interaction (Figure 1-4), the effect of the key predictor (neighbourhood walkability) on the dependent variable (transport walking) is in different directions for different levels of the moderator (pedestrian environment), creating a cross-over effect. In ordinal interaction, the moderator either increases the effect or dampens the effect of the various groups of the key predictor on the dependent variable. Depending on the outcome, the ordinal interaction is one of two sub-types: i) synergistic interaction and ii) buffering interaction. ‘A synergistic interaction effect (Figure 1-5) occurs when a change in the level of the moderator variable, i.e., pedestrian environment, enhances the relationship between neighbourhood walkability and transport walking. A buffering interaction effect (Figure 1-6) occurs when a change in the level of pedestrian environment reduces the magnitude of the relationship between transport walking and neighbourhood walkability. Among these patterns of interactions, synergistic interaction is more intuitive and reflects the potential for harnessing the health potential of neighbourhood walkability by providing a supportive pedestrian environment. While this research focuses on identifying synergy, it does not overlook the fact that other patterns of interactions are also possible.  13  Figure 1-4. Disordinal interaction between neighbourhood walkability and the pedestrian environment.   Figure 1-5. Synergistic interaction between neighbourhood walkability and the pedestrian environment.                                                                                                                                                                                                                                                                                                                                                                                           14  Figure 1-6. Buffering interaction between neighbourhood walkability and the pedestrian environment.   1.8 Objectives, Hypotheses, and Rationale 1.8.1 Objective 1 To examine whether there is an interaction between neighbourhood walkability and the pedestrian environment in explaining walking and physical activity in children, teens, and older adults. 1.8.1.1 Hypothesis Street design qualities and neighbourhood walkability interact to affect physical activity. The interaction effect is most synergistic when both street design qualities and neighbourhood walkability are high (Figure 1-7).                                                                                                                                                                                            15  Figure 1-7. Conceptual model showing the interaction between neighbourhood walkability and pedestrian environment.  1.8.1.2 Rationale The disparity in pedestrian environment types can attenuate the positive effect of high neighbourhood walkability (Neckerman et al., 2009). Likewise, the lack of a favourable pedestrian environment can make a neighbourhood less conducive for walking (Thornton et al., 2016). This is particularly important around schools with poor pedestrian environments but with highly favourable neighbourhood-level conditions for walking (Zhu & Lee, 2008). Well-designed street facilities, along with features like separation from vehicle traffic and network 16  connectivity, can make a predominantly residential neighbourhood conducive for walking (Adkins et al., 2012). Studies have also found that design features of pedestrian environments can explain more variation in walking behaviour compared to macro-level features such as neighbourhood walkability (Foltête & Piombini, 2007). Thus, it is highly possible that pedestrian environment features and neighbourhood walkability can synergistically affect physical activity.  1.8.2 Objective 2 To examine whether the interaction between neighbourhood walkability and the pedestrian environment to explain walking and physical activity varies by gender and income across the three age groups.  1.8.2.1 Hypothesis The interaction between neighbourhood walkability and pedestrian environment is more synergistic in males than females and is synergistic in high-income neighbourhoods compared to low-income neighbourhoods (Figure 1-8).  1.8.2.2 Rationale Several studies have found that the association between neighbourhood walkability and physical activity varies by gender and income. Women are less physically active than men (Y. S. Lee, 2005; Poitras et al., 2016) and are less responsive to their built environment compared to men. Likewise, people living in low-income neighbourhoods are also less likely to be physically active. Therefore, among people living in neighbourhoods with high walkability and good pedestrian environments, men will show a higher level of physical activity than women, and high-income neighbourhood residents will show a higher level of physical activity than the residents in low-income neighbourhoods.  17  Figure 1-8. Conceptual model of the variation in interaction between neighbourhood walkability and pedestrian environment based on gender and income.   1.8.3 Objective 3 To examine whether the interaction between neighbourhood walkability and the pedestrian environment to explain walking and physical activity varies by parental/family support, self-efficacy, and safety perception of the three age groups.  18  1.8.3.1 Hypothesis 1 In children and teens, the interaction between street design qualities and neighbourhood walkability is synergistic when their parents support them to do physical activity, whereas, in older adults, interaction is synergistic when their self-efficacy and social support are high (Figure 1-9).  Figure 1-9. Conceptual model of the variation in the in interaction between neighbourhood walkability and pedestrian environment based on various psychosocial factors.   19  1.8.3.2 Rationale Studies have shown that parents act as role models for physical activity for children and teens. A higher level of physical activity in parents is positively associated with physical activity in children (Edwardson & Gorely, 2010). Likewise, support from parents has also been found to have a significant positive association with children’s physical activity (Mendonça et al., 2014; Pyper et al., 2016; Zecevic et al., 2010). Therefore, children living in a neighbourhood with high walkability and good street design features will have an even higher level of physical activity if their parents highly support their children to be physically active.  Self-efficacy has been found to moderate the relationship between neighbourhood walkability and physical activity in older adults. Even when their neighbourhoods have the same level of walkability, older adults with higher self-efficacy and better social support spend more time engaging in moderate and vigorous physical activity compared to those with lower self-efficacy and social support (J. A. Carlson et al., 2012). These findings suggest that self-efficacy and social support are potential moderators of physical activity and neighbourhood walkability relationship in older adults.  1.8.3.3 Hypothesis 2 In children and teens, the interaction between street design qualities and neighbourhood walkability is synergistic when their parents have a high perception of neighbourhood safety, whereas, in older adults, the interaction is synergistic when their perception of neighbourhood safety is high (Figure 1-9).  20  1.8.3.4 Rationale Parents’ perceptions of neighbourhood safety can affect their children’s engagement in physical activity. A higher perceived safety risk by parents could constrain physical activity participation in children and teens (Carver et al., 2008; Foster et al., 2014).  Likewise, older adults are less likely to participate in physical activity because of safety concerns. Concerns about traffic safety, pedestrian safety, and personal safety can act as barriers to engaging in physical activity. Therefore, safety concerns can moderate the synergistic effect of street design qualities and neighbourhood walkability on physical activity.  1.9 Research Context  This study used data from three studies conducted in three different parts of the United States (Figure 1-10) Baltimore, Maryland-Washington, DC; Seattle/King County, Washington; and San Diego County, California. The three studies are the Neighborhood Impact on Kids (NIK) study, the Teen Environment and Neighborhood (TEAN) study, and the Senior Neighborhood Quality of Life Study (SNQLS). The NIK study was conducted between September 2007 and January 2009 in Seattle/King County, Washington, and San Diego County, California. It examined the impact of neighbourhood environment on physical activity and eating behaviour in children aged 6–11. The TEAN study was conducted between 2009 and 2011 in Baltimore, Maryland-Washington, DC, and Seattle/King County, Washington. TEAN was an observational study of neighbourhood environment and physical activity patterns in teens between ages 12 and 16. The SNQLS is another study on older adults that provided data for this research. The SNQLS is also an observational study that was conducted in the Seattle/King County, Washington, and Baltimore-Washington DC regions of the U.S. between 2005 and 2008. The primary aim of the SNQLS was to examine the relationship between neighbourhood 21  walkability and various health outcomes among residents living in neighbourhoods that differed in walkability and income. Appropriate institutional review boards approved all the studies.   Various publications have come out of these studies.1 Though these studies have broader goals and objectives, some of the findings from these studies are relevant to this research. For instance, Cain et al. (2014) used the data from the three studies to examine the contributions of street-scale built environment features to physical activity (transport-related, leisure and neighbourhood, objective) in four age groups: children, teens, adults, and older adults. They  1 The list of publications can be found on this website: http://sallis.ucsd.edu/publications.html  Figure 1-10. Map showing the locations of the study areas. 22  found that street-scale built environment features are associated with physical activity across all four age groups. Though their study controlled for neighbourhood walkability, it did not examine the interaction between neighbourhood walkability and street design qualities. Another study by Wang et al. (2017)examined the interacting associations of psychosocial and built-environment variables with adolescents’ active transportation using data from TEAN. Their study found a positive interaction between neighbourhood walkability and self-efficacy, and between neighbourhood recreational facilities (parks and recreational facilities) and self-efficacy. Wang et al. (2017) also examined the interaction between street-scale built environment features and psychosocial factors. They found the barrier for physical activity moderates the relationship between street design features and active transport to school. Wang et al. (2017) examined interaction in separate models, i.e., using different models for neighbourhood walkability, neighbourhood recreational facilities and street design qualities. Similar results were found in the SNQLS study by Carlson et al. (2012). They looked at the interaction between psychosocial and neighbourhood built environment factors (only neighbourhood recreation environment and neighbourhood walkability) in explaining older adults’ physical activity. Though all these studies have examined the interaction between neighbourhood built environment features and various psychosocial factors, a comprehensive analysis of the pattern of interaction between neighbourhood walkability and street design features is still lacking. By comprehensively analyzing the pattern of interaction between neighbourhood walkability and street design features in explaining physical activity in children, teens, and older adults, this study covers the areas not addressed in the previous studies.  23  1.10 Organization of the Dissertation The rest of the dissertation is organized into six chapters. Chapter 2 provides a theoretical background for this dissertation by discussing the socio-ecological model of health behaviour, the interaction between the built environment, demographic and psychosocial factors. It also provides examples of different cases showing a combination of neighbourhood walkability and pedestrian environment. Chapter 3 reviews the literature on the built environment and walking. The chapter starts by discussing the research on the built environment and walking from an urban design, transportation planning and public health perspective, which is then followed by literature on walking as physical activity and physical and social correlates of physical activity in teens, children and older adults. Chapter 4 discusses the method used in the dissertation. The chapter begins by describing the data sources, study design to collect the data, variables used in this study, and analytical approach. Chapter 5 provides descriptive statistics summarizing the socio-demographic characteristics of the study participants. It also provides the bivariate relationship between dependent and independent variables. The inferential statistics are presented in Chapter 6. Finally, Chapter 7 discusses the results of the statistical models and provides some examples of physical activity patterns in neighbourhoods with different levels of walkability and pedestrian environment. It also discusses the limitation of the research, implication for planning and policymaking, and implications for future research.  24  CHAPTER 2: THEORETICAL BACKGROUND  2.1 Overview This chapter provides the theoretical background that underpins the proposed research. It does so in four sections. First, it discusses the ecological model of health behaviour. One of the central tenets of the ecological model is the multilevel range of effects of the various built environment and psychosocial factors. Second, it discusses the synergistic role of neighbourhood walkability and pedestrian environment on walking and physical activity. It then expands to the broader theoretical framework of the ecological model by discussing the interaction between built environment features and various psychosocial factors on physical activity. The final section discusses the different neighbourhoods (or place types) based on the combination of neighbourhood walkability and the pedestrian environment.  2.2 Socio-Ecological Model of Health Behaviour Ecological models provide a guiding framework for addressing health problems at the population level. Ecological models assume that behaviour has multiple levels of influence, ranging from the intrapersonal level (e.g., biological, psychological) to the interpersonal level (e.g., social, cultural) to broader levels of community and public policy. Different strands of work contributed to developing the ecological model (Sallis et al., 2008), such as environmental psychology, social ecology, systems theory, operant theory, and social cognitive theory.   25   Early work on an ecological model of behaviour was done by Bronfenbrenner (Bronfenbrenner, 1993), who focused on the neighbourhood factors that affect children’s development. McLeroy, Bibeau, Steckler, and Glanz (1988) later adapted Bronfenbrenner’s ecological model to study health behaviour by concentrating on various layers of factors that affect health behaviour. They argued that five factors influence the patterns: i) intrapersonal factors, ii) interpersonal processes and primary groups, iii) institutional factors, iv) community factors, and v) public policy (Figure 2-1). Intrapersonal factors include the characteristics of individuals, such as attitudes, self-concept, knowledge, etc. The interpersonal process and primary groups include families, workgroups, and friend circles, which can be formal and informal. Institutional factors include institutions that have their own rules and regulations; community factors include linkages between institutions and organizations; and public policy includes local, state, and national laws and policies. Incorporating the various levels of the environment, ecological models are distinct from other behavioural models that limit their focus to the individual characteristics and proximal influences and do not consider the broader community, organizational, and policy setting within which behaviours occur (Flay & Petraitis, Figure 2-1. The conceptual framework of the ecological model. (Source: McLeroy, Bibeau, Steckler, & Glanz, 1988) 26  1994). Such a view of health promotion has significant implications not only for theory development and basic research but also for public policy, community intervention, and program evaluation (Stokols, 1992; Stokols et al., 2003). Four core concepts of ecological models are essential (Sallis et al., 2008) for research and practice. The first concept argues that behaviours are influenced by factors at multiple levels, such as intrapersonal, interpersonal, community, and policy levels. The second concept contends that the factors influencing behaviours interact across levels, which implies that these factors act together. The third concept proposes multilevel intervention for the most effective behaviour change. The fourth concept states that ecological models are best used when the behaviour being studied is specific (Sallis et al., 2008). Based on these core concepts, researchers from the fields of psychology, transportation planning, economics, and environmental planning have synthesized findings and proposed an ecological model for creating active living communities (Sallis et al., 2006). This model (Figure 2-2) provides a framework for studying physical activity by adding three distinguishing features to the general ecological model. First, the model specifies four physical activity domains; this reflects that physical activity occurs in various contexts. Second, some influences like mass media can influence behaviour irrespective of the settings where behaviours take place. Third, the social and cultural environments influence behaviour at multiple levels.  Several studies have used the ecological approach to study physical activity. The findings from these studies support the core concepts of the ecological model of physical activity. For instance, a selective review of studies looking at the environmental correlates of physical activity argues that the relationship can be better understood when the environmental measures closely match the behaviour of interest and the setting in which the behaviour takes place (Giles-corti et 27  al., 2005). The ecological model is also used in some studies to examine the relative influence of individual, social, and physical environment determinants of physical activity (Bolívar et al., 2010; Giles-Corti & Donovan, 2002). Likewise, some studies have also examined the interactive effects of the physical environment and psychosocial attributes of physical activity (Ding, Sallis, et al., 2012). Though a comprehensive analysis of the ecological model is still lacking, the model provides a robust framework for understanding physical activity (Sallis et al., 2008). Figure 2-2. Ecological model of physical activity (Source: Sallis et al., 2006).  28  2.3 Synergistic Effects of Multiple Macro-Built Environment Factors on Walking Behaviour  Studies have acknowledged that the features of the built environment collectively relate with (Cervero & Kockelman, 1997; Ewing & Cervero, 2010; Saelens & Handy, 2008) and affect (Frank et al., 2019) walking behaviour. The famous “Ds” of urban planning—density, diversity, design, destination accessibility and distance to transit are effective when they act collectively (Cervero & Kockelman, 1997; Ewing & Cervero, 2010). Living in areas with high density, higher land-use diversity, and functional street connectivity is found to be associated with more walking. However, to date, little work has been done to directly assess how macro-level walkability and micro-level pedestrian design features interact and how potentially synergistic relationships impact travel patterns.  This is particularly important when considering policy implications where investing in pedestrian design features in areas where walkability is high may yield considerable payback in shifting travel patterns and increasing walking.  However, these features of the built environment usually occur jointly and are highly correlated, creating methodological challenges to examine interaction. Therefore, composite indices are commonly used to examine the collective effect, e.g., the walkability index (Frank, Sallis, et al., 2010). The variables used in such indices are based on prior literature, and various studies have validated their usefulness. The walkability index, a composite index of residential density, intersection density, land-use mix, and floor area ratio within a 1-km street network buffer, is consistently associated with walking behaviour.  It can be argued that if the built environment features collectively contribute to enhancing the built environment's walkability, it is redundant to stratify the built environment features at different spatial scales and study the interactions of the same concept of walkability measured at 29  the different spatial levels. The first part of the argument is conceptually correct as it is in line with the existing literature. However, the second part of the argument has some flaws from the planning and policy perspectives. Macro-scale built environment characteristics, such as regional transit accessibility, are related to long-term plans and policies at the regional level. Neighbourhood walkability, which reflects the meso-scale built environment characteristics, is relevant for community plans and policies commonly formulated by cities. Street-scale design features, often referred to at the micro-scale built environment features, reflect the walkability at a much finer scale. Street-scale design features are often part of short-term neighbourhood plans and policies. The regional plans and policies at metropolitan levels are not always legislative or mandatory. They lack enough power to control the built environment. In contrast, plans and policies at the city and neighbourhood levels have more power to control the built environment at the neighbourhood and street level (Heath et al., 2006; Steinmetz-Wood et al., 2020). As mentioned before, from a fiscal and pragmatic perspective, street-scale plans and policies are less demanding compared to neighbourhood-level and regional plans and policies; street-scale plans can also be implemented at a lower cost. Therefore, by examining the interaction of the built environment features measured at the different spatial scales, urban planners can identify significant street design features that can help effectively increase the walkability of the built environment at a lower cost. A high correlation between built environment features poses methodological challenges in assessing the interactions between them. However, evidence supports traces of interaction between built environment features. For instance, suitable street design qualities on routes to transit stations can increase the walking distance in transit-oriented development (Park, Choi, et al., 2015; Park, Deakin, et al., 2015). Mixed land use and land-use diversity around transit 30  stations have been associated with higher levels of walking (Huang et al., 2017). People are more likely to walk even in a predominantly highly residential neighbourhood if there are well-designed street facilities such as separation from vehicular traffic and street connectivity (Adkins et al., 2012). Modifying the street by adding pedestrian infrastructure can enhance walkability, thereby increasing walking (Jensen et al., 2017). In addition to positive interaction, also called synergy, between built environment features, studies have found that having a poor pedestrian environment can reduce the positive effect of high neighbourhood walkability (Lovasi et al., 2009; Neckerman et al., 2009). Such a problem is common in low-income communities and communities with racial minorities, which are highly walkable but lack appropriate pedestrian environments (Thornton et al., 2016). Such conditions put children at higher risk of being physically inactive (Zhu & Lee, 2008). These studies suggest that built environment features interact with each other while explaining walking behaviour and physical activity. Moreover, they present substantial evidence of a synergistic relationship between built environment characteristics.  Recently, researchers have been increasing interest to examine synergy between built environment features using data reduction techniques. Latent class analysis is one of the statistical approaches used to create discrete profiles using specific numbers of variables. Each profile contains a range of values from the variables used to create it, e.g., one profile may have one variable with high values and another variable with low values; other profiles may have high values for all variables. Such profiles indirectly mimic complex interactions involving multiple variables that are correlated with each other (Lanza & Rhoades, 2013). Neighbourhood profiles can be created using a combination of various factors like residential density, land-use diversity, and intersection density. Multiple studies have used such an approach to find a relationship 31  between neighbourhood profiles and walking behaviour (Hobbs et al., 2018; Timperio et al., 2017; Todd et al., 2016). For instance, a study found that people living in a neighbourhood with high walkability, high transit accessibility, and high numbers of recreational facilities have higher physical activity than those living in a neighbourhood with low walkability, low transit accessibility, and low numbers of recreational facilities (Todd et al., 2016). Though these methods provide an efficient tool to model the cumulative effects of various built environment features, they have some limitations. Interpreting results from data-driven methods can be challenging for policymakers because it is hard to isolate the effects of individual components within each neighbourhood typology (Frank, Sallis, et al., 2010). Thus, the magnitude of interaction cannot be assessed directly. Because data reduction techniques are highly dependent on the data distribution, neighbourhood profiles created from such an approach may not be readily generalized. More importantly, most of the data reduction techniques have focused on features that determine neighbourhood walkability.  Though there is usually a high correlation among built environment features such as density, commercial floor area, and land-use mix, which determine neighbourhood walkability, some of the key built environment characteristics are not necessarily highly correlated. These features include regional accessibility, neighbourhood walkability, and street design qualities. A neighbourhood with high regional accessibility may not have enough land-use diversity, residential density, or intersection density to be highly walkable. “Park and ride” is a typical example of such a condition. Likewise, an area with high neighbourhood walkability may not always have attractive street design qualities. A combination of such situations can have a significant impact on walking and physical activity. As discussed before, few studies have examined the effects of such a combination of neighbourhood-level built environment features 32  on walking and physical activity using data reduction techniques. However, a comprehensive study of the synergistic relationship among these built environment characteristics has not been evaluated. In addition, the effects of such phenomena on vulnerable populations like children, teens, and older adults have not received enough attention.  2.4 The Interactions Between the Built Environment and Psychosocial Factors Though built environment features synergistically affect walking behaviour and physical activity patterns, they do not tell the whole story about factors that affect walking behaviour (Reyer et al., 2014). In addition to the effect of the built environment, socio-demographic and psychosocial factors also play a significant role. The built environment has a differential association with physical activity based on income, gender, and ethnicity (Richardson et al., 2017). Psychosocial factors such as self-esteem, self-efficacy, and parental and social support have also been found to moderate the relationship between the built environment and physical activity. Beenackers, Kamphuis, Mackenbach, Burdorf, and Lenthe (2013) proposed two possible interaction mechanisms when evaluating the interaction between the built environment and psychosocial cognitions to explain physical activity. Psychosocial cognitions toward physical activity include attitude, self-efficacy, intention and social influence towards physical activity. The first mechanism states that the built environment is less critical for the decision to walk for those who have more positive psychosocial cognitions toward physical activity. When this interaction exists, people with less positive psychosocial cognitions will benefit more from a supportive environment (Beenackers et al., 2013). For instance, people who are less inclined to be active due to psychosocial factors like social support, barriers, and attitudes may walk more because of improved neighbourhood walkability (Ding, Sallis, et al., 2012; Van Dyck et al., 33  2009). The other mechanism assumes a synergy between environmental factors and psychosocial cognitions; the environment is more critical in the decision to walk for people with more positive cognitions. This means that the beneficial effects on walking of having positive psychosocial cognitions and living in a stimulating environment would strengthen each other. For instance, having a walking partner and a supportive built environment may be particularly useful in facilitating older adults’ physical activity (J. A. Carlson et al., 2012). Likewise, active transportation to school among adolescents may be higher if the neighbourhood built environment and psychosocial factors are higher (X. Wang et al., 2017). Therefore, these interactive mechanisms highlight psychosocial factors that modify the relationship between the built environment and physical activity.  2.5 Place Types Based on Neighbourhood Walkability and the Pedestrian Environment 2.5.1 Place Types Based on Neighbourhood Walkability Neighbourhood walkability is a mix of four components that capture various aspects of local accessibility. The components are: i) residential density, ii) commercial floor area ratio, iii) street connectivity, and iv) land-use mix. A neighbourhood need not have the highest scores for each of these components to qualify as a walkable neighbourhood. It is therefore important to maintain a balance across all components of walkability to make neighbourhoods more walkable. Although it may be harder to define various place types based on the different combinations of walkability components, a clear distinction can be made between places that are at opposite ends of the walkability spectrum.  34  2.5.1.1 Low-Walkability Neighbourhoods Low-walkability neighbourhoods tend to be designed with low residential density with predominantly single-family homes, nonexistent or widely spread apart commercial land uses (e.g., big box stores in large parking lots), a low degree of street connectivity (larger block sizes, many cul-de-sacs), and a homogenous land-use mix, i.e., residential areas separated from commercial regions separated from recreational areas. Low-walkability neighbourhoods mostly include ex-urban areas, but areas within a city with predominantly single-family land use also fall into this category (Figure 2-3).   2.5.1.2 High-Walkability Neighbourhoods High-walkability neighbourhoods generally have a more compact urban form characterized by medium to high residential densities such as townhouses and apartments, a Figure 2-3. Aerial view of a low-walkability neighbourhood.  35  concentration of nearby commercial (retail) establishments, a functional mix of land uses, and a high degree of connectivity (i.e., small block sizes, higher number of intersections). Walkable neighbourhoods can include neighbourhoods with high-density mid-rise building forms with maximum mixed uses and a highly connected street network and neighbourhoods with high-density high-rise buildings with vertically mixed uses and good street connectivity. These kinds of neighbourhoods provide optimal environments for walking. They tend to be present in and around major urban centers and downtown areas (Figure 2-4) and some historic neighbourhoods designed in the pre-World War I era.    2.5.2 Place Types Based on the Pedestrian Environment Unlike the components of neighbourhood walkability, the pedestrian environment reflects different aspects of the street environment. Such components can be broadly categorized into Figure 2-4. Aerial view of a high-walkability neighbourhood. 36  three sections: i) route-level components, ii) segment-level components, and iii) crossing-level components. Within these sections, various design features are included.  2.5.2.1 Neighbourhoods with Poor Pedestrian Environment Neighbourhoods with a poor pedestrian environment often have less pedestrian-friendly streets. The streets in such neighbourhoods usually have fewer commercial shops, restaurants, banks, etc. There may be a poor connection between streets and buildings, with larger parking stalls in front of street buildings. The buildings may not be well maintained and are often vandalized by graffiti and litter. Sidewalks may not be present or may be minimally present with bumps and few or no streetlights, benches, or trees. The crossings and intersections may lack markings, curb cuts, or proper signage, making them less useful for pedestrians. The most obvious examples of such neighbourhoods are the suburban neighbourhoods built after World War II, which were oriented toward the automobile (Figure 2-5). However, neighbourhoods closer to or in downtown areas can also have streets that are not pedestrian-friendly.  Figure 2-5. Street with poor pedestrian environment.  37  2.5.2.2 Neighbourhoods with Good Pedestrian Environment Neighbourhoods with a good pedestrian environment tend to have streets with more pedestrian-friendly features. Such streets can include various destinations to walk to, an appealing street elevation with a good connection between the inside and outside of the buildings, vibrant building colours, a good social environment, appropriately designed sidewalks, and crossings and intersections with good streetlights, benches, and trees. Neighbourhoods with the good street design are often in urban centres (Figure 2-6) but can also be found in new suburban developments and destination retail centres like outlet malls.  Figure 2-6. Street with good pedestrian environment.  38  2.5.3 Place Types Based on Neighbourhood Walkability and Pedestrian Environment Neighbourhood walkability and the pedestrian environment reflect different aspects of a neighbourhood’s conduciveness for walking. The combination of these two aspects creates a different environment for pedestrians. For instance, high-walkability neighbourhoods with good street design qualities may have a different pedestrian environment than high-walkability neighbourhoods with poor street design qualities (Figure 2-7). Some such examples are discussed in the following section.  2.5.3.1 High Neighbourhood Walkability but Poor Pedestrian Environment An example of a highly walkable neighbourhood with a poor street design is a situation where there is an appropriate residential density with good mixes of different land uses, an optimal amount of commercial uses, and a street network with high connectivity. However, on Figure 2-7. Types of neighbourhoods based on walkability and pedestrian environment. 39  the pedestrian environment side, this neighbourhood may lack the design and aesthetic qualities as well as the utilities, infrastructures, and services that make pedestrians comfortable walking around. The streets may have many commercial and retail destinations to walk to, but the buildings may not be well maintained, the street elevation may not be welcoming, and the sidewalks, crossings, and intersections may be poorly designed. There may be no traffic-calming measures in the streets, and the streets themselves may be poorly lit and lack seating, trees, or any attractive architectural or landscape features. Neighbourhoods with a high level of social deprivation and a higher concentration of low-income ethnic minorities in urban areas often have streets that are not pedestrian-friendly. Sometimes, downtown core can also include such environment (Figure 2-8).   Figure 2-8. Example of high neighbourhood walkability and poor pedestrian environment. 40  2.5.3.2 Low Neighbourhood Walkability but Good Pedestrian Environment Neighbourhoods with low walkability but good street design qualities often have a mix of street design components and features geared to create family-friendly urban living in the suburbs. Such a development pattern is usually found in new suburban residential developments with minimal retail destinations but good pedestrian environments such as sidewalks, streetlights, trees, benches, crossing design, local traffic control, etc. (Figure 2-9). Other examples include the newly developed suburban downtowns and retail centres (outlet malls) where people must drive to reach the destination and then can walk once they’ve arrived.    2.5.3.3 Low Neighbourhood Walkability and Poor Pedestrian Environment Neighbourhoods with low walkability and poor street design include most of the development following World War II. Such developments are characterized by low-density Figure 2-9. Example of low neighbourhood walkability and good pedestrian environment.  41  single-family residential developments that lack commercial space or have spread-out big-box stores amid large parking lots, and have poor street connectivity with a higher number of dead-end streets and cul-de-sacs along with a large separation between land uses (commercial, retail, recreation, etc.) (Figure 2-10). The streets in such neighbourhoods are designed for cars and have a higher number of lanes with a higher speed limit, fewer sidewalks, and limited crossing and intersection amenities.   Figure 2-10. Example of low neighbourhood walkability and poor pedestrian environment. 42  2.5.3.4 High Neighbourhood Walkability and Good Pedestrian Environment Neighbourhoods with high walkability and good street design qualities provide an ideal environment for walking (Figure 2-11). These neighbourhoods have good local accessibility to various destinations, diverse land uses, and good street connectivity. In addition, the streets in such a neighbourhood have appealing aesthetics and social environments that attract more pedestrians. There is an excellent visual connection between the streets and the insides of the buildings. Sidewalks are well designed with a good number of streetlights, seating areas, and trees. There is a safe buffer between the street and sidewalk, and traffic speed is lower with properly marked street crossings and lights. Such neighbourhoods are more common in urban areas, especially in cities that have focused their plans and policies to create more walkable environments in their urban neighbourhoods.  Figure 2-11. Example of high neighbourhood walkability and good pedestrian environment.  43  2.6 Summary Ecological models provide a guiding framework for addressing health problems at the population level. Ecological models assume that multiple levels of factors collectively affect health behaviour. These factors range from intrapersonal level (e.g., biological, psychological) to interpersonal level (e.g., social, cultural) to broader levels of community and public policy. Several studies have used the ecological approach to study physical activity. The built environment also influences physical activity at various levels. The Synergy between neighbourhood walkability and pedestrian environment can be analyzed using the ecological model of health behaviour. Neighbourhood level plans are implemented over a larger period, whereas street-level plans are short term plans. From a fiscal and pragmatic perspective, street-scale plans and policies are less demanding compared to neighbourhood-level plans and policies; street-scale plans can also be implemented at a lower cost. Modifying the street by adding pedestrian infrastructure can enhance walkability, thereby increasing walking. In addition to the effect of the built environment, socio-demographic and psychosocial factors also play a significant role. The built environment has a differential association with physical activity based on demographic and psychosocial factors.  Neighbourhood walkability is a mix of four components that capture various aspects of local accessibility. The components are i) residential density, ii) commercial floor area ratio, iii) street connectivity, and iv) land-use mix. A neighbourhood need not have the highest scores for each of these components to qualify as a walkable neighbourhood. It is therefore important to maintain a balance across all components of walkability to make neighbourhoods more walkable. Unlike the components of neighbourhood walkability, the pedestrian environment reflects different aspects of the street environment. Such components can be broadly categorized into 44  three sections: i) route-level components, ii) segment-level components, and iii) crossing-level components. Within these sections, various design features are included. Neighbourhood walkability and the pedestrian environment reflect different aspects of a neighbourhood’s conduciveness for walking. The combination of these two aspects creates a different environment for pedestrians. For instance, high-walkability neighbourhoods with good street design qualities may have a different pedestrian environment than high-walkability neighbourhoods with poor street design qualities.                 45  CHAPTER 3: LITERATURE REVIEW 3.1 Overview This chapter consists of four sections. The first section discusses how urban designers, transportation planners, and public health practitioners view the relationship between the built environment and walking. The second section discusses walking as a form of physical activity by reviewing the evidence supporting the health benefits of walking. The third section examines the built environment factors correlated with walking behaviour and physical activity. Built environment variables at three different spatial scales (metropolitan, neighbourhood, and street) and their relationship with walking and physical activity are discussed in this section. The fourth section discusses the various built environment, personal and psychosocial factors that affect physical activity in children, teens, and older adults.  3.2 The Nexus Between the Built Environment and Walking: Views from Urban Design, Transportation Planning, and Public Health The built environment is a multi-dimensional concept that comprises urban design, land use, and the transportation system and encompasses human activity patterns within the physical environment (Handy et al., 2002). Urban planners have long been interested in understanding how the built environment affects travel behaviour. Within this realm, the focus on the effect of the built environment on walking behaviour is comparatively recent. However, in a short time, there has been a tremendous amount of research examining the relationship between the built environment and walking behaviour. Urban design, public health, and transportation planning are the three fields that evaluate the relationship between the built environment and walking behaviour (Figure 3-1). However, their approaches, theoretical underpinnings, and interests in 46  understanding the relationship between the built environment and walking behaviour are different.   Research in urban design focuses on making places lively by increasing pedestrian activities. Following a normative approach, urban designers are mostly interested in understanding how people use spaces and propose design solutions appropriate for the users' needs. Urban designers, therefore, perceive the urban form and walking relationship as a design problem (Choi, 2012).  Various theories from prominent urban design scholars have laid significant groundwork for understanding the link between urban form and behaviour. Christopher Alexander, Jane Jacobs, William Whyte, Kevin Lynch, and Jan Gehl are prominent urban design scholars whose work has had a significant effect on urban design and planning. For instance, Christopher Alexander’s seminal books The Timeless Way of Building (Alexander, 1979) and Pattern Language (Alexander, 1977) provide theoretical explanations and illustrative Figure 3-1 Conceptual model showing “walking” as a nexus between urban design, transportation planning, and public health. 47  examples of patterns for developing towns, neighbourhoods, houses, gardens, and rooms. Kevin Lynch’s work (1960) on urban form discusses how people perceive urban form. Likewise, Jane Jacob’s work (1961) and William Whyte’s work (1980) respectively discuss neighbourhood vitality and social life in urban spaces due to increased pedestrian activities. These works have made a significant impact on urban design and planning practice.  Unlike urban designers, who are interested in making spaces lively and vibrant, transportation planners are interested in understanding how people make decisions about using different modes of transportation to get from one location to another. Transportation planners use economics theories (McFadden, 1973) and geography (Chapin, 1974; Hägerstrand, 1970) to understand how people make such decisions. Random utility theory is a dominant theory used to study travel behaviour; its fundamental notion assumes that people make rational choices and always aim to attain maximal utility with minimal cost (McFadden, 1973). However, people do not always make rational decisions (Foerster, 1979; Simon, 1959; Tversky & Kahneman, 1974). To address such conditions, transportation planners use behavioural economics theories, such as prospect theory (Tversky & Kahneman, 1974). These theories help transportation planners understand how people decide to manage their time schedules for departure (Jou et al., 2008), choose routes to get from one place to another (Ben-Elia & Shiftan, 2010), and prefer one mode over others on a day-to-day basis (Stathopoulos & Hess, 2012).  In addition to daily travel decision making, transportation planners are also interested in understanding how a decision made in other life domains affects travel behaviour. These domains include a choice of residential location, workplace location, auto ownership decisions, family formation, decisions about participating in labour, etc. (Salomon & Ben-Akiva, 1983). These decisions are made over a long period and pose significant challenges in studying how 48  such choices affect travel behaviour. Using the mobility biography concept, transportation planners have started to evaluate travel demands over individuals' lifetimes. Within mobility biographies, interrelated partial biographies can be embedded, such as a residential biography, employment biography, and household biography (Scheiner, 2007). A mobility biography provides a broader lens to understand how different incidents at different stages of life can influence travel behaviour. It also helps transportation planners understand how decisions people make in other domains of their lives affect their travel behaviour. Using various concepts of decision making, transportation planners seek to understand travel behaviour.  In public health research, walking is a form of physical activity with potential health benefits. Public health researchers are interested in understanding why people are physically active or inactive, and they use this evidence for public health interventions (Bauman et al., 2012). Various health behaviour models are used to evaluate the factors that affect engagement in physical activities, such as personal factors, interpersonal factors, and community factors that influence a person’s participation in physical activity. For instance, the Theory of Planned Behavior (TPB) is a personal health behaviour model. TPB argues that a person’s attitude toward a behaviour is a much better predictor of that behaviour (e.g., physical activity) than their attitude toward the object (e.g., physical activity-related health morbidities) at which the behaviour is directed (Fishbein & Ajzen, 1975). Likewise, to account for factors other than the personal, researchers must consider interpersonal factors such as the family, friends, and co-workers who make up an individual’s social network. The Social Cognitive Theory (SCT) is one of the health behaviour models that examine human behaviour by analyzing the interplay between personal, behavioural, and environmental factors (Bandura, 1971). One of the significant contributions of the SCT to behaviour study is “self-efficacy.” Self-efficacy can be defined as the conviction that 49  one has to successfully perform a specific action required to produce the desired outcome. Self-efficacy is a core construct of the SCT; various studies have validated its predictive ability in the context of physical activity, including walking behaviour (M. D. Young et al., 2014). Socioecological models (or ecological models) are another class of health behaviour models, and they are based on the idea that behaviour has multiple levels of influence. With various models, public health research provides a multitude of behaviour models rich with psychological and social factors that affect walking behaviour.  Though urban designers, transportation planners, and public health researchers have different approaches, theoretical assumptions, and interests in understanding the factors that affect walking behaviour, each of these fields provides a unique contribution to understanding the relationship between the built environment and walking behaviour. Urban design research provides abundant theoretical literature on the understanding of the built environment from a design perspective. For example, “wayfinding” is an important concept in urban design that facilitates people to navigate successfully in an urban environment (Ewing et al., 2005; Mullen et al., 2016) through appropriate design of the pedestrian environment. Wayfinding is particularly important for children's independent mobility (Riazi et al., 2019) and seniors' mobility for getting from one location to another (Mishler & Neider, 2017). Similarly, research in transportation planning provides an econometric framework for understanding the mechanisms that affect travel behaviour. The random utility theory has been used to study mode choice behaviours in children, teens and seniors (Ermagun et al., 2015; Kamargianni & Polydoropoulou, 2013; Liu et al., 2020; Ulfarsson & Shankar, 2008). And public health research offers a theoretical understanding of how various social and psychological factors influence health behaviour. Various social and psychosocial factors such as income, social support, and perception of safety 50  have been associated with walking and physical activity in children, teens and older adults. Though these fields view walking behaviour differently, they make a unique contribution to understanding walking behaviour. 3.3 Walking as a Form of Physical Activity Physical activity includes any bodily movement produced by skeletal muscles that result in energy expenditure (Caspersen et al., 1985). This definition of physical activity incorporates various activities that could range from highly structured activities such as sports to casual activities like a stroll in a park. These activities can be categorized based on their context, type, or intensity. Regardless of the type of activity, the amount of caloric expenditure reflects the actual physical activity measure. Metabolic Equivalents (MET2) is one of the standard metrics used to measure the energy expenditure of various physical activities, and this metric is also used to create categories based on the intensity of an activity, i.e., light, moderate, or vigorous. Though METs may not be able to capture the nuances associated with activities that vary in intensity (such as walking, which can be light, moderate, and vigorous), various empirical studies have assessed the health benefits of these three categories of physical activities using METs as the unit of measurement. Walking is a form of physical activity that can belong to all three categories of intensity-based physical activity—a stroll is light-intensity physical activity, brisk walking (e.g., walking as if late for a meeting, walking to catch a bus, etc.) is moderate-intensity physical activity, and walking uphill or with a load is vigorous-intensity physical activity. While walking uphill or carrying a load is more aligned with occasional activity such as hiking, both a light stroll and  2 MET is the ratio of the work metabolic rate to the resting metabolic rate. One MET is defined as 1 kcal/kg/hr and is roughly equivalent to the energy cost of sitting quietly.  51  brisk walking are more relevant for promoting physical activity. A stroll at 1–2 miles per hour can include recreational walking, and brisk walking at 3-4 miles per hour can consist of utilitarian walking like walking for transit. Light or moderate walking is associated with numerous health benefits (Hart, 2009; Matthews et al., 2007). Likewise, walking is also a low-risk physical activity, making it attractive for older adults and some mobility-challenged groups (Colbert et al., 2000).  Since walking is accessible to various demographics and does not require much dedicated time, specialized equipment, or a specific place, it is an available physical activity for all age groups. Walking is also negatively associated with long-term weight gain (Penny Gordon-Larsen et al., 2008). In addition to physical health benefits, walking can also have positive psychological and emotional effects (Marselle et al., 2013); for instance, walking in a natural setting can enhance mood (Roe & Aspinall, 2011). Walking can lead to more social interaction, which enhances social well-being (Kim & Yang, 2017); for example, “walking school bus” is a particularly useful means of socialization for children and their peers (Baslington, 2008). Lastly, through the indirect pathway, walking as a transportation mode improves air quality by reducing auto dependency.  Walking can be a practical physical activity for positive health (Stamatakis et al., 2018). Walking is perhaps the most efficient health promotion strategy—especially when coupled with transportation accessibility and social equity. Several studies have found that walking to transit is useful to help sedentary populations meet recommended physical activity levels (Besser & Dannenberg, 2005; Freeland et al., 2013; Lachapelle et al., 2011; MacDonald et al., 2010; Rissel et al., 2012; Saelens et al., 2014). The prevalence of a sedentary lifestyle is high in low-income and minority groups (Besser & Dannenberg, 2005; Bostock, 2001). Since low-income groups 52  cannot afford to allocate time for other kinds of physical activities, walking or taking transit to work can be a feasible alternative. Most importantly, studies suggest that other forms of physical activity do not compensate for physical activity done through transit use (Saelens et al., 2014). Likewise, people who walk for transportation purposes also engage in various physical activities (Ibid.). With all its benefits and convenience, walking is a useful utilitarian physical activity. 3.4 Built Environment Correlates of Walking 3.4.1 Metropolitan Scale 3.4.1.1 The Built Environment at Metropolitan Scale Studies examining the relationship between the built environment and walking behaviour at a metro scale often use urban sprawl as an indicator of urban compactness. Though urban planners started using the term “urban sprawl” as early as the 1920s (Nechyba & Walsh, 2004), a systematic approach to defining and measuring urban sprawl did not emerge until the late 1990s. Ewing (1997) characterizes sprawl by the degree of intensity of four features: i) low-density development, ii) commercial strip development, iii) expanses of single-use developments, and iv) leapfrog development. These features reflect whether an urban area has poor accessibility and whether it is automobile-dependent. Various works have quantitatively measured urban sprawl. Early works used density as a proxy for measuring urban sprawl, which resulted in some inconsistent findings. Acknowledging that urban sprawl results from the interactions of various factors of the built environment, later works have developed composite indices. Galster et al. (2001) developed a measure of urban sprawl using eight dimensions of land use: i) density, ii) continuity, iii) concentration, iv) clustering, v) centrality, vi) nuclearity, vii) mixed uses, and viii) proximity. They defined sprawl as a condition of land use represented by low values on one or more of these dimensions. 53  Ewing, Pendall, and Chen (2003) used a two-step approach to create a sprawl index. First, they used a data reduction technique called principal component analysis to develop indices that capture four urban form dimensions, including density, land-use mix, activity centring, and street accessibility. Then they combined these components to create a composite index. Hamidi, Ewing, Preuss, and Dodds (2015) updated the work by Ewing et al. (2003) by adding new variables to the four components and validating the overall index with new data. Like urban sprawl, regional accessibility is used to assess the relationship between the built environment and travel behaviour. Regional accessibility is defined as the ease of accessing major centres (jobs, retail, or recreational spaces) in an urban area (in some cases, just the distance to the central business district) and is primarily determined by the availability of transportation services. Various studies have used different methods to assess regional accessibility. The regional accessibility measures can be divided into three broad categories: cumulative opportunities measures, gravity-based measures, and utility-based measures (Handy & Niemeier, 1997). Cumulative opportunities measures count the number of opportunities reached within a given travel time (or distance). Gravity-based measures extend cumulative opportunities measures of regional accessibility by weighing opportunities by their intensity and the impedance or travel cost to the opportunity. Utility-based regional accessibility measures are based on random utility theory, in which the probability of an individual making a choice, i.e., picking a destination, depends on the utility of that choice relative to the utility of all choices. Regional accessibility measures are essential for urban planners. These measures are used to assess how people from different socio-demographics access urban services.  54  3.4.1.2 Built Environment Correlates of Walking at Metropolitan Scale Urban sprawl is often used to evaluate the distance travelled in a car, i.e., per capita vehicle miles travelled. A substantial body of literature supports the relationship between urban sprawl, travel behaviours, and various physical activity-related health outcomes. Studies that have evaluated the relationship between urban sprawl and walking have found an inverse relationship. Women living in highly sprawling areas are less likely to spend time walking, jogging, or running (James et al., 2013). Likewise, living in an area with low population density, a standard proxy for urban sprawl, can lead to a lower level of walking (Garden & Jalaludin, 2009). The evidence linking urban sprawl with walking and physical activity is not as substantial as the evidence linking urban sprawl with physical activity-related health outcomes.  Like urban sprawl, regional accessibility is also associated with walking and travel behaviour. High regional accessibility can lead to low per capita vehicle miles and less time spent in the car (Ewing & Cervero, 2001). Walking for transportation is also higher among people who live in areas with high regional accessibility (Cho & Rodríguez, 2015). People who live far away from major urban centres spend less time in transport-related walking compared to people living closer to the centre (S. C. Brown et al., 2014). Regional accessibility reflects the ease of accessing major services in an urban centre. A robust transit network with excellent access to the regional centre can lead to higher rates of walking and transit use (Frank, Greenwald, et al., 2010). Additionally, light rail transit can also lead to higher amounts of walking in residents living near the station (Huang et al., 2017).  55  3.4.2 Neighbourhood Scale 3.4.2.1 The Built Environment at Neighbourhood Scale Though metropolitan-scale built environment features reflect the general urban form associated with regional travel behaviour (or commuting), a significant body of research focuses on understanding the relationship between neighbourhood-scale built environment features and travel behaviour. Most of the research examining the relationship between the built environment and walking (including physical activity and travel behaviour) is heavily focused on the neighbourhood built environment. However, before the 1990s, planners used metropolitan-level built environment features (such as density) to study travel behaviour (Cervero & Kockelman, 1997). Two forces drove the shift from metropolitan-scale studies toward neighbourhood-scale studies. First, movements like new urbanism, transit-oriented development, and traditional town planning gained popularity in the 1990s and advocated that travel demand can be controlled through density, diversity, and design while planning neighbourhoods. Second, software technology developments, especially in geographic information systems, to collect and analyze large-scale disaggregated data, made it possible to capture neighbourhood-scale built environment features. Because any study that uses spatial data is prone to vulnerabilities because of the modifiable area unit problem (MAUP), the conceptualization of a neighbourhood is critical to understanding its relationship with walking behaviour. MAUP refers to the sensitivity of analytical results to the definition of spatial units for which data are collected (Fotheringham & Wong, 1991). Although there are various theoretical definitions of the spatial nature of the neighbourhood in different fields, all seem to view a neighbourhood as a geographical construct of a place, defined around the home and everyday activities, centred on schools, community 56  centers, parks, or retail services (Moudon et al., 2006). Urban planners define a neighbourhood based on walkability to gauge whether a neighbourhood environment has a favourable walking infrastructure (p. Ibid.). Studies usually use a 1-km street network buffer around an individual’s home as a neighbourhood definition (Frank et al., 2008; Moudon et al., 2007). However, the buffer sizes may also differ based on demographic characteristics (Villanueva et al., 2014) of the study population. 3.4.2.2 Built Environment Correlates of Walking at Neighbourhood Scale Earlier studies examining the relationship between neighbourhood built environment and travel behaviour used density and mixed uses as indicators of neighbourhood built environment (Frank & Pivo, 1994). Additionally, density, diversity, and design were also used as the three core neighbourhood built environment characteristics (“3Ds”). Early studies explored how the three built environment characteristics influenced the choice of different modes of transportation (Cervero & Kockelman, 1997). Later works added demography, distance to transit, and destination accessibility (Ewing & Cervero, 2001, 2010). Simultaneously, the concept of walkability was also becoming popular among urban form and travel behaviour researchers.  Though the term “walkability” has a recent history going back to the 1990s, the use of the word “walkable” has a long history that goes back to the 18th century (Forsyth, 2015). Walkability measures the conduciveness of a neighbourhood built environment for walking. Various composite indices to measure neighbourhood-scale built environments have been developed and validated empirically (Frank et al., 2005; Kuzmyak et al., 2006; Walk Score, 2021).  Frank et al.’s (2005) walkability index is one of the first walkability indices created using detailed built environment data on the measures of the land-use mix, residential density, and intersection density. This work builds on their prior work on the neighbourhood correlates of 57  walking and bicycling (Saelens, Sallis, & Frank, 2003). The indices developed after that have used one or the other combinations of density, connectivity, and land use mix. Among them, two are the most used indices. First is the walkability index developed by Frank et al. (2010), a combination of the normalized score of net residential density, commercial floor area ratio, intersection density, and land-use mix, all associated with walking behaviour. The index has been validated in various studies and has efficiently predicted walking behaviour (Cerin et al., 2007; Frank, Sallis, et al., 2010). Second is the Walk Score, which is also increasingly used to measure neighbourhood built environment. It is an online gravity-based metric that gives points to various destinations near a location. It then adds the points and normalizes them to create a score that ranges from 1 to 100. Multiple studies have assessed Walk Score’s reliability and validity and have found it a useful tool (Carr et al., 2011; Duncan et al., 2011). One significant difference between the two indices is that the walkability index captures the built environment's three-dimensionalities by measuring the ratio between commercial building floor area and land area where the building stands. In contrast, the Walk Score only uses the number of destinations accessible within a neighbourhood.  The walkability index has gained popularity in various fields ranging from real estate (Boyle et al., 2014; Gilderbloom et al., 2015; Walk Score, 2021) to transportation planning (Manaugh & El-Geneidy, 2011) to public health (Frank, Sallis, et al., 2010). Depending on the area of research and research questions, the definition of walkability also varies (Dörrzapf et al., 2019; Ramakreshnan et al., 2021; Shashank & Schuurman, 2019; H. Wang & Yang, 2019). Shashank & Schuurman (2019) reviewed the nuances in the definition and measurement of walkability based on behaviour (transport walking vs leisure walking) researchers are interested in and how such variation can create differences in neighbourhood walkability at the local level. 58  However, there is a general theme that can be seen in most walkability-related studies, i.e., walkability is a measure of neighbourhood built environment’s friendliness for walking.  Many studies have established the relationship between neighbourhood built environment and travel behaviour (Freeman et al., 2013; Manaugh & El-Geneidy, 2011; Owen et al., 2007). Kockelman (1997), one of the early studies, examined the relationship of travel behaviour to accessibility, land-use mix, and land-use balance and found that land-use balance was positively associated with the probability of walking. Rajamani, Bhat, Handy, Knaap, and Song (2003) found some relationship between land-use diversity and walking; they discovered that mixed uses have the potential to substitute driving with walking. Land topography also influences the choice of travel mode. Rodríguez and Joo (Rodríguez & Joo, 2004) found a one-minute increase in walking time due to the slope of local terrain lowered the odds of walking between 15% and 18.5%. Likewise, having destinations located within walking distance in the neighbourhood also increases walking probability (Koohsari et al., 2014).  Research looking at the neighbourhood-scale built environment and walking has mostly focused on the neighbourhood around the home. Several studies have also looked at the association between walking and the built environment around schools (Gallimore et al., 2011; Zhu & Lee, 2008). Relatively few studies have looked at the built environment around the workplace, where adults spend most of their time (Forsyth & Oakes, 2014; Schwartz et al., 2009). Some studies have shown that though the home is a common origin or destination of walking trips, a substantial number of these trips do not originate or terminate at home (Millward et al., 2013). A study by Forsyth and Oakes (2014) found a modest relationship between workplace built environment—including housing density and commercial density—and walking for transportation.  59  Self-selection is often cited as an essential factor by researchers examining the relationship between a neighbourhood-scale built environment feature and walking behaviour (Cao et al., 2006)(Cao, Handy, & Mokhtarian, 2006; Frank, Saelens, Powell, & Chapman, 2007; Handy, Cao, & Mokhtarian, 2005). Self-selection states that people choose to live in a neighbourhood that matches their preferences, both toward the neighbourhood and travel. Not addressing self-selection can lead to biased results, i.e., reverse causation. Cao, Mokhtarian, and Handy (2008) mention nine approaches to address self-selection: direct question, statistical control, instrumental variables models, sample selection models, propensity score, joint discrete choice models, structural equation models, mutually dependent discrete choice models, and longitudinal designs. In the same study, the authors found that the 38 studies they reviewed had significant associations between the features of the neighbourhood-scale built environment and walking behaviour even after accounting for self-selection.  There is a growing interest among walkability researchers to contextualize walkability based on the needs of various sociodemographic groups (Adkins et al., 2017; Alves et al., 2020; Liao et al., 2020; Richardson et al., 2017). For example, Golan et al. (2019) created a walkability index specially target for women using a mixed-method. Women’s Walkability Index and found it to be inversely correlated with Walk Score. They discovered that Walk Score showed high walkability in areas with high crime and homelessness density. They concluded that Walk Score inaccurately represents women’s walkability (Golan et al., 2019, p. 501). In another study, Alves et al. (2020) have proposed a conceptual framework for developing a walkability index targeted for the elderly population called Walkability Index for Elderly Health (WIEH). The WEIH is divided into four steps: (1) Analyzing public spaces and characterizing their quality for walking, (2) considering the existence of slopes and stairs, (3) calculating different routes for the elderly 60  in their daily routines, or when going to points of interest, and (4) selecting the “heart-friendly route” for elderly people (Alves et al., 2020, p. 01). The school walkability index, specifically targeted for children, was developed by S. Lee et al. (2020). They used a combination of GIS data and data collected through street audits. Using a combination of fine-grained street-level data and aggregate neighbourhood data using GIS is gaining interest among walkability researchers (Gu et al., 2018; Park et al., 2017; Yin, 2017).  3.4.3 Street Scale 3.4.3.1 The Built Environment at Street Scale Street-scale features of the built environment related to walking behaviour are relatively unexplored compared to the metropolitan-scale and neighbourhood-scale features of the built environment. Collecting information on the street-scale features requires a considerable amount of resources (both time and money). It is also hard to collect such data for a large geographic area. Measuring the experience of walking in the street is even more challenging. These challenges have limited empirical research examining the relationship between street-scale built environment features and walking behaviour.  Despite the challenges of measuring and collecting data on street-scale features of the built environment, researchers have developed tools and protocols (Active Living Research, 2015; Brownson et al., 2009). Most of the tools require fieldwork to collect data. These tools can be used to collect information on various aspects of streets; some tools use as few as 15 items (Sallis et al., 2015), and others use as many as 162 (Marlon G. Boarnet et al., 2006). Many of the tools are modified to adapt to individual research requirements on logistical grounds; some are modified to meet targeted users' needs. For instance, the Microscale Audit of Pedestrian Streetscapes (MAPS) tool has three different versions that target three different user groups. The 61  first version (MAPS-Full) is a 120-item survey for research use. The second version (MAPS-abbreviated) 60 items is for the use of researchers and advanced practitioners. The third (MAPS-mini) has 15 items and is designed for use by practitioners, advocacy groups, and community members. Various studies have used the audit tools to examine the validity of these tools (Adkins et al., 2012; M. G. Boarnet et al., 2011; Borst et al., 2008; Gallimore et al., 2011; Guo & Loo, 2013; Pikora et al., 2006).  3.4.3.2 Built Environment Correlates of Walking at Street Scale One of the first street audit tools (Pikora et al., 2002, 2006) found a significant relationship between street-scale built environment features and walking. A subsequent study used a different street audit tool (Cerin et al., 2006) and found a positive relationship between walking for transport purposes and the presence of a diversity of destinations, residential density, walking infrastructure, aesthetics, traffic safety, and low crime levels. It also found a positive relationship between walking for recreation and aesthetics, mixed destinations, and residential density. Physical street infrastructure like sidewalks, pedestrian crossings, and traffic signs are essential factors for predicting walking behaviour and physical activity (M. G. Boarnet et al., 2011). Likewise, trees, street furniture, and architectural treatments create a favourable environment for walking. Having functional street-crossing amenities, sidewalk characteristics, and street segment characteristics, with an aesthetically pleasing social environment, can lead to higher rates of walking and higher levels of physical activity in all age groups (Millstein et al., 2013). Having appealing street design qualities can also increase the number of people in the street, creating a lively environment that can induce more walking (Ameli et al., 2015; Ewing et al., 2005; Neckerman et al., 2013).  62  Though studies examining the association between the street-scale built environment and walking are limited and evidence of the relationship is weak, the results are encouraging (Millstein et al., 2013). Especially with an increase in the potential for collecting big data in street-scale built environment research, it is becoming easier to collect street scale data using an online mapping tool. Google Street View, Google Maps, and Bing Maps provide easy access to street-scale data and do not require time-consuming field visits. Several studies have examined the agreement between data collected through field visits and data collected virtually through Google Street View, Google Maps, and Bing Maps (or other virtual tools). These studies found that the virtual tools have the potential to collect data for sizeable geographic areas conveniently (Badland et al., 2010; Ben-Joseph et al., 2013; Clarke et al., 2010; N. Edwards et al., 2013; Kelly et al., 2013). Virtual tools have their limitations because of an inability to capture minute spatial details such as litter and broken glass, temporal details such as street activity, and qualitative measures like street condition (Badland et al., 2010; Ben-Joseph et al., 2013). However, such limitations could be complemented using supplemental data sources like GIS databases (Badland et al., 2010).  3.5 Factors Affecting Physical Activity in Children, Teens, and Older Adults Vulnerable groups are subpopulations that often lack the necessary physical capabilities, educational backgrounds, communication skills, or financial resources to adequately safeguard their health (Shi & Stevens, 2010). The vulnerability involves several interrelated factors such as individual capacities and actions, the availability or lack of individual and instrumental support, and neighbourhood and community resources that may facilitate or hinder personal coping and interpersonal relationships (Mechanic & Tanner, 2007). Mechanic and Tanner (2007) argue that vulnerability is a cumulative effect of various factors over the course of a lifetime; these include 63  socio-economic factors, race, social networks, personal limitations, and the location of their residence. There are various vulnerable groups; this study, however, discusses the factors affecting physical activity participation in children, youths, and older adults since they are the vulnerable populations in physical activity research (Dubbert et al., 2004).  Despite documented evidence suggesting the multiple health benefits of physical activity, a significant segment of the population does not meet the recommended levels of physical activity—children, teens, and older adults are more likely to be physically inactive. As discussed at the beginning of this document, participation in physical activity at an early age plays a significant role in determining lifelong physical activity patterns (Faulkner et al., 2009; Malina, 2001). Despite being the most physically active segment of the population (Dubbert et al., 2004), children and teens have demonstrated a decreasing trend in physical activity participation (Tremblay et al., 2014). This shift is common during adolescence, which is a transitional phase from youth to adulthood (Kwan et al., 2012), and a relatively sedentary phase of life. Therefore, this life stage is crucial to the development of physically active lifestyles.  Older adults who are on the other side of the age spectrum are at high risk of losing mobility independence as they age; this may create mobility impairment, resulting in low physical activity and poor physical health. Despite the numerous health benefits of physical activity, older adults are also one of the least physically active groups (Kerr et al., 2012); older adults are more likely to be physically inactive if they have a chronic condition (Watson et al., 2016). Ensuring that older adults have healthy and physically active lives is one of the significant challenges for public health practitioners. 64  3.5.1 Children and Teens 3.5.1.1 Built Environment Factors The built environment can influence physical activity in children and teens in various ways. The first is related to active transport to school, which is the most common effect of the built environment on physical activity patterns in children and teens (Mandic et al., 2016; Pang et al., 2017). Living in an urban environment with high walkability and high regional accessibility via a transit system, teens can accumulate more physical activity through active transport to school compared to those living in a suburban area (Frazer et al., 2015). The second relates to access to locations like parks, gyms, etc. where children and teens can participate in physical activity (Frank et al., 2007). Easy access to physical activity locations can also increase active transportation (Sallis et al., 2004). The third includes neighbourhood streets, sidewalks, and cul-de-sacs where children can play (Jones et al., 2009). In addition to all these mechanisms, the built environment can also influence physical activity in children through their parents’ behaviour. For instance, parents who decide to live in a walkable neighbourhood may choose not to own a car, which may lead to more walking for their children.  Along with these measures of the built environment, street-scale features have also been found to have a significant association with physical activity in children and teens. The pedestrian environment can be made more conducive to walking and engaging in physical activity for children and teens. A safe street with more destinations, better aesthetics, and better functionality can encourage walking (Rodríguez et al., 2014). Features like sitting space, streetlights, and windows facing towards the street can increase “eyes on streets,” which can enhance walking and “free play” (Veitch et al., 2008). A better pedestrian environment can also improve pedestrian safety from traffic, which is a significant concern in a neighbourhood with 65  high commercial area and diverse land uses. Safety concerns related to street incivilities and disorder can discourage walking in children and teens (Nasar et al., 2015). The safety concern is particularly significant for independent mobility in children and teens and is more pronounced in girls compared to boys. The pedestrian environment, therefore, has a significant relationship with physical activity in children and teens.  3.5.1.2 Sociodemographic and Psychosocial Factors 3.5.1.2.1 Sociodemographic Factors Age, gender, ethnicity, and family/neighbourhood income are some of the sociodemographic factors associated with physical activity in children and teens (Sallis & Owen, 1998). As children and teens age, they are less likely to engage in physical activity. This trend can start while the individual is transitioning from childhood to adolescence (Farooq et al., 2018; Verloigne et al., 2012), as well as from adolescence to adulthood (Armstrong & Welsman, 1997), and puts children and teens at a higher risk of being physically inactive. Furthermore, the decline in physical activity with age is higher among girls (Craggs et al., 2011). Gender Findings from various studies have consistently shown a gender-specific difference in engagement in physical activity among children and teens (Armstrong & Welsman, 1997; Farooq et al., 2018; Lampinen et al., 2017; Seabra et al., 2013; Telford et al., 2016; Verloigne et al., 2012). Telford et al. (2016) argue that the possible explanation for lower physical activity among girls could be lower participation in organized sport, less social support to engage in physical activity and lower enjoyment level when doing physical activity. Preference for outdoor activities also tends to be higher among boys than girls (Korpela et al., 2002). Girls are also more likely to engage in relatively low-intensity, rhythmic physical activity such as dancing 66  (Lampinen et al., 2017). The difference in physical activity may be related to the difference in attitudes towards physical activity in boys and girls' social circles. According to Seabra et al.(2013), parental support and peer acceptance are positive correlates of girls' physical activity, whereas perceived physical competence is a significant factor for boys’ physical activity. A similar argument is presented by Laird et al. (2018), who studied the underlying mechanism of how social support affects physical activity in girls. Income In addition to age and gender, income is also a significant predictor of physical activity in children and teens (Katzmarzyk et al., 2018; Sterdt et al., 2014). Children from low-income households tend to show lower physical activity levels (Drenowatz et al., 2010; Tandon et al., 2012). Similarly, teens from low-income households are also more likely to have a lower level of physical activity (Burns et al., 2020; Butcher et al., 2008; Lämmle et al., 2012; Raudsepp & Viira, 2000). Lower household income can influence physical activity in various ways. Children from low-income households may have limited access to play equipment ( bikes and jumping ropes) and more physical activity restrictions (Tandon et al., 2012). Similarly, parents’ work schedule, lack of interest and competing commitments can also act as physical activity barriers in children from low-income households (Finkelstein et al., 2017). Children from low-income families also tend to have lower levels of organized physical activity (such as supervised activities organised by sports clubs) and girls are at higher risk of having a lower level of physical activity (Lampinen et al., 2017). Parents with low household income also may lack enough financial resources for their children to access private recreational facilities for organized physical activity (Chang & Kim, 2017; Finkelstein et al., 2017).  67  Neighbourhood income is also associated with physical activity in children and teens (Romero, 2005). Both social and physical environments in low-income neighbourhoods can influence physical activity in children and teens. One of the significant barriers to physical activity in children and teens in low-income neighbourhoods is the perception of safety. Fear of crime in low-income communities can limit participation in physical activity, even for normal age adults (Richardson et al., 2017). Lack of affordable recreational facilities, illicit activity in public spaces and lack of information are some of the barriers to children and teens' physical activity in low-income neighbourhoods (Finkelstein et al., 2017). Lower perception of the quality of recreational facilities in low-income neighbourhoods can discourage teens from participating in physical activity. Neighbourhood disorder and lower social capital in low-income neighbourhoods may also act collectively to influence children and teens (Van Loon et al., 2014).  Though low family income and low neighbourhood income have shown an inverse relationship with children and teens' physical activity, some studies have found conflicting results. Studies have found that levels of transport walking in higher among children and teens from low-income households and low-income neighbourhoods (L. Rothman et al., 2018). Frank et al. (2007) found a higher level of active transport among teens from low-income neighbourhoods. Similar results were found by (Pabayo et al. (2012), who used longitudinal data to examine the collective influence for poverty and safety on transport walking among children. Lower-income is associated with lower access to a car, leading to higher transport walking as a necessity (Frank et al., 2007; L. Rothman et al., 2018; Van Loon et al., 2014). While this may lead to positive health benefits from increased walking levels in low-income neighbourhoods, it may put children at higher risk of injury due to poor pedestrian infrastructures in low-income 68  neighbourhoods (Thornton et al., 2016). This raises health inequity, which needs to be addressed, putting a higher burden of harmful conditions low-income populations are exposed to (Pabayo et al., 2012).  3.5.1.2.2 Psychosocial Factors Apart from these demographic factors, various psychosocial factors can influence physical activity in children and teens. Family environment plays a significant role in children’s and teens’ physical activity since these stages of life are the formative time for the development of behaviour that can have a lifelong impact (Bingham et al., 2016; Edwardson & Gorely, 2010; Lloyd et al., 2014; Rhodes, Guerrero, et al., 2020). Parents can act as gatekeepers and influence physical activity in children and teens in multiple ways. For example, parental perception of neighbourhood safety and physical activity support has a significant role in children and teens' physical activity. These factors can act as both facilitators and inhibitors of physical activity in children and teens.  Parental Perception of Safety  Parents’ perceptions of safety can affect their children’s engagement in physical activity, especially independent mobility (Carver et al., 2008; Foster et al., 2014; Page et al., 2010). Perception of safety can include various factors such as traffic safety, crime safety and stranger danger (Panter et al., 2008). Fear of traffic fatalities is an important safety factor that parents are most concerned with, and it is crucial for walking to school and playing outdoors (Ahlport et al., 2006; Gielen et al., 2004). Parents who perceive higher traffic speed in their neighbourhood can limit their children from playing outdoors (Jago et al., 2009), leading to lower outdoor physical activity (M. Luo et al., 2020). A study by Jelleyman et al. (2019) found that parents tend to be 69  neutral about the risk associated with independent mobility but be more deterred by the risk associated with road safety.   Perception of neighbourhood crime is another factor that may limit independent mobility and physical activity in children and teens. However, perception of crime can be measured in various which may lead to inconsistent findings on its role on physical activity in children and teens (Ding et al., 2011; Giles-Corti et al., 2009). According to Kneeshaw-Price et al. (2013), perception of crime can include general crime and disorder and stranger danger. Studies that have used an objective measure of crime and disorder have found it to be a barrier for physical activity in children and teens (Gómez et al., 2004; S. H. Kneeshaw-Price et al., 2015; Molnar et al., 2004). Though the objective measure of crime tends to predict behaviour more consistently than the perceived measure, crime perception is accepted as a barrier for physical activity in children and teens.  Stranger danger is another aspect of the neighbourhood environment that can influence engagement in physical activity and independent mobility in children and teens (Carver et al., 2008). Stranger danger is perhaps one of the critical factors influencing physical activity, along with traffic safety (Carver et al., 2008; Christian et al., 2015; Faulkner et al., 2010; Malone & Rudner, 2011; Shaw et al., 2015). Parental concern regarding strangers harming their kids is one of the re-occurring issues in most independent mobility literature (Francis et al., 2017). Stranger danger can include multiple factors such as unwelcome approaches by strangers, abduction, assault, molestation, even murder, and the perpetrator can be adults, teenagers or older children (Carver et al., 2008). Various studies have argued that parental perception of stranger danger, especially related to kidnapping by children by strangers, maybe exaggerated than actual statistics (Ding, Bracy, et al., 2012; Ikeda et al., 2019). Parents perceiving higher stranger danger 70  tend to limit their children from engaging in the unsupervised activity and independent mobility and prefer sedentary activities that need minimal supervision. Kyttä et al. (2015) argue that a lack of social trust may influence stranger danger. Similar arguments are echoed in other papers (Foster et al., 2014; Francis et al., 2017; Shaw et al., 2015). Neighbourhood friendliness and creating a better sense of community could be a factor to reduce the perception of stranger danger (Alparone & Pacilli, 2012; Lin et al., 2017; Pacilli et al., 2013).  Perception of safety can have a differential impact on girls compared to boys (Carver et al., 2010; Mitra et al., 2014). Parents tend to be more protective about their daughters, which counteracts independent mobility (Bennetts et al., 2018; Foster et al., 2014; Ghekiere et al., 2017). Parents tend to allow their sons' independent mobility at an earlier age than daughters (Hillman et al., 1990) and less likely to let daughters walk or bike without adult supervision (McDonald, 2012). Lower level of independent mobility among girls at an early age can negative impact, such as internalized prohibitions, fears, and anxiety related to independent mobility in neighbourhood which can have a significant impact on their urban mobility throughout life (Pacilli et al., 2016).   Parental Support for Physical Activity Children and teens spend a significant amount of their formative years in the family environment in their parents' care (Rhodes et al., 2013).  Parents can play a dual role of ‘role models’ and ‘gatekeepers’ to participate in physical activity (Bauer, Neumark-Sztainer, et al., 2011; Gustafson & Rhodes, 2006). The parents’ own physical activity levels can also significantly affect their children’s physical activity. While the relationship between parent and children's physical activity level is inconsistent, parental support has shown a significant 71  association with children and teens physical activity (Beets et al., 2010; Edwardson & Gorely, 2010; Gill et al., 2018; Gustafson & Rhodes, 2006; Langer et al., 2014; Rhodes, Perdew, et al., 2020; Trost & Loprinzi, 2011).  Parental support for physical activity can take various forms. In their review of 34 studies on parental support, Gustafson & Rhodes (2006) identified three forms of parental support: encouragement, involvement and facilitation. Another systematic review by (Rhodes, Perdew, et al. (2020) also identified similar parental support elements, including encouragement, logistical support, and co-participation activity and other components like providing information, home-environment facilitation, spectating, and supervision. A study by King et al. (2008, p. 374) found that adolescents who received parental encouragement to exercise and who had an exercising friend engaged in significantly more days of physical activity than did their counterparts. Pyper et al. (2016) examined the impact of different parental support behaviours on child physical activity. They found that parents who take their child to places where they can be active were more than twice as likely to have their child meeting physical activity guidelines. Parent-child co-participation in physical activity is associated with higher levels of physical activity in children (Dlugonski et al., 2020) and lower sedentary time (Määttä et al., 2018). Active support of children is the most reliable parental influence of child physical activity behavior (Rhodes et al., 2013).  Parental support tends to be effective in younger children and tends to decrease with age (Bauer, Laska, et al., 2011; Davison & Jago, 2009; Gustafson & Rhodes, 2006). Lau et al.(2016) examined temporal variation in parental influences on adolescent physical activity and found adolescents perceived significantly less parental encouragement and instrumental support and reported fewer active peers as they got older. This inverse relationship may be related to older 72  children's autonomy (Bauer, Laska, et al., 2011). However, this may not be a practical approach since physical activity level tends to decline in adolescence. Perception of parental support for physical activity plays a vital role in adolescents' physical activity (De la Torre-Cruz et al., 2019). Parental support, therefore, may be crucial to maintain physical activity in adolescents.  3.5.2 Older Adults 3.5.2.1 Built Environment A broad range of built environment factors can significantly facilitate or constrain physical activity patterns in older adults. These factors range from macro-level features like population density and land-use mix to microscale features like street seating, lighting, and trees. Built environment features can affect physical activity in older adults through transport-related walking and leisure walking. For instance, easy access to neighbourhood services is more likely to increase transport walking among older adults (Adams et al., 2012). Likewise, having nearby parks and recreational spaces and an excellent pedestrian environment can also lead to higher rates of leisure walking and physical activity in older adults (Buman et al., 2013).  Older adults are more likely to walk for transportation if their neighbourhood has reasonable access to retail and commercial destinations, such as grocery stores (A. C. King et al., 2011). In addition, a pedestrian environment with curb cuts and no physical incivilities, etc., can also help increase transport walking in older adults (Sallis et al., 2015). The pedestrian environment created by microscale built environment features such as curb cuts, street seating, lighting, and trees can help mitigate the traffic generated by high density and mixed uses, which are essential for neighbourhood walkability. Studies have shown that having the right pedestrian environment can increase transport walking in older adults (Cain et al., 2014). Also, fear of falling is a common psychological factor that constrains older adults from walking (Rantakokko 73  et al., 2009); having a pedestrian built environment that helps boost confidence for walking is therefore a significant way to increase physical activity in older adults.  The features of the built environment also influence leisure walking in older adults. Studies have shown that older adults are more likely to walk for leisure than for transport. Likewise, leisure walking is a common form of physical activity among older adults (Ghani et al., 2016), especially among older adults who are sedentary (Siegel et al., 1995). Though neighbourhood walkability has been found to be a significant predictor of transport walking in older adults, its relationship with leisure walking is not clear (A. C. King et al., 2011). This could be because neighbourhood walkability is based on access to destinations that focus more on transport or utilitarian walking. However, studies have found that recreational walking is higher among older adults when the microscale built environment features create a supportive pedestrian environment (Buman et al., 2013; Sallis et al., 2015).   3.5.2.2 Sociodemographic and Psychosocial Factors 3.5.2.2.1 Sociodemographic Factors In addition to the built environment features, various sociodemographic factors influence the physical activity of older adults. Age, gender, income, ethnicity, health status, and vehicle ownership are some personal factors that influence physical activity in older adults. Due to increasing physical frailty, older adults are less likely to engage in physical activity than those in the early stages of older age. Likewise, the gender difference in physical activity engagement is well established across all age groups. Women are less likely to be physically active compared to men. At the same time, women's average life expectancy is higher than in men (Austad & Fischer, 2016). Though women have an advantage over men on the longevity side, this longer 74  life expectancy puts them at higher risk of living a life with higher morbidities caused by lack of physical activity. Apart from age and gender, income is another critical personal factor influencing older adults' physical activity. Studies have found that older adults with higher incomes spend more time engaged in moderate and vigorous physical activity than those with lower incomes.   Gender Gender difference in physical activity in older adults has been documented in various studies (Ferreira et al., 2010; Notthoff et al., 2017). As discussed in the earlier section, this difference starts as early as childhood, which can permeate the later stage of life (The Lancet Public Health, 2019). Mobility limitations increase with ageing, increasing decline in physical activity level at an older age. This decline is higher among women compared to men (Musselman & Brouwer, 2005). Mottram et al. (2008) examined mobility limitation using a population survey of 18,497 British older adults and found that mobility limitation in women higher than men. Additionally, the decline in physical activity during older age is also higher in women than men (Ferreira et al., 2010; Pollard & Wagnild, 2017). A nationally representative sample showed that regular walking decreases with age in women (Reis et al., 2008). Lee (2005) found that women tend to spend less time in total walking and leisure-time physical activity. Another study by Berger et al. (2005) looked at the impact of retirement on physical activity and found women less likely to meet the recommended level of physical activity, which is similar to the results found by Zhao et al.(2011) who looked at the physical activity in U.S. older adults with diabetes mellitus. The additional finding from the Berger et al. (2005) study was that physical activity loss due to retirement might not be compensated by physical activity from leisure after retirement. This is particularly important because most older adults spend a retirement life, and 75  ways to compensate for the loss of work-related physical activity may be necessary. However, some studies have found conflicting results showing men spending less time doing the physical activity than females (Black et al., 2015; Stephan et al., 2011).   However, the gender difference in older adults’ physical activity is not consistent.  There may be various reasons for lower levels of physical activity in older women compared to older men. One possible explanation could be related to the focus on meeting the recommended physical activity guidelines. Amagasa et al. (2017) found the older women are less active when physical activity is measured using the standard guidelines but are more active when total physical activity is measured.  A similar pattern was observed in a systematic review of physical activity in older adults done by Sun et al. (2013). However, unlike Amagasa et al. (2017), Sun et al. (2013) examined studies that used self-reported total physical activity. The inconsistency in differences in physical activity in older adults could be related to types of physical activity studied, i.e., high-intensity physical activity (such as exercise and sport most common among men) versus low-intensity activity (such as household/gardening work most common among women) (Notthoff et al., 2017).  Income Income status tends to have a differential relationship with health status across various demographics, with lower income groups at disproportionately higher risk of adverse health status across all age-groups (van Zwieten et al., 2018). And this differential impact can culminate into multiple morbidities at the later stage of life (Economou & Theodossiou, 2011). Lower-income is also found to be associated with a higher risk of mobility impairment in older adults (Nordstrom et al., 2007; Webber et al., 2010). Satariano et al.(2012) identified four public health burden related to limited mobility. First, poor health outcomes due to limitations in walking and 76  driving. Second, health issues and injuries related directly limited to mobility limitation. Third, social isolation due to lack of social contact. Fourth, lower community participation and engagement. All these public health risk put lower income older adults at higher risk of the poor health status and limited mobility.  Low-income households tend to have less disposable income for structured physical activity, and the neighbourhood environment may play a significant role in their physical activity (Chudyk et al., 2017). At the same time, lower-income older adults may not be able to afford a car to meet their mobility needs making walking or taking transit a necessity for daily needs(Grant et al., 2010; Van Holle et al., 2014). Low household income often coexists with lower neighbourhood income and such neighbourhoods lack pedestrian supportive environment for walking (Zandieh et al., 2016). Higher level of walking, even as a necessity, can increase contribute to physical activity levels. However, navigating street environment for daily errand can increase the risk of fall and traffic related injury in older adults which can low income seniors at disproportionately higher health risk.  3.5.2.2.2 Psychosocial Factors Self-efficacy  Self-efficacy is an important psychosocial determinant of physical activity in all age groups (De Bourdeaudhuij & Sallis, 2002; Dilorenzo et al., 1998; Mcauley et al., 2011; Motl et al., 2002). Self-efficacy is a core tenet of the Social Cognitive Theory which has been used to study health behaviours in various contexts (M. D. Young et al., 2014). Higher self-efficacy for physical activity increases participation in older adults' physical activity (Harris et al., 2009; Perkins et al., 2008). Older adults with higher self-efficacy for walking spend more time walking for transportation and leisure purposes (Thornton, Kerr, et al., 2017b). A higher self-efficacy for 77  walking has a positive association with walking in disadvantaged populations (Cleland et al., 2010).  Higher self-efficacy for physical activity can offset some of the personal barriers to engage in physical activity. For example, Gallagher et al.(2015) found that self-efficacy for being physically active despite individual physical activity barriers is significantly associated with neighbourhood walking duration with somewhat stronger effects among fallers than non-fallers. Similarly, another study by Maly et al. (2007) examined the role of self-efficacy on walking performance in older adults with osteoarthritis and found that self-efficacy mediates the effect of pain, stiffness, and age on walking performance.  Social Support The World Health Organization (2002) has recognized social support as a critical factor for healthy ageing. Studies have shown that social support is key to healthy and active living across age groups and populations (Pels & Kleinert, 2016). Social support for seniors is vital since they are more likely to have a smaller social circle due to the loss of family members and friends, leading to a high risk of being isolated and lonely (World Health Organization, 2002). Loneliness and social isolation can lead to low physical activity in older adults (Shankar et al., 2011). Having a strong family and social support network, however, can attenuate some of the effects of the barriers to physical activity (Lindsay Smith et al., 2017), such as lack of interest for physical activity in older adults (Moschny et al., 2011; Nelson et al., 2007).  Social support can be from family, friends, acquaintances or special groups and be associated differently based on the types of physical activity and study population. A study by (Bopp et al., 2004) found a positive association between social support and physical activity in older rural African American and Caucasian women. A relatively older study on Canadian older women found that having active friends and encouragement from a social circle is associated 78  with physical activity (Cousins, 1995). In another study by (Orsega-Smith et al., 2007), social support from friends and not the family was related to leisure-time physical activity. A positive association between social support and walking was found by Ory et al. (2016). An interesting thing in Ory et al. work was that they combined both human and animal support, i.e., having a dog to measure social support. Using a combined score for family support and friend support, Thornton et al. (2017a) found social support positively related to walking for errands, leisure/exercise and accelerometer-measured daily moderate to vigorous physical activity in older adults. Though there seems to be variation in how social support is measured and the behavioural context of its use, the existing evidence shows that social support for physical activity is significant for older adults’ physical activity.  Perception of Safety Perception of safety related to crime, traffic, and pedestrians are significant factors for outdoor and neighbourhood physical activity in older adults (Bracy et al., 2014). Lower perception of safety can limit mobility in the neighbourhood, thereby increasing social isolation among older adults.  Studies have shown the lower perception of safety is associated with lower levels of physical activity and walking (Barnett et al., 2017; Li et al., 2005; Mendes de Leon et al., 2009; Piro et al., 2006; Sugiyama & Thompson, 2008; Weber Corseuil et al., 2012; Wilcox et al., 2003). A meta-analysis by Barnett et al.(2017) found a strong association between safety from crime and physical activity; another study by (Van Cauwenberg et al. (2014) found older adults’ television viewing time inversely related with higher safety from crime. A positive relation between perceived safety and physical activity was also observed in another study by Wilcox et al. (2003). Similarly, Piro et al. (2006) examined if the perception of crime and physical activity 79  relationship varies by gender and found that perceived fear of crime was inversely related to physical activity only in older women. In addition to physical activity, perception of safety is also associated with walking in older adults (Barnett et al., 2017; Li et al., 2005; Mendes de Leon et al., 2009; Won et al., 2016; Zandieh et al., 2016). Won et al.’s review (2016) found crime-related safety consistently associated with walking. Interestingly, their study did not find a consistent relationship between traffic-related safety and walking. A positive relationship between safety from crime and walking in older adults was also observed by Maisel (2016) and Barnett et al. (2017).  While various studies have found a significant relationship between the perception of safety and physical activity in older adults, some studies have found different results. For example, Tucker-Seeley et al.(2009) examined the relationship between perceived neighbourhood safety and physical activity in older adults. They found a significant positive relationship between higher perception of safety and leisure-time physical activity, which became non-significant after controlling for health status. Similarly, Barnett et al.’s (2017) systematic review did not find a significant association between walking and perception of traffic safety, but there was a significant association with crime safety. Similar results were observed in Won et al.’s review (2016). According to Bracy et al. (2014), the mixed results suggest that the relation of crime, traffic and pedestrian safety with physical activity may not be direct, and they may be acting as moderators. For example, Li et al. (2005) between the number of street intersections and perceptions of safety from traffic for neighbourhood walking in older adults. Additionally, perception of safety can also indirectly affect physical activity via self-efficacy (Gallagher et al., 2012) and even act as a mediator of the relation between neighbourhood built environment and walking (Van Dyck et al., 2013).  80  3.6 Summary Urban design, public health, and transportation planning are the three fields that evaluate the relationship between the built environment and walking behaviour. However, their approaches, theoretical underpinnings, and interests in understanding the relationship between the built environment and walking behaviours are different. Research in urban design focuses on ways to make places lively by increasing pedestrian activities. Transportation planners are interested in understanding how people make decisions about using different transportation modes to get from one location to another. In public health research, walking is a form of physical activity with potential health benefits. Walking behaviour lies at the nexus of the three fields. Though these fields view walking behaviour differently, they all acknowledge the health benefits of walking. Walking is accessible to various demographics and does not require much dedicated time, specialized equipment, or a specific place; it is an available physical activity for all age groups. Walking has the potential to be a practical physical activity for positive health. Walking is perhaps the most efficient health promotion strategy—especially when coupled with transportation accessibility and social equity. Likewise, people who walk for transportation purposes also engage in various physical activities. With all its benefits and convenience, walking is a useful physical activity. The built environment correlates of walking can be studied at three spatial scales: i) Metropolitan scale, ii) Neighbourhood scale, iii) Street scale. Studies examining the relationship between the built environment and walking behaviour at a metropolitan scale often use urban sprawl to indicate urban compactness. At the neighbourhood scale, the research focuses on walkability and its relationship with walking. The association of urban design features of the 81  pedestrian environments with walking behaviour is studied by research done at the street scale. The physical activity patterns in children, teens and older adults are affected by the built environment features at all three spatial levels. However, the neighbourhood scale and pedestrian scale built environment features tend to have a stronger association with physical activity in the three age groups. In addition to the built environment, various personal and psychosocial factors also play an essential role in physical activity in children, teens and older adults. These include factors like gender, income, social support and perception of safety.                 82  CHAPTER 4: METHODOLOGY 4.1 Overview This chapter presents the data sources, built environment measures, and data analysis approach used in this research. The first section discusses the data sources, their study designs and the psychometric properties of the data collected by the survey questionnaires. The second section discusses the variables used in this research. The third section elaborates on the analytical approach used to examine the hypotheses.  4.2 Data Sources This dissertation used the baseline data from three cohorts: Neighborhood Influences on Kids (NIK), Teen Environment and Neighborhood (TEAN), and Senior Neighborhood Quality of Life Study (SNQLS). The NIK study is a longitudinal observational study evaluating neighbourhood-level activity and nutrition environments' associations with children’s weight statuses and obesity, including physical activity (Frank et al., 2012). The study was conducted in Seattle/King County, Washington, and San Diego County, California. The study used both self-reported and objective measures of physical activity. The TEAN study focused on teens and is an observational study conducted in the Baltimore, Maryland, and Seattle regions with teenage children between ages 12 and 16 and sought to evaluate how residence location affected their health-related behaviour (J. A. Carlson et al., 2014). The SNQLS is aa observational study of ambulatory adults 66 years of age or older living in Seattle/King County, Washington, and Baltimore, Maryland (Shigematsu et al., 2009). The study's main aim was to evaluate the multiple health outcomes among residents living in neighbourhoods that differed in walkability and income. The study sites were selected because of the availability of detail parcel level data and variability in neighborhood walkability(Frank, Sallis, et al., 2010; A. C. King et al., 2011; 83  Sallis et al., 2009). These three studies used a similar study design and sampling strategies to those used by Frank et al. (2010) in the Neighbourhood Quality of Life Study (NQLS).  The NQLS study examined the relationships between the neighbourhood built environment, socio-economic characteristics, health outcomes (Frank, Sallis, et al., 2010) in the Baltimore, Maryland and Seattle, Washington region in the USA. The study stratified neighbourhoods based on walkability and median household income before recruiting participants. Census block group was used to select neighbourhoods. Stratification was done to maximize variation in the built environment and demographic characteristics of the sample (Sallis et al., 2009). The census blocks were grouped into deciles based on the walkability index to create low-walkability (2nd, 3rd, 4th deciles) and high-walkability (7th, 8th, 9th deciles) neighbourhoods. Similarly, using the 2000 census information, the census block groups were categorized into low-income (2nd, 3rd, 4th deciles) and high-income neighbourhoods (7th, 8th, 9th deciles). The 1st and 10th deciles were not included to avoid the outlier effect, and the 5th and 6th decile were omitted to create separation between the high and low groups. The high and low “walkability” and income characteristics of each block group were combined (low-/high-walkability × low-/high-income) to create a list of block groups that fit into one of four categories. Other studies have used such a stratified sampling approach to examine the travel behaviour and physical activity of residents living in high and low neighbourhood walkability (Sallis et al., 2004). 4.2.1 Neighbourhood Impacts on Kids (NIK) 4.2.1.1 Neighbourhood Selection The NIK study used stratified sampling to select neighbourhoods for recruiting participants to maximize the variation in physical activity environments and nutrition 84  environments. The physical activity environment is based on data that include the walkability index and playability index. The walkability index3 is a composite index created by summing normalized residential density, street connectivity, commercial floor area ratio, and land-use mix for each census block (Frank, Sallis, et al., 2010). Likewise, the playability index is based on park proximity and park quality as assessed by the Environmental Assessment of Public Recreation Spaces (Saelens et al., 2006). The nutrition environment is based on information about fast-food restaurant proximity and concentration and supermarket proximity.  The study used physical activity environment and nutrition environment measures to create a 2x2 matrix of high/low physical activity environments and nutrition environments. The high/low settings were defined through specific numeric thresholds for the high and low categories for each region (Frank et al., 2012). The study used census block group boundaries to define neighbourhoods for two reasons: i) it is the lowest level of census geography with publicly available census data, and ii) the built environment at this geography is more homogeneous than at larger levels of census geography (Frank et al., 2012).   4.2.1.2 Recruitment The study used a 2x2 matrix of the physical activity environments and nutrition environments to identify block groups with households that had children aged 6–11 years and contrasting physical activity and nutrition environments. A commercial marketing firm provided the contact information of the households that fell within the sampling framework. Families were randomly selected, and letters were mailed to potential participants. A week after the message  3 It is discussed in more detail in the section on built environment measures. 85  was mailed, researchers followed up by telephone to provide details about the study, evaluate eligibility, and assess parents and children's willingness to participate.  Those who agreed to participate in the study were scheduled for an in-office or in-home measurement visit. The study recruited participants from September 2007 to January 2009. Children and parents had to meet four conditions to be eligible to participate in the study: i) live in the neighbourhood at least five days per week; ii) be able to participate in moderate-intensity physical activity; iii) be free of underlying medical conditions related to obesity; iv) not be involved in medical treatment that had a substantial impact on growth (e.g., growth hormone treatments). Additionally, parents needed to be legal guardians, and only one child per household could participate. 4.2.1.3 Data Collection Instruments 4.2.1.3.1 Accelerometers to Measure Objective Physical Activity Children were asked to wear an accelerometer (MTI GT1M Actigraph accelerometer) for seven days in a row for at least 10 hours per day during waking hours. After the participants returned the accelerometer, data were downloaded and screened to filter invalid data. Valid data included hours that did not have any zero counts for more than 20 minutes in a row. Children were asked to re-wear the accelerometer if it was not worn for at least ten valid hours on at least six days, including one weekend day. 4.2.1.3.2 Survey Questionnaire Parents completed a separate questionnaire that asked questions regarding eating habits, physical activity, and various psychological and social (psychosocial) factors related to physical 86  activity (see Appendix A  for questions used in this study).4 Parents reported how often their child walked or biked to various locations, including recreation centres and friends’ houses. The intra-class coefficient for the survey questions ranged from 0.41 to 0.61 (Joe et al., 2012b).  Similarly, parents completed five questions about how often their child was physically active in settings near their home, such as on a nearby street, sidewalk, or cul-de-sac for neighbourhood physical activity. The intra-class correlation coefficient for the survey questions ranged from 0.64 to 0.88 (Joe et al., 2012b). The questionnaire's reliability has been tested, and it has shown good test-retest reliability (Grow et al., 2008).  The psychosocial variables included factors related to parent’s perceptions of safety and parental support for physical activity. Safety-related questions were adapted from the Neighbourhood Environment Walkability Scale for Youth (NEWS-Y)(Rosenberg et al., 2009). Perception of safety included questions related to traffic safety (3 items), pedestrian safety (3 items), crime safety (1 item) and stranger danger (4 items). Test-retest intra-class correlation coefficients of these items ranged from 0.61 to 0.78 (Rosenberg et al., 2009). Parents also responded to four questions, which asked how often they or another adult in the household supported their kids to participate in physical activity. The questions were adapted from previous work by which showed a test-retest reliability of 0.79 (Sallis et al., 1987).   4 The full survey questionnaire can be found at http://www.seattlechildrens.org/research/child-health-behavior-and-development/saelens-lab/measures-and-protocols/  87  4.2.2 Teen Environment and Neighborhood (TEAN) 4.2.2.1 Neighbourhood Selection The TEAN used the walkability index and median income to create the 2x2 matrix for neighbourhood stratification. This method is similar to the one used by Frank et al. (2010) in the NQLS study. Census block groups were used as neighbourhood boundaries. Block group income was based on 2000 census data. Census block groups with income below $50,907 for the Baltimore-Washington DC four-county study region and below $56,676 for Seattle-King County were classified as low-income neighbourhoods. A combination of low and high neighbourhood walkability and income was designed to recruit participants for the study. Unlike the NQLS study, TEAN study included all deciles for neighbourhood selection. This as was due to challenge in recruiting enough eligible participants.  4.2.2.2 Recruitment The households were randomly selected and contacted by mail and telephone for recruitment. The study purchased a list of families with adolescents 12–16 years of age from a marketing company to identify potential participants. The sampling was designed to be balanced by age and gender and represent the ethnic distributions in the study areas. The participants were ineligible if they had any conditions that would affect their ability to engage in physical activity, any eating disorders, or any developmental disabilities. A letter about the study informed parents and adolescents; eligibility was determined during telephone calls and written informed consent and adolescent assent were provided. Qualified households were mailed an accelerometer and a GPS device with instructions to wear the devices for one week during all waking hours. The GPS data were not used in this dissertation.   88  4.2.2.3 Data Collection Instruments 4.2.2.3.1 Accelerometers to Measure Objective Physical Activity Teens were required to wear the accelerometers for at least a week. Different models (models 7164/71256 or GT1M/GT3X) of accelerometers were used, but the measurements did not vary by instrument model (Cain et al., 2014). The validity of the accelerometer readings was evaluated after the respondents returned the accelerometers. Readings with 30 minutes of continuous zero counts were considered invalid. Likewise, participants with less than five days of readings were asked to re-wear the device for the number of missing days.  4.2.2.3.2 Survey Questionnaire Two questionnaires were sent to each household—one for adolescents and one for a parent—asking various questions related to physical activity, dietary habits, socio-demography, and various psychosocial factors.5 Adolescents’ transport-related physical activities were assessed through nine questions inquiring about the frequency of transportation by active modes (walk or bike) to several places such as indoor recreation or exercise facilities, a friend or relative’s house, outdoor recreation facilities, food stores or restaurants/cafes, other retail stores, non-school social or educational activities, public transportation stops, work, and other places. The questions were adapted from the study by Frank et al., 2001. The test-retest ICCs of the questions range from 0.47 to 0.82 (Cerin et al., 2014). Likewise, leisure and neighbourhood physical activity was assessed using five questions asking how often they were physically active in locations near home, such as on a nearby street, sidewalk, or cul-de-sac. The items used to measure physical activity have been tested for reliability, and the results show fair to good  5 The full survey questionnaire can be found at http://sallis.ucsd.edu/measure_tean.html.  89  reliability of the questions (Grow et al., 2008). The test-retest ICC's range from .31 to .82. (Cerin et al., 2014; Joe et al., 2012a).  In addition to physical activity, adolescents and their parents were also asked questions about the perception of safety and parental support for physical activity. Perception of safety was reported by parents, whereas teens reported parental support for physical activity. The perception of safety was measured using the Neighborhood Environment Walkability Scale for Youth on four scales. Like the survey of children (NIK study), parents reported safety perceptions related to traffic, pedestrian, crime, and stranger danger. The test-retest intercorrelation coefficients (ICCs) range from 0.61 to 0.78 (Rosenberg et al., 2009) .  Parental support for physical activity included three items measured on a five-point ordinal scale ranging from 0 to 4 (0=never, 1=rarely, 2=sometimes, 3=often, 4=very often). The items captured information such as how often an adult in the household encouraged participating in sports or physical activity, provided transportation to play sport, etc. These items are based on prior literature (Sallis et al., 2002) and have an ICC of 0.79 (Cerin et al., 2017).  4.2.3 Senior Neighborhood Quality of Life Study (SNQLS) 4.2.3.1 Neighbourhood Selection Neighbourhoods were selected following a similar approach to the one used in the TEAN and NQLS studies. Census block groups were chosen to represent high and low walkability based on the walkability index and high and low-income levels based on the income data (A. C. King et al., 2011). Census block with income below $56,676 were grouped as low income neighbourhoods. Similar to the TEAN study, all income and walkability deciles were included to create high and low categories. A 2x2 matrix with quadrants of high/low walkability and 90  income—high walkable/high income, high walkable/low income, low walkable/high income, and low walkable/low income—was used to recruit participants for the study.  4.2.3.2 Recruitment After the neighbourhoods were identified, the contact information of potential participants was purchased from a marketing company. Letters with a description of the study were mailed to eligible households within the selected neighbourhoods. Eligible participants had to be 66 years of age or older, able to complete the survey in English, and able to walk more than 10 feet at a time independently. If the initially contacted person did not agree or was not eligible to participate in the study, another adult from the same household was invited. Potential participants were contacted by phone to review the study’s purpose and to exclude individuals with cognitive impairments that could affect their ability to perform tasks required for the study. An accelerometer was mailed to each qualified participant, who wore it for at least seven days. At the end of the seventh day, they filled out a survey by mail, online, or telephone interviews. The second survey was conducted six months after the first survey to reduce the burden on participants and control for variation in physical activity due to weather.  4.2.3.3 Data collection instruments 4.2.3.3.1 Accelerometers to Measure Objective Physical Activity Each participant wore the accelerometers (model 7164 or 71256; Actigraph, LLC, Fort Walton Beach, Florida) on an elastic belt with the device positioned over their right hip for seven days. This was done at two time points six months apart. Participants were asked to re-wear the accelerometer if an assessment contained less than five valid days or less than six valid hours across seven days, or if the accelerometer had malfunctioned during data collection (King et al., 2011). Accelerometers collected data in 60-s epochs.  91  4.2.3.3.2 Survey questionnaire Participants completed the survey questionnaire asking about their demographic information, physical activity, and various psychosocial factors.6 Participants completed a modified version of the CHAMPS questionnaire, which includes items specific to walking for transportation and leisure. The validity of the CHAMPS questionnaire has been extensively evaluated for the older population (STEWART et al., 2001) and has shown good test-retest reliability (r = 0.76). For transport-related walking, participants were asked two separate questions about the number of times in a week they walked and biked to or from various locations to do errands. For leisure walking and walking for exercise, they were also asked about the time they walked for leisure or exercise each week.  Psychosocial factors included perception of safety, social support for physical activity, and self-efficacy. Perception of safety was assessed using four-point ordinal scale measuring three aspects: i) traffic safety (3 items), ii) pedestrian safety (7 items), iii) personal safety (7 items). All subscales have shown moderate to high test-retest reliability (α = 0.58 to 0.80) (Saelens, Sallis, Black, et al., 2003). Social support for physical activity was measured by asking participants to rate support from their family and friend to do physical activities using two separate sets of questions. Each set contained four items measured on an ordinal scale (0= Never, 4= Very often) (α=0.67) (Sallis et al., 1987). Participants were asked to rate their confidence to walk ½ block, four blocks, and ten blocks on a 10-point scale. The measure of self-efficacy for walking has shown good internal consistency (α = 0.90) and test-retest reliability ( r = 0.67) for older adults (Garcia & King, 1991) to measure self-efficacy for physical activity.   6 The full survey questionnaire can be found at http://sallis.ucsd.edu/measure_snqls.html  92  4.3 Variables used in this study 4.3.1 Dependent Variables Three different measures of physical activity were used for this study: objective physical activity, self-reported transport-related physical activity, and leisure and neighbourhood physical activity. Because of the skewness in the distribution of the objective physical activity (see Appendix E for data distribution) and the ordinal scale used to capture self-reported physical activity (walking/biking) in kids and teens, binary variables were used.  4.3.1.1 Self-Reported Physical Activity 4.3.1.1.1 Transport-Related Physical Activity For the NIK and TEAN cohorts, transport-related physical activity was assessed using questions that asked whether the participants walked or biked to nine common locations (including recreation centres and a friend’s house) in their neighbourhoods. The responses were recorded on an ordinal scale. The SNQLS cohort was asked to report the number of times they walked or biked for transportation. All the variables were dichotomized into binary outcomes. For SNQLS and NIK data, the final measure indicated whether the cohorts walked or biked for transport. For TEAN data, it indicated whether the participants walked or biked to five or more destinations in their neighbourhood. (Refer to the Appendix A  for the questionnaire).  4.3.1.1.2 Leisure and Neighbourhood Physical Activity The SNQLS cohort reported the number of times they walked for leisure during a week. And the NIK and TEAN cohort reported how often they were physically active in settings near their homes, such as nearby streets, sidewalks, or cul-de-sacs. The NIK and TEAN questionnaires are based on a scale with good test-retest reliability (Grow et al., 2008) (Refer to the Appendix A  for the questionnaire used for each cohort). These indicators were also 93  categorized into binary variables reflecting the older adults walked for leisure on not (SNQLS data) and whether children and teens are active in 3 or more locations near their homes.  Children and teens who were active in 3 or more location near their home are referred as “active in neighbourhood.” 4.3.1.2 Objective Physical Activity Objectively measured moderate to vigorous physical activity (MVPA) was used as an outcome variable. Even though there is increasing evidence on the potential health benefits of light-intensity physical activity (Lawman & Wilson, 2014; Loprinzi, 2013), MVPA has been consistently found to have significant health benefits and, in some cases, better health outcomes than light-intensity physical activity (García-Hermoso et al., 2017). For children and teens, MVPA was scored using Freedson youth age-specific cut-points with a 4-MET moderate-intensity threshold (TROST et al., 2002). Since school-time physical activity could inflate the results, only physical activity data from non-school hours were used, i.e., 3 pm to 11 pm on weekdays and whole days on the weekend. For older adults, accelerometer data were converted to minutes in MVPA using the Freedson adult cut-point, 3-MET (Freedson et al., 1998), and average daily minutes of MVPA were calculated. The resulting data were converted into a binary variable using 30-minute cut-points for older adults and teens and 60-minute cut-points for children. The cut-points of 30 minutes per day and 60 minutes per day are based on the guidelines recommended for engaging in physical activity for adults and children, respectively. The TEAN cohort had very few participants who met the 60-minutes-per-day threshold. Therefore, 30 minutes per day was used as a threshold to create a binary dummy variable.  94  4.3.2 Independent Variable 4.3.2.1 Neighbourhood Walkability All three studies (NIK, TEAN, and SNQLS) used the walkability index developed by Frank et al. (2010) to measure neighbourhood built environment. The neighbourhood boundary in this study is defined as the 1 km street network buffer around a participant’s place of residence. Figure 4-1 shows a diagram of the 1-km street-network buffer (solid blue line) that was used to develop the walkability index.  The walkability index is a function of four components—net residential density, commercial floor area ratio (FAR), land-use mix and intersection density—which captures whether a neighbourhood has a physical environment that supports walking. Net residential density measures the dwelling unit concentration within a specific area. Higher net residential density values indicate a higher number of dwelling units relative to the residential land area. Commercial FAR is the ratio between the total floor area of a building to the land area. Commercial FAR allows researchers to capture the three-dimensional urban form. Land-use mix measures the balance between the building floor areas of six land uses Figure 4-1. Street network buffer used to compute walkability index (Source: Frank et al., 2013). 95  (retail, entertainment/recreation, civic/educational, office, single-family residential, and multi-family residential). Values range from 0 to 1, with values closer to 1 indicating greater evenness among the six land uses (Frank et al., 2005). Intersection density is a measure of street connectivity within a given area. Smaller city blocks tend toward higher intersection density values and increased ease of travel to destinations. The walkability index is computed by adding up the normalized scores of residential density, commercial FAR, intersection density, and land-use mix within a 1-km street network buffer around a postal code. Likewise, the buffer can be created using the exact address or postal code centroid (Figure 4-1).  4.3.2.2 Street-Scale Built Environment Features The street-scale built environment data for the three studies were collected using the Microscale Audit of Pedestrian Streetscapes (MAPS) (Millstein et al., 2013). MAPS was developed to collect audit data on the pedestrian environment and walkability in neighbourhoods. MAPS is primarily an observation-based street audit tool adapted from the Analytic Audit Tool (Brownson et al., 2004) and the findings of the Healthy Aging Network (Kealey et al., 2005 as cited in Millstein et al. 2013) on older adults.  MAPS comprises three main sections: overall route, street segments, crossings (Table 4-1). Cul-de-sac is an additional section used for children and teens. Route-level variables show the characteristics of the whole route, such as land use and destinations, street amenities, and aesthetic and social aspects. Each segment includes the street section on the route between street intersections. Segment-level variables include sidewalks, slope, buffers between street and sidewalk, bicycle facilities, building aesthetics, trees, setbacks of buildings from the sidewalk or street, and building height. Street-crossing variables are measured at every intersection or crossing along the route. Variables measured at street crossings include crosswalk markings, the 96  width of crossings, curbs, crossing signals, and pedestrian protection. Cul-de-sacs include variables like size and condition of the surface area, slope, and amenities within cul-de-sacs; these data were collected only when one or more cul-de-sacs were present within 400 meters of the participant’s home. Table 4-1 MAPS section descriptions. (Source: Millstein et al., 2013) 4.3.2.2.1 Data Collection Method The MAPS data were collected along a ¼-mile route starting at a study participant’s home and walking toward the nearest pre-determined destination, although not necessarily reaching a destination. If the destination was reached before the ¼-mile mark, raters continued toward the next identified destination and ended ratings there. If the destination was not reached in ¼ mile, the route ended after the segment that included the ¼-mile endpoint.  Microscale section Description Route Approximately ¼ mile from a participant’s home toward a predetermined destination. Included components of land use and destinations, streetscape, aesthetics, and social environment Consisted of varying numbers of segments and crossings within the ¼ mile. Segment A section of a street between two crossings If street name changed, a new segment started. There were up to 8 segments per route Crossing A crossing occurred when the rater went through an intersection, whether a pedestrian crossing existed or not. There were up to five crossings per route. Cul-de-sac* A cul-de-sac or dead-end street had to be within 400 feet of a participant’s home. The cul-de-sac was usually (but not always) the dead-end part of the participant’s street There were up to two cul-de-sacs per route. Note: * Cul-de-sac data is collected only of children and teens. Cul-de-sac score is not included in MAPS grand score calculation.    97  Table 4-2: Destinations definition for MAPS survey.  Cluster type Description Retail cluster Convenience store, market, grocery, specialty, dollar, retail, warehouse, pharmacy, video stores, fast food, restaurant, coffee, bagel, buffet, fast casual restaurants FROM food enumeration databases AND civic, doctor-dentist, entertainment, large retail, museum, neighbourhood retail, recreational, and super large retail FROM tax assessor parcel data. Auto parts stores and liquor stores were removed. Does not include parks or schools. Park  One acre or larger in size and its boundary intersects a participant’s one-kilometre road-network based buffer around their home. School type Elementary, middle and high schools, college/university, parochial/religious. Source: https://drjimsallis.org/Documents/Measures_documents/MAPS%20Manual_v1_010713.pdf Data were collected during 2009 and 2010 by certified raters after rigorous 3-days training. Raters were certified only after they were able to successfully assess a minimum of four routes with inter-rater reliability ≥ 95%. Data were collected starting for a participant’s home and walking towards a predetermined destination. As shown in Table 4-2, destinations include commercial retail clusters, parks, schools or other destinations. Eligible destinations for NIK, TEAN and SNQLS included clusters of more than 3 commercial locations, parks, schools. The raters walked on a designated route along the same side of the street. Route level items were collected for the entire ¼ mile. Segment level information was collected for each street segment along the route; a similar approach was taken for crossing that came along the route. Any cul-de-sacs or dead-end street within 400 feet from a participant’s home were also audited using the cul-de-sac section. Cul-de-sac data were collected only for children and teens. 4.3.2.2.2 The Conceptual Approach to the Scoring System The scoring system used by MAPS is based on expert consensus from the study team members (MAPS principal investigators), prior literature and evidence supporting factors 98  thought to influence people’s perceptions of their physical activity environments. The factors include safety, aesthetics, destinations, arterial or thoroughfare roads, land use, recreational facilities, transportation environment, and social environment/physical incivilities. Using this framework resulted in a hierarchical scoring system (Figure 4-2; Figure 4-3; Table 4-3). All sections have positive and negative valence scores based on the expected effects on physical activity (e.g., crosswalk amenities are positive features, and crossing impediments are negative features). Negative valence scores (higher scores indicate more negative attributes) are subtracted from positive valence scores (higher scores indicate more positive attributes) to create overall section scores. MAPS items and subscales have demonstrated predominantly moderate to excellent inter-rater reliability (Millstein et al., 2013).  4.3.2.2.3 Converting Item Scores to Subscales and Valence scores7 All the items level data for the four main MAPS sections were sorted into groups through consensus among the core investigators and prior literature. The core investigators also established a scoring convention to simplify the scoring process. Most of the data were collected using a binary scale to reflect the presence or absence of street design features (e.g., presence of trash bin, benches, bike racks, etc.). Some items measured counts using an ordinal scale, such as the number of driveways or alleys in a route (none, 1-2, 3-5, 6+). In such cases, data were dichotomized or trichotomized based on the prior literature and the appropriateness of the items in the group. In some cases, there was scores were weighted as well. This was done to calculate residential mix used to construct positive land use subscale (Millstein et al., 2013).    7 Full information about data scoring and syntax is available in following link: https://drjimsallis.org/Documents/Measures_documents/MAPS_Full_DataDictionary.pdf  99  Subscales were created based on their expected relation with physical activity—positive or negative association. For example, the sum of the score for crosswalk amenities, good curb quality and having intersection control sign was combined to create a positive crossing sub-scale. The negative sub-scale for crossing included the combined score for road width and crossing impediment. Sub-scale items were added to create positive and negative valence scores. Positive sub-scales were used to create positive valence scores, whereas negative sub-scales were added to create negative valence scores. Overall section scores were calculated by subtracting negative valence scores from positive valence scores. For route, this resulted in three sub-section scores, i) Destination and Land use score, ii) Streetscape score, iii) Aesthetics and Social score. The sub-section scores were added to calculate the overall score for the route. The total MAPS score was calculated by adding the overall score for route, crossing and segment. The cul-de-sac items were not included in the maps score because of the lack of consensus on their relationship with physical activity (Millstein et al., 2013).  4.3.2.2.4 Domain-Specific Scores The tiered scoring system used by the MAPS tool allows creating behaviour-specific scores (Cain et al., 2018). Two separate scores were calculated, i) MAPS score for active transport, ii) MAPS score for leisure walking. The MAPS score for active transport was calculated by adding positive crossing, positive segment, positive streetscape (no Aesthetics/Social) and positive land use/destination. Similarly, the MAPS score for leisure walking was created by combining the scores for public rec land uses, private rec land uses, positive aesthetics/social, sidewalks, buffers, bike infrastructure, building aesthetics/design and trees. Both these scores were based on the consensus among the key researchers of the MAPS study.  100   Figure 4-2 MAPS scoring structure and summary of inter-rater reliability: Route section. (Source: Millstein et al., 2013)   101   Figure 4-3 MAPS scoring structure and summary of inter-rater reliability: Segments and Crossings sections (multiple surveys per route). (Source: Millstein et al., 2013)   102  Table 4-3. MAPS valence and subscales. Route Street segments Crossings/ intersections Cul-de-sac Destinations  & land use Streetscape characteristics Aesthetics & social characteristics       Positive characteristics Positive characteristics  • Residential mix   • Building height-setback • Crosswalk amenities  • Shops   • Sidewalk • Curb quality  • Restaurant/entertainment   • Buffer • Intersection control  • Institutional service   • Bike infrastructure   • Government service   • Building aesthetics/design   • Public recreation   • Trees   • Private recreation   • Building height-road width ratio   • Parking   Negative characteristics Negative characteristics  • Transit Stops   • Sidewalk obstructions/hazards • Road width     • Slope • Impediments  Valence and  overall score Valence and  overall score Valence and  overall score Valence score Valence score  • Positive  • Positive  • Positive  • Positive  • Positive   • Negative • Negative • Negative • Negative • Negative  • Overall score • Overall score • Overall score    Overall score route Overall score segment Overall score crossing Overall score cul-de-sac MAPS Grand Score   103  4.3.2.3 Psychosocial Variables 4.3.2.3.1 Parents’ Support for Physical Activity (For NIK and TEAN Sample) Parents’ support for physical activity was measured using four questions in the NIK cohort and three questions in the TEAN cohort. Parents were asked how often in a week they or any adults in the household encouraged their children to do physical activity, participated in physical activity with their children, provided transportation to get to a location for physical activity, and watched their children do physical activity. The ordinal scale was used to measure the items, and a composite score was created by averaging the responses.  4.3.2.3.2 Self-Efficacy for Walking (For SNQLS Sample) Self-efficacy for walking in older adults was measured by three items in which participants reported their levels of confidence to walk a half-block, four blocks, and ten blocks on a 10-point scale (from “not confident at all” to “absolutely confident”). These items have shown good reliability and internal consistency (Garcia & King, 1991). An average score from the three items was calculated for the analysis. 4.3.2.3.3 Social Support for Physical Activity (For SNQLS Sample) Social support for physical activity was assessed by two categories of items: family support and support from friends, acquaintances, or co-workers. A previous study has validated the items (Sallis et al., 1987). Family support for physical activity was measured by four questions which asked participants to rate on a five-point scale (“never” to “very often”) how often during the past six months their family i) walked or exercised with them, ii) encouraged them to do physical activity, iii) made positive comments about the participant’s physical appearance, and iv) criticized or made fun of them for walking or exercising. Similar items were 104  used to assess support from friends, acquaintances, or co-workers. The responses were averaged to create a composite score.  4.3.2.3.4 Parents’ Safety Perceptions (For NIK and TEAN Sample) Parents’ perception of neighbourhood safety was assessed using four scales adapted from the Neighborhood Environment Walkability Scale for Youth (Rosenberg et al., 2009) with an additional measure of stranger danger. Three items measured traffic safety: i) Traffic makes it unpleasant to walk, ii) the speed of traffic on most streets is usually slow, and iii) most drivers go faster than the posted speed limits. Likewise, pedestrian safety was assessed by three items: i) Streets have good lighting at night, ii) walkers and bikers can be easily seen by people in their homes, and iii) there are crosswalks and signals on busy streets. The parents’ perception of crime rates was used to measure safety from crime. Stranger danger safety was measured by four items that asked about the fear of being taken or hurt by a stranger in various places: i) in a local park; ii) on local streets; iii) in my yard, driveway, or apartment common area; or iv) in my neighbourhood. Response options ranged from 1 (strongly disagree) to 4 (strongly agree), with higher numbers representing perceptions of greater safety except for stranger danger, in which higher numbers reflect the opposite. The average scores of items within each category were used for the analysis.  4.3.2.3.5 Safety Perception (For SNQLS Sample) Older adults responded to various questions about their perceptions of traffic safety, pedestrian safety, and safety from crime based on a modified version of the Neighborhood Environment Walkability Scale targeted for older adults. Traffic safety was assessed by a three-item scale that asked participants to rate their perception of the quantity and speed of traffic on their neighbourhood streets. Pedestrian safety was assessed using seven items that asked 105  participants to rate their perceptions of how safe it is to walk in their neighbourhoods, specific to issues such as safe intersections, sidewalks, and crosswalks. The mean of each safety measure was used for analysis.  4.3.2.4 Control Variables  4.3.2.4.1 Socio-Demography Factors Age, gender, income, and ethnicity were used as socio-demographic variables. For NIK and TEAN, these variables were reported by parents, whereas the SNQLS participants self-reported the information.  4.3.2.4.2 Self-Reported Mobility Impairment  SNQLS participants also self-reported any lower-extremity mobility impairment, which was assessed using the previously validated Late-Life Function and Disability Instrument (Sayers et al., 2004). Eleven items were used to measure mobility impairment. An average of the items was used for the analysis.  4.3.2.4.3 Weather Patterns Temperature and precipitation levels were obtained using the dates when participants wore the accelerometers. Daily temperature and precipitation data were obtained from the National Oceanic and Climate database (Climate Data Online, 2020), which provides historical daily temperatures from various weather stations. For each cohort, the weather data from all the relevant weather stations were collected, averaged, and linked with the dates when the participants wore the accelerometers.  4.3.2.4.4 Neighbourhood Self-Selection Neighbourhood self-selection was assessed by asking the cohort if they had chosen their neighbourhood because of the ease of walking it offered. In the NIK and TEAN cohort, parents 106  responded to this question, whereas in the SNQLS cohort, the response was given by the participants themselves. SNQLS cohort had a larger number of missing data and was not recommended to be used in the analysis by the study team.  4.3.2.4.5 Regional Transit Accessibility The regional transit accessibility score measured the number of regional destinations accessible within 45 minutes, more than 90% of the time during morning peak hours. The bus and rail service frequency and travel time were obtained from the historic General Transit Feed Service data, which is publicly available data (GTFS Data Exchange, 2016). The 1-minute time gap was used for each departure between 7 am and 9 am. The number of destinations accessible within 45 minutes was counted for each departure, and there were 120 departures. The results from each departure were used to calculate the number of regional destinations accessible within 45 minutes for at least 90% of the total departures. The regional destinations were identified using the regional planning documents for each of the study regions. 4.4 Data Analysis Plan 4.4.1 Overall Analysis Plan Because of the differences in study design and measurement techniques, the data for the three age groups were analyzed separately. For each of the three datasets, three outcomes were analyzed in separate models. Figure 4-4 shows a framework for grouping the data for analysis. Regressions models were built in three steps (Figure 4-5) for each outcome. Two-way interaction between neighbourhood walkability and MAPS grand was tested in the first step. In the second step, hypotheses related to the second objective were tested. This included three-way interactions between i) walkability index, maps grand and gender, and ii) walkability index, maps grand and neighbourhood income. Since the data on children (NIK study) did not have neighbourhood 107  income information, household income was used to test the interaction. The interactions between walkability index, MAPS grand and various psycho-social factors (objective 3) were tested in the third step. The psycho-social factors included social support for physical activity and perception of safety. The final models included only significant interactions, i.e., psycho-social factors that did not have significant interactions were not included as control variables in the final model. However, gender and income were included final model regardless of their interaction with walkability and MAPS grand score.  In addition to the overall MAPS score, behaviour-specific (also called domain-specific) MAPS scores were also used to examine the hypotheses. These included MAPS scores related to active transportation and MAPS scores related to leisure walking. The MAPS score for active transport was used for walkability or biking for transport, whereas the MAPS score leisure was used for leisure walking/walking for exercise (SNQLS) and being active in the neighbourhood (NIK and TEAN). Both scores were used in the cases of objective physical activity. Domain-specific analysis followed the same approach used for the MAPS grand score.          108       Figure 4-5 Steps for testing interactions based on study objectives and hypotheses. Figure 4-4. Data grouping plan for datasets and outcome of interest. 109  4.4.2 Supplementary Analysis Plan The interactions between neighbourhood walkability and lower-order MAPS scores (valence scores and subscales) were tested to inform the MAPS grand score results. This was done after the final model using MAPS grand scores was identified. For example, if the final model using the MAPS grand score found a significant interaction between neighbourhood walkability, MAPS grand score, and gender, separate models were further developed to examine the interactions between neighbourhood walkability, gender, and the lower-order MAPS scores. The lower-order MAPS scores include valence scores and subscale scores.  4.4.3 Data Analysis 4.4.3.1 Regression Analysis Bivariate relationships between the dependent variables and independent variables were evaluated through multiple t-tests and chi-square tests. Separate logistic regression models were created for the three outcomes. A hierarchical modelling framework was used to account for the potential clustering effect among the participants sampled from the same census block. All continuous variables were mean centred. A stepwise regression method was used to develop the final model. Generalized linear models were built using the Penalized Quasi-Likelihood (PQL) method. The main effect models for the walkability index and MAPS grand score were fitted before testing the hypotheses. Next, the models for each hypothesis were tested incrementally. Significant interaction effects in preceding hypotheses were included in the experimental hypothesis (Figure 4-5). A p-value less than 0.05 was used as a cut-off threshold for significance. All significant results were adjusted for the False Discovery Rate (FDR) to account for the bias resulting from multiple tests. Benjamini-Hochberg method was used to adjust for the FDR (Benjamini & Hochberg, 1995). Only those results with FDR adjusted p-values less than 0.05 are 110  presented and discussed. For valence score and subscale analysis, a lower threshold of 0.01 was used to avoid chances of false positive results.   4.4.3.2 Model Fit and Robustness Test For each model, a Recursive Operating Curve (ROC) curve was created, and the Area Under the Curve (AUC) was calculated and reported, which is recommended for logistic regression (Monsalves et al., 2020). Higher AUC represents higher model accuracy (Hanley & McNeil, 1982), but it also requires additional analysis like setting up threshold value to measure precision level. For this study, only overall AUC was calculated.  Additionally, Variance Inflation Factors (VIF) was used to assess multicollinearity in the final model. As a general rule of thumb, a threshold of 10 was used as a cut-off for high collinearity is used by various studies (Dormann et al., 2013; Kutner et al., 2005). However, some studies have also suggested using a relatively smaller cut-off value (O’Brien, 2007; Vatcheva et al., 2016). A higher threshold of 10 was used because this study uses interaction terms that tend to have higher VIF. High VIF in models involving interaction and product terms is not usually a problem (Allison, 2012). VIF for the final model were calculated in two ways, with and without interaction terms. The results are included in the appendix (Appendix C  ) as supplementary material.  For all final models using MAPS grand score, separate models were fitted using different R-packages to check model sensitivity. Sensitivity tests were done using the Laplace approximation technique and the gauss-Hermite quadrature techniques that have better accuracy over the PQL method but are a less flexible method and often have convergence problems (Bolker et al., 2009). The results of these models are reported as supplementary material in the 111  appendix (Appendix D  ). Sensitivity tests were not done for domain-specific MAPS scores and MAPS subscale and valence score results.    4.4.3.3 Visualizing Interaction Effect All significant interactions were plotted using the log-odds as well as the predicted probabilities. These plots and graphs were created by holding categorical covariates at the reference group and continuous covariates at the mean.  4.5 Summary This chapter discussed the data from three studies used in this dissertation: Neighbourhood Influences on Kids (NIK), Teen Environment and Neighbourhood (TEAN), and Senior Neighbourhood Quality of Life Study (SNQLS). These studies were conducted in Seattle, San-Diego, and Baltimore regions in the United States of America using a similar study design and sampling strategies to those used by Frank et al. (2010). The studies collected both self-reported and objective measures of physical activity along with data on neighbourhood walkability and pedestrian environment. The following two chapters present the results of the analysis, discuss the implication for planning and policymaking.          112  CHAPTER 5: RESULTS: DESCRIPTIVE STATISTICS 5.1 Overview This chapter summarizes the results of exploratory data analysis. It provides a descriptive summary of various demographic variables, key predictors, and moderators corresponding to each dataset. The results of two-sample tests and correlations between moderators are also presented. The chapter is divided into three sections that match the datasets: NIK, TEAN, and SNQLS. The results are organized in two subsections within each section: i) descriptive statistics and ii) bivariate relationship between the outcome and key variables.  5.2 Sample 1: Neighborhood Influences on Kids (NIK) 5.2.1 Descriptive Statistics Table 5-1 shows a summary of participant characteristics. A total of 757 children participated in the study, and the average age of participants was about nine years. A high proportion of the participants were white (81.77%). There was a balance in gender proportion; however, most children were from high-income households (80.79%). Almost half of the participants’ parents (45.45%) reported that they chose their current neighbourhood because of the ease of walking. The average score for parental support for physical activity was 2.88 (SD = 0.73). Likewise, the mean scores for the parents’ perception of traffic safety, pedestrian safety, crime in the neighbourhood, and stranger danger were 2.55 (SD = 0.61), 2.71 (SD = 0.65), and 3.32 (SD = 0.77), and 2.16 (SD = 0.73) respectively. About 83% of the children walked or biked for transportation; more than half (about 55%) of the children were active in the neighbourhood, such as playing in neighbourhood streets, sidewalks, etc. Less than a quarter (22.04%) of the total number of children spent at least 60 minutes doing moderate to vigorous physical activity daily during non-school hours.  113  Table 5-1 NIK: Descriptive summary of demography, psychosocial factors, outcome and other covariates.  Variables Mean (Standard Deviation)/ Count (Percentage) (N = 757) Age (Range 6 -11) 9.16 (±1.59) Ethnicity Non-white 138 (18.23%) White 619 (81.77%) Gender Male 378 (49.93%) Female 379 (50.07%) Household Income Low (< $ 60,000)  136 (19.21%) High (≥ $ 60,000) 572 (80.79%) Neighbourhood preference for walking No 396 (54.55%) Yes 330 (45.45%) Average Temperature (Celsius) (Range: -8 – 28) 10.10 (±8.13) Average Precipitation (Millimeters) (Range: 0 – 22.89) 3.24 (±4.44) Regional Transit Accessibility (Range: 0 – 6) 0.61 (±1.14) Parental Support for Physical Activity (Range: 1 – 5) 2.88 (±0.73) Perceived Traffic Safety (Range: 1 – 4) 2.55 (±0.61) Perceived Pedestrian Safety (Range 1 – 4) 2.71 (±0.65) Perceived Crime (Range: 1 – 4) 3.32 (±0.77) Perceived Stranger Danger (Range: 1-4) 2.16 (±0.73) Study Site San Diego 364 (48.08%) Seattle 393 (51.92%) Walk or Bike for Transport No 117 (16.27%) Yes 602 (83.73%) Neighbourhood Physical Activity Not Active in Neighbourhood 318 (44.23%) Active in Neighbourhood 401 (55.77%) Non-school Hours Daily Moderate to Vigorous Physical Activity (MVPA) (≥ 60 minutes) No 566 (77.96%) Yes 160 (22.04%) Note: Number of missing observations: Household Income = 49; Neighbourhood Preference for Walking = 31, Average Temperature and Average Precipitation = 55; Regional Transit Accessibility = 4; Parental Support for Physical Activity = 39; Perceived Traffic Safety = 36, Perceived Pedestrian Safety = 38; Perceived Crime = 38, Perceived Stranger Danger = 37; Walk or Bike for Transport= 38; Neighbourhood Physical Activity = 38, Non-School Hours Daily MVPA = 31.  114  5.2.2 Bivariate Relationship  Two-sample tests were conducted to examine the group differences in neighbourhood walkability, MAPS grand score, and other key variables based on the outcomes: i) active transport, ii) neighbourhood physical activity, and iii) MVPA non-school hours. Table 5-2 shows the results of active transportation. As seen in the table, the mean score for neighbourhood walkability and MAPS score was higher for the groups who walked or biked for transport, and the difference in the mean was significant. There was no significant difference in gender and household income between the two groups. There was a significant difference in mean scores for the parents’ perceptions of traffic safety, pedestrian safety, and crime between the two groups. These scores were higher in the groups that walked or biked for transportation. The difference in perception of stranger danger was marginally significant. Parental support for physical activity was not significantly different between the two groups.  Table 5-2. NIK: Bivariate relationship between neighbourhood walkability, MAPS scores, and other key variables based on active transport. Variables Non-walkers or non-bikers n= 117 Walkers or bikers n= 602 p-value Neighbourhood Walkability -0.69 (±2.39) 0.15 (±2.88) 0.003 MAPS Grand Score 12.66 (±5.04) 13.93 (±5.47) 0.02 Gender   0.84 Male 58 (49.57%) 305 (50.66%)  Female 59 (50.43%) 297 (49.34%)  Household Income   0.8 Low (< $ 60,000)  23 (20.00%) 112 (18.98%)  High (≥ $ 60,000) 92 (80.00%) 478 (81.02%)  Parental Support for Physical Activity 2.77 (±0.72) 2.90 (±0.72) 0.064 Perceived Traffic Safety 2.44 (±0.66) 2.57 (±0.60) 0.038 Perceived Pedestrian Safety 2.51 (±0.71) 2.74 (±0.63) 0.0003 Perceived Crime 3.09 (±0.90) 3.36 (±0.74) 0.0006 Perceived Stranger Danger 2.28 (±0.78) 2.14 (±0.72) 0.060 115  The differences in neighbourhood walkability, MAPS grand score, and other key variables between those who were active in the neighbourhood and those who were not active in the neighbourhood are shown in Table 5-3. There was a significant difference in neighbourhood walkability and MAPS grand score between the children who were active in their neighbourhoods and those who were not active in their neighbourhoods. Children who were active had lower neighbourhood walkability (mean = −0.31, SD = 2.65) compared to those who were not active in their neighbourhoods (mean = 0.38, SD = 2.97). However, the MAPS grand score was higher for the children who were not physically active in their neighbourhoods. There were significant differences in household income, parental support for physical activity, and parents’ perceptions of traffic, pedestrian safety, crime in the neighbourhood, and perceived stranger danger between the two groups. Table 5-3. NIK: Bivariate relationship between walkability, MAPS Grand Score, and other key variables based on neighbourhood physical activity. Variables Not Active in Neighbourhood (n = 318) Active in Neighbourhood (n = 401) p-value Neighbourhood Walkability 0.38 (±2.97) -0.31 (±2.65) 0.001 MAPS Grand Score 14.20 (±6.02) 13.30 (±4.80) 0.026 Gender   0.11 Male 150 (47.17%) 214 (53.37%)  Female 168 (52.83%) 187 (46.63%)  Household Income   < 0.0001 Low (< $ 60,000)  82 (26.37%) 54 (13.71%)  High (≥ $ 60,000) 229 (73.63%) 340 (86.29%)  Parental Support for Physical Activity 2.74 (±0.73) 3.00 (±0.70) < 0.0001 Perceived Traffic Safety 2.43 (±0.64) 2.64 (±0.57) < 0.0001 Perceived Pedestrian Safety 2.60 (±0.67) 2.80 (±0.62) < 0.0001 Perceived Crime 3.19 (±0.83) 3.42 (±0.71) < 0.0001 Perceived Stranger Danger 2.29 (±0.74) 2.06 (±0.71) < 0.0001 Note: Neighbourhood physical activity refers to being active in various locations around home.  116  Table 5-4 shows the results of the group difference tests based on MVPA during non-school hours measured by an accelerometer. There were no significant differences in neighbourhood walkability and MAPS grand score between the children who spent 60 minutes or more doing MVPA daily during non-school hours compared to those who didn’t. There were significant gender and income difference between the two groups. A higher proportion of boys (66.88%) spent at least 60 minutes daily doing MVPA, whereas only one-third of the total number of girls (33.12%) did. Likewise, a higher proportion of children from high-income households (86.36%) spent at least 60 minutes each day doing MVPA outside of school. Parental support was also significantly higher among the physically active children (mean = 3.11, SD = 0.67). There were no differences in perceived traffic safety, pedestrian safety, crime and stranger danger between the two groups.   Table 5-4. NIK: Bivariate relationship between walkability, MAPS Grand Score, and other key variables based on non-school hours of daily moderate to vigorous physical activity (MVPA) measured by accelerometer. Variables MVPA < 60 min/day (n = 566) MVPA ≥ 60 min/day (n = 160) p-value Neighbourhood Walkability 0.00 (±2.77) -0.13 (±2.93) 0.59 MAPS Grand Score 13.73 (±5.43) 13.58 (±5.33) 0.76 Gender < 0.0001 Male 256 (45.23%) 107 (66.88%)  Female 310 (54.77%) 53 (33.12%)  Household Income 0.049 Low (< $ 60,000)  113 (20.77%) 21 (13.64%)  High (≥ $ 60,000) 431 (79.23%) 133 (86.36%)  Parental Support for Physical Activity 2.82 (±0.73) 3.11 (±0.67) < 0.0001 Perceived Traffic Safety 2.55 (±0.61) 2.54 (±0.64) 0.85 Perceived Pedestrian Safety 2.69 (±0.64) 2.77 (±0.68) 0.19 Perceived Crime 3.33 (±0.77) 3.30 (±0.79) 0.70 Perceived Stranger Danger 2.16 (±0.74) 2.16 (±0.70) 0.99 117  5.3 Sample 2: Teen Environment and Neighborhood (TEAN) 5.3.1 Descriptive Statistics Table 5-5 shows a summary of participant characteristics. The total sample size was 928. The participants' average age was about 14 years, and more than half were white (66.34%). There was a balance in gender and neighbourhood income in the sample. Almost half of the participants’ parents (45.36%) reported that they chose their current neighbourhood because of the ease of walking, and the average score for parental support for physical activity was 2.26 (SD=0.99). The mean scores for parents' perception of traffic safety and pedestrian safety in the neighbourhood were 2.58 (SD = 0.58) and 2.83 (SD = 0. 65). About one-third (32.721%) of the total sample walked or biked for transportation, and almost half (48.06%) of the total sample were active in the neighbourhood, such as neighbourhood streets, sidewalks, etc. However, only 17.65% of the total number of children spent at least 30 minutes doing moderate to vigorous physical activity daily during non-school hours.  Table 5-5. TEAN: Descriptive summary of demography, psychosocial factors, outcome, and other covariates. Variables Mean (Standard Deviation)/ Count (Percentage) (N = 928) Age (Range: 12 – 15) 14.09 (±1.40) Ethnicity Non-white 310 (33.66%) White 611 (66.34%) Gender Male 460 (49.57%) Female 468 (50.43%) Neighbourhood Income* Low 463 (49.89%) High 465 (50.11%) Neighbourhood preference for walking No 507 (54.63%) Yes 421 (45.37%) 118  Variables Mean (Standard Deviation)/ Count (Percentage) (N = 928) Average Temperature (Celsius) (Range: -8.89 – 27.26) 7.48 (±7.27) Average Precipitation (Millimetres) (Range: 0 – 38.64) 4.50 (±4.60) Regional Transit Accessibility (Range: 0 – 7) 0.77 (±1.44) Parental Support for Physical Activity (Range: 0 - 4) 2.26 (±0.99) Perceived Traffic Safety (Range: 1 – 4) 2.58 (±0.58) Perceived Pedestrian Safety (Range: 1 - 4) 2.83 (±0.65) Perceived crime (low crime rate) Strongly disagree 58 (6.26%) Somewhat disagree 148 (15.98%) Somewhat agree 369 (39.85%) Strongly agree 351 (37.90%) Perceived Stranger Danger (Range: 1 – 4) 1.99 (±0.73) Study Site Baltimore 485 (52.26%) Seattle 443 (47.74%) Walk or Bike for Transport No 623 (67.28%) Yes 303 (32.72%) Neighbourhood Physical Activity Not Active in Neighbourhood 481 (51.94%) Active in Neighbourhood 445 (48.06%) Non-school Hours of Daily MVPA (≥ 30 minutes) No 709 (82.35%) Yes 152 (17.65%) Note: Number of missing variables: Ethnicity = 7; Average Temperature and Average Precipitation = 48; Family Support for Physical Activity = 2; Perceived Traffic Safety = 1, Perceived Pedestrian Safety = 2; Perceived Crime = 2, Perceived Stranger Danger = 2; Walk or Bike for Transport= 2; Neighbourhood Physical Activity = 2, Non-School Hours of Daily MVPA = 67. *Low income refers to income below $50,907 for the Baltimore-Washington DC four-county study region; $56,676 for Seattle-King County     5.3.2 Bivariate Relationship The differences in neighbourhood walkability, MAPS grand score, and other key variables based on active transport (walk or bike) are presented in Table 5-6. The results show that participants who walked or biked for transport had higher neighbourhood walkability (mean = 0.77, SD = 3.27) and MAPS scores (mean = 4.70, SD = 8.28). There were gender differences 119  between the two groups, with a higher proportion of boys walking or biking (55.12%) than girls (44.88%). There was no significant difference in neighbourhood income between the two groups. Parental perceptions of traffic and pedestrian safety scores were higher among participants who walked or biked for transport and were significantly different from those who did not walk or bike for transport. Similarly, perceived stranger danger was lower for participants who walked or biked for transport.  Table 5-6. TEAN: Bivariate relationship between neighbourhood walkability, MAPS grand score, and other key variables based on active transport. Variables Non-walkers or bikers (n = 623) Walkers or bikers  (n = 303) p-value Neighbourhood Walkability -0.43 (±2.52) 0.77 (±3.27) < 0.0001 MAPS Grand Score 3.33 (±8.02) 4.70 (±8.28) 0.019 Gender 0.021 Male 292 (46.87%) 167 (55.12%)  Female 331 (53.13%) 136 (44.88%)  Neighbourhood Income* 0.89 Low 310 (49.76%) 153 (50.50%)  High 313 (50.24%) 150 (49.50%)  Parental Support for Physical Activity 2.22 (±1.01) 2.34 (±0.96) 0.077 Perceived Traffic Safety 2.54 (±0.59) 2.66 (±0.56) 0.005 Perceived Pedestrian Safety 2.80 (±0.66) 2.90 (±0.62) 0.038 Perceived crime (low crime rate) 0.22 Strongly disagree 46 (7.38%) 12 (3.99%)  Somewhat disagree 101 (16.21%) 47 (15.61%)  Somewhat agree 246 (39.49%) 122 (40.53%)  Strongly agree 230 (36.92%) 120 (39.87%)  Perceived Stranger Danger 2.05 (±0.74) 1.88 (±0.71) 0.001 *Low income refers to income below $50,907 for the Baltimore-Washington DC four-county study region; $56,676 for Seattle-King County   Neighbourhood walkability was lower among the participants who were active in their neighbourhoods (mean = −0.29, SD= 2.56), whereas the MAPS grand score was higher in the 120  same group (mean=2.79, SD=7.44) (Table 5-7). A high proportion of boys (57.75%) were more active in their neighbourhoods, and a high percentage of girls (58%) were not active in their neighbourhoods; this difference was significant. There was a marginally significant difference in neighbourhood income between participants who were active in their neighbourhoods and those who were not. Parental support for physical activity was also higher in participants who were active in their neighbourhoods (mean = 2.49, SD = 0.91). The parental perceptions of traffic, pedestrian safety, crime in the neighbourhood, and perceived stranger danger were not different in the two groups.  Table 5-7. TEAN: Bivariate relationship between neighbourhood walkability, MAPS scores, and other key variables based on neighbourhood physical activity. Variables Not Active in Neighbourhood (n= 481) Active in Neighbourhood (n= 445) p-value Neighbourhood Walkability 0.19 (±3.07) -0.29 (±2.56) 0.010 MAPS Grand Score 4.69 (±8.62) 2.79 (±7.44) 0.0005 Gender < 0.0001 Male 202 (42.00%) 257 (57.75%)  Female 279 (58.00%) 188 (42.25%)  Neighbourhood Income* 0.087 Low 254 (52.81%) 209 (46.97%)  High 227 (47.19%) 236 (53.03%)  Parental Support for Physical Activity 2.04 (±1.02) 2.49 (±0.91) < 0.0001 Perceived Traffic Safety 2.55 (±0.61) 2.61 (±0.55) 0.12 Perceived Pedestrian Safety 2.84 (±0.67) 2.82 (±0.64) 0.71 Perceived crime (low crime rate) 0.47 Strongly disagree 31 (6.44%) 27 (6.09%)  Somewhat disagree 85 (17.67%) 63 (14.22%)  Somewhat agree 191 (39.71%) 177 (39.95%)  Strongly agree 174 (36.17%) 176 (39.73%)  Perceived Stranger Danger 2.02 (±0.73) 1.96 (±0.73) 0.18 Note: Neighbourhood physical activity refers to being active in various locations around home. *Low income refers to income below $50,907 for the Baltimore-Washington DC four-county study region; $56,676 for Seattle-King County  121   There were significant differences in neighbourhood walkability, gender, neighbourhood income, and parental support between the participants who spent at least 30 minutes daily doing MVPA (measured by an accelerometer) during non-school hours (Table 5-8). The average neighbourhood walkability (mean = 0.44, SD = 3.66) was higher in participants who spent at least 30 minutes daily doing MVPA during non-school hours. A higher proportion of the boys (73.03%) were active, doing MVPA for 30 minutes daily during non-school hours. Likewise, a higher percentage of participants from the high-income neighbourhoods (57.89%) met the 30-minute daily threshold of MVPA during non-school hours. Parental support for physical activity (mean = 2.66, SD = 0.95) was also higher in participants who were doing 30 minutes or more doing MVPA outside of school. Table 5-8. TEAN: Bivariate relationship between neighbourhood walkability, MAPS scores, and other key variables based on non-school hours of daily moderate to vigorous physical activity (MVPA) measured by accelerometer. Variables MVPA < 30 min/day (n = 709) MVPA ≥ 30 min/day (n = 152) p-value Neighbourhood Walkability -0.15 (±2.67) 0.44 (±3.66) 0.022 MAPS Grand Score 3.66 (±8.04) 4.03 (±8.39) 0.61 Gender < 0.0001 Male 320 (45.13%) 111 (73.03%)  Female 389 (54.87%) 41 (26.97%)  Neighbourhood Income* 0.040 Low 366 (51.62%) 64 (42.11%)  High 343 (48.38%) 88 (57.89%)  Parental Support for Physical Activity 2.17 (±0.97) 2.66 (±0.95) < 0.0001 Perceived Traffic Safety 2.57 (±0.58) 2.66 (±0.57) 0.090 Perceived Pedestrian Safety 2.83 (±0.65) 2.80 (±0.64) 0.55 Perceived crime (low crime rate) 0.13 Strongly disagree 40 (5.64%) 13 (8.61%)  122  Variables MVPA < 30 min/day (n = 709) MVPA ≥ 30 min/day (n = 152) p-value Somewhat disagree 120 (16.93%) 16 (10.60%)  Somewhat agree 284 (40.06%) 60 (39.74%)  Strongly agree 265 (37.38%) 62 (41.06%)  Perceived Stranger Danger 2.00 (±0.74) 1.96 (±0.69) 0.52 *Low income refers to income below $50,907 for the Baltimore-Washington DC four-county study region; $56,676 for Seattle-King County.  5.4 Sample 3: Senior Neighborhood Quality of Life Study (SNQLS) 5.4.1 Descriptive Statistics The total sample size for the study was 367 older adults (Table 5-9). The average age of the participants was about 74.98 years. There was gender balance in the sample, which was 50.68% female, but it had an overrepresentation of white older adults (83.84%). Almost half of the sample was from low-income neighbourhoods (48.50%). The mean scores for family support for physical activity, perceived traffic safety, pedestrian safety, and personal safety were 1.68, 2.82, 2.61, and 3.51, respectively. Of the total sample, 40.44% of participants walked for transportation, 70.19% walked for leisure, and 15.83% spent at least 30 minutes in daily MVPA. Table 5-9. SNQLS: Descriptive summary of demography, psychosocial factors, outcome, and other covariates.  Variables Mean (SD)/ Count (Percentage) (N = 367) Age (Range: 66 – 79) 74.98 (±6.62) Ethnicity Non-White 59 (16.16%) White 306 (83.84%) Gender Male 181 (49.32%) Female 186 (50.68%) Neighbourhood Income* Low 178 (48.50%) High 189 (51.50%) Average Temperature (Celsius) (Range: -4.19 – 18.22) 7.60 (±6.55) 123  Variables Mean (SD)/ Count (Percentage) (N = 367) Average Precipitation (Millimeters) (Range: 0 – 48.75) 6.27 (±7.36) Regional Transit Accessibility (Range: 0 – 7) 1.20 (±1.95) Self-efficacy for walking (Range: 1 – 10) 8.60 (±2.44) Mobility Status (Range: 0 – 100) 57.96 (±18.04) Family Support for Physical Activity (Range: 0 – 4) 1.68 (±1.34) Perceived Traffic Safety (Range: 0 – 4) 2.82 (±0.66) Perceived Pedestrian Safety (Range: 0 – 4) 2.61 (±0.46) Perceived Personal Safety (Range: 0 – 4) 3.51 (±0.51) Active Transport (Walking or Biking) No 215 (59.56%) Yes 146 (40.44%) Leisure Walking No 107 (29.81%) Yes 252 (70.19%) Daily Moderate to Vigorous Physical Activity (≥ 30 minutes) No 303 (84.17%) Yes 57 (15.83%) Note: Number of missing variables: Ethnicity = 2; Average Temperature and Average Precipitation = 7; Family Support for Physical Activity = 6; Perceived Traffic Safety = 7, Perceived Pedestrian Safety = 7; Perceived Personal Safety = 7, Active Transport = 6; Leisure Walking = 8, Non-School Hours of Daily MVPA = 7. *Low income refers to income below $56,676.   5.4.2 Bivariate Relationship The neighbourhood walkability (mean = 0.82, SD = 2.50) and MAPS grand score (mean = 14.80, SD = 8.25) were higher among participants who walked for transportation compared to those who did not walk (Table 5-10). There were no significant differences in gender and neighbourhood income between the two groups. Overall, the scores for self-efficacy for walking (mean = 9.08, SD = 1.91), mobility status (mean = 61.69, SD = 18.33), family support for physical activity (mean = 1.85, SD = 1.37) and pedestrian safety (mean = 2.71, SD = 0.45) were higher among the participants who walked for transportation and the differences were significant.  124  Table 5-10. SNQLS: Bivariate relationship between neighbourhood walkability, MAPS scores, and other key variables based on active transport.  Do not Walk or Bike for Transport (n= 215) Walk or Bike for Transport (n= 146) p-value Neighbourhood Walkability -1.21 (±2.25) 0.82 (±2.50) < 0.0001 MAPS Grand Score 9.21 (±6.11) 14.80 (±8.25) < 0.0001 Gender 0.33 Male 111 (51.63%) 67 (45.89%)  Female 104 (48.37%) 79 (54.11%)  Neighbourhood Income* 0.83 Low 103 (47.91%) 72 (49.32%)  High 112 (52.09%) 74 (50.68%)  Self-efficacy for walking 8.27 (±2.70) 9.08 (±1.91) 0.002 Mobility Status 55.43 (±17.43) 61.69 (±18.33) 0.001 Family Support for Physical Activity 1.56 (±1.31) 1.85 (±1.37) 0.044 Perceived Traffic Safety 2.82 (±0.64) 2.82 (±0.69) 0.97 Perceived Pedestrian Safety 2.54 (±0.46) 2.71 (±0.45) 0.0006 Perceived Personal Safety 3.53 (±0.50) 3.47 (±0.51) 0.24 *Low income refers to income below $56,676.   For leisure walking (Table 5-11), there was a significant difference in neighbourhood walkability, self-efficacy for walking, mobility status, family support for physical activity, and pedestrian safety between the two groups (those who walked for leisure vs. those who didn’t). Participants who walked for leisure or exercise had high scores for neighbourhood walkability (mean = -0.21, SD = 2.64), self-efficacy for walking (mean = 8.87, SD = 2.16), mobility status ( mean = 59.73, SD = 17.29), family support (mean = 2.066, SD = 1.377), and pedestrian safety (mean = 2.65, SD = 0.48) compared to those who did not walk for leisure.     125  Table 5-11. SNQLS: Bivariate relationship between neighbourhood walkability, MAPS scores, and other key variables based on leisure walking.  Do not walk for leisure  (n = 107) Walk for leisure (n=252) p-value Neighbourhood Walkability -0.86 (±2.30) -0.21 (±2.64) 0.029 MAPS Grand Score 10.51 (±7.47) 11.88 (±7.60) 0.12 Gender 1 Male 53 (49.53%) 124 (49.21%)  Female 54 (50.47%) 128 (50.79%)  Neighbourhood Income* 1 Low 52 (48.60%) 121 (48.02%)  High 55 (51.40%) 131 (51.98%)  Self-efficacy for walking 7.98 (±2.91) 8.87 (±2.16) 0.002 Mobility Status 53.68 (±19.25) 59.72 (±17.29) 0.004 Family Support for Physical Activity 1.11 (±1.18) 1.92 (±1.34) < 0.0001 Perceived Traffic Safety 2.82 (±0.62) 2.83 (±0.67) 0.97 Perceived Pedestrian Safety 2.52 (±0.40) 2.65 (±0.48) 0.011 Perceived Personal Safety 3.58 (±0.41) 3.48 (±0.54) 0.065 *Low income refers to income below $56,676.  The group difference test showed marginal gender differences and significant neighbourhood income differences between the two groups based on daily MVPA (Table 5-12) measured by an accelerometer. A higher proportion of male participants (61.40%) spent 30 minutes doing MVPA daily compared to female participants (38.60%), and a large proportion of the participants living high-income neighbourhoods (71.93%) spent 30 minutes doing MVPA daily compared to those living in low-income neighbourhoods (28.07%). Like scores for transport walking and leisure walking, the scores for self-efficacy for walking (mean = 9.87, SD = 0.42), mobility status (mean = 71.67, SD = 11.71) and family support for physical activity (mean = 2.11, SD = 1.42) were higher among the participants who spent at least 30 minutes daily doing MVPA. The scores for perceived traffic safety (mean = 2.99, SD = 0.61), perceived 126  pedestrian safety (mean = 2.74, SD = 0.42), and personal safety (mean = 3.65, SD = 0.40) were also higher among the participants who met the 30-minute daily MVPA threshold. Table 5-12. SNQLS: Bivariate relationship between neighbourhood walkability, MAPS scores, and other key variables based on daily moderate to vigorous physical activity (MVPA) measured by accelerometer.  Daily MVPA < 30 mins/day (n= 303) Daily MVPA ≥ 30 mins/day (n= 57) p-value Neighbourhood Walkability -0.35 (±2.61) -0.57 (±2.11) 0.55 MAPS Grand Score 11.31 (±7.60) 11.89 (±6.48) 0.59 Gender 0.061 Male 144 (47.52%) 35 (61.40%)  Female 159 (52.48%) 22 (38.60%)  Neighbourhood Income* 0.0008 Low 159 (52.48%) 16 (28.07%)  High 144 (47.52%) 41 (71.93%)  Self-efficacy for walking 8.41 (±2.51) 9.87 (±0.42) < 0.0001 Mobility Status 55.50 (±17.61) 71.67 (±11.71) < 0.0001 Family Support for Physical Activity 1.61 (±1.31) 2.11 (±1.42) 0.010 Perceived Traffic Safety 2.80 (±0.66) 2.99 (±0.61) 0.043 Perceived Pedestrian Safety 2.59 (±0.47) 2.74 (±0.42) 0.031 Perceived Personal Safety 3.49 (±0.52) 3.65 (±0.40) 0.022 *Low income refers to income below $56,676.  5.5 Summary This chapter presented the results of exploratory analysis of the three datasets by tabulating descriptive statistics and conducting tests of bivariate relation between outcome and key independent variables. A total of 757 children participated in the NIK study; 928 teens in the TEAN study; and 367 in the SNQLS study. There was a balance in gender proportion in all three studies. All three studies had a higher percentage of the participants being white. About 83% of children, 37.72% of teens, 40.43% of older adults walked or biked for transportation.  More than half of the children (about 55%) and almost half of teens (48.05%) were active in the neighbourhood, such as playing in neighbourhood streets, on sidewalks, etc. Less than a quarter 127  (22.04%) of the total number of children spent at least 60 minutes doing moderate to vigorous physical activity daily during non-school hours. However, the percentage was much lower in teens and older adults. Only 17.65% of the total number of children spent at least 30 minutes doing moderate to vigorous physical activity daily during non-school hours; 15.83% of the older adults spent 30 minutes doing moderate to vigorous physical activity daily. Bivariate tests showed differences in the built environment and psychosocial factors in all age groups for transport walking, neighbourhood physical activity and leisure walking. A high percentage of females compared to males reported low physical activity. Psychosocial factors were also high among the physically active participants. The following chapter provides the results of multivariate models that offer a better picture of the collective effect of neighbourhood walkability and pedestrian environment on physical activity.         128  CHAPTER 6: RESULTS: INFERENTIAL STATISTICS This chapter presents the results of the statistical models fitted to test the hypotheses outlined in the first chapter. Results are grouped according to the study sample, starting with children (NIK study; Section 6.1), teens (TEAN study; Section 6.2) and older adults (SNQLS study; Section 6.3). For each sample, results are presented based on the outcome or the behaviour of interest.  Figure 6-1 shows a schematic diagram showing the organization of results in three sections for each dataset. As shown in the figure, the results from each sample are organized based on the three outcomes.  Table 6-1 shows the organization of the results for MAPS grand scores and MAPS domain-specific scores for each outcome. The organization of the results of supplementary analyses using the MAPS valence scores and sub scales is shown in Table 6-2.   Figure 6-1. Schematic diagram showing the organization of results. Result for each dataset is presented separately.  129  Table 6-1. Sample table showing the organization of table for reporting the MAPS grand and the domain specific MAPS score results.   Fixed Effects of Key Variables Base Model Objective 1: Objective 2 & Objective 3:   2-Way Interaction  Final Model This column includes key variables used in the study analysis. This column reports results of model with walkability index, MAPS score and other covariates.     This column reports the result with 2-way interaction between walkability index and MAPS Grand/MAPS domain specific score.    This column reports the result with significant 3- way interactions between walkability index, MAPS Grand/MAPS domain specific score and demographic and psychosocial factors.  The psychosocial factors that do not have significant interaction are not retained as covariates.  Random Effects & Model Information (This section shows the estimates of random effects and other information relation to the models)  Table 6-2. Sample table showing the organization of table for reporting the MAPS valence score and sub scale results (supplementary analysis).  Fixed Effects of Key Variables  ‡MAPS Valence score & MAPS Sub Scale  MAPS Score† This column reports the results for MAPS sub-score or MAPS valence scores which have significant interaction with walkability index and Demographic or Psychosocial factor.   The demographic/psychosocial factors are identified from the final model identified in the MAPS grand score analysis.   The number of this column depends on the number of significant interactions identified.    Walkability index  Parent Support   MAPS score† × Walkability index   MAPS score† × Demographic or Psychosocial factor  Walkability index × Demographic or Psychosocial factor  MAPS Score† × Walkability index ×   Demographic or Psychosocial factor Random Effects & Model Information (This section shows the estimates of random effects and other information relation to the models) Note: † MAPS Score refer to the relevant MAPS valence score and MAPS Sub scale that show significant interaction, i.e., the ‡ MAPS label in each column).  130  6.1 Sample 1: Neighbourhood Influences on Kids (NIK)  6.1.1 Transport Walking or Biking  There were no significant interactions for transport walking or biking in children.  The interactions with MAPS valence scores and subscales were not tested as no significant interactions were observed using the MAPS grand score.    6.1.2 Neighbourhood Physical Activity  None of the interactions were significant in the case of neighbourhood physical activity in children.  6.1.3 Objective Physical Activity 6.1.3.1 MAPS Grand Score  There was no significant two-way interaction between MAPS grand score and neighbourhood walkability (objective 1) (Table 6-3). The three-way interactions related to objective 2: i) walkability index × MAPS grand score × Gender, and ii) walkability index × MAPS grand score × Household Income, were not significant. Out of the five psychosocial factors (traffic safety, pedestrian safety, perceived crime, stranger danger and parental support for physical activity) examined for the third objective, only parental support for physical activity had a significant interaction with the walkability index and MAPS grand score (Log Odds = 0.03; SE = 0.01, p-value = 0.02) (the final model in Table 6-3). The main effects of walkability index (Log Odds = -0.34; SE = 0.09, p-value = 0.00) and MAPS grand score (Log Odds = 0.18; SE = 0.04, p-value = 0.00) were also significant. However, it should also be noted that parallel models that used different approximation techniques yielded non-significant interactions (Appendix D  ).  131  Table 6-3. NIK: Results of the regression model for objective MVPA ≥ 60 min/day during non-school hour using MAPS grand score.  Fixed Effects of Key Variables  Base Model Objective 1:  2-Way Interaction Objective 2 & Objective 3:  Final Model Log Odds p-value†   Log Odds p-value†   Log Odds p-value†   (SE) (SE) (SE) MAPS Grand 0.17 0.000  0.16 0.000  0.18 0.000  (0.03) (0.03) (0.04) Walkability Index  -0.327 0.000 −0.33 0.000  −0.340 0.001  (0.08) (0.08) (0.09) MAPS Grand × Walkability Index     0.01 0.788 0.00 0.904  (0.01) (0.01) Parent Support         0.24 0.340  (0.20) MAPS Grand × Parent Support         −0.05 0.336  (0.04) Walkability Index × Parent Support         0.12 0.193  (0.08) MAPS Grand × Walkability Index ×         0.03 0.022 Parent Support (0.01) Random Effects & Model Information       Intercept (SD) 3.42 3.43 3.46 Residual (SD) 0.40 0.40 0.39 AUC 99% 99% 99% Num. obs. 679 679 678 Num. groups 571 571 571 Note: All models adjusted for age, income, gender, ethnicity, neighbourhood preference, temperature, precipitation, and regional accessibility. †- p-values were adjusted for False Discovery Rates and rounded to 3 decimal places.  Acronyms: MVPA: Moderate to Vigorous Physical Activity; SE: Standard Error; SD: Standard Deviation; AUC: Area Under the Curve.  Refer Appendix B  for full model results; Appendix C  for VIF for final model. 132    Figure 6-2 shows a graphical representation of the three-way interaction between neighbourhood walkability, MAPS grand score, and parental support for physical activity. The figure shows a synergistic interaction between neighbourhood walkability and MAPS grand score for children from families with higher support for physical activity, and a buffering interaction between neighbourhood walkability and MAPS grand score for children from Figure 6-2. Walkability index, MAPS grand score, and parental support for physical activity explaining the likelihood of spending at least 60 minutes doing MVPA daily during non-school hours. High and Low refer to values 1 standard deviation above and below the mean.    133  families with lower support for physical activity. A similar trend is seen in the predicted probability graph shown in Figure 6-3.    6.1.3.2 Domain-Specific MAPS Scores Domain-specific interactions were also tested using the MAPS score for active transport (MAPS Active Transport) and MAPS score for leisure (MAPS Leisure). MAP active transport did not have any significant interaction. However, the MAPS leisure score showed a similar pattern seen in MAPS grand score (Table 6-3). The final model showed a significant Walkability index × MAPS leisure score × Parental support interaction (Log Odds = 0.05; SE = 0.02, p-value = 0.037) (Table 6-4).  Figure 6-4 shows a graphical representation of the interaction using high and low walkability index values and parental support. Figure 6-5 shows graphs with predicted probability of spending at least 60 minutes/day during non-school hours at the different levels of walkability index and MAPS leisure scores.  Figure 6-3. Predicted probabilities of objective physical activity at different levels of walkability and MAPS grand scores for high and low parental support. High and Low refer to values 1 standard deviation above and below the mean.   134  Table 6-4: NIK: Results of the regression model for objective MVPA ≥ 60 min/day during non-school hour using MAPS leisure score.  Fixed Effects of Key Variables  Base Model Objective 1:  2-Way Interaction Objective 2 & Objective 3:  Final Model Log Odds p-value†   Log Odds p-value†   Log Odds p-value†   (SE) (SE) (SE) MAPS Leisure 0.15 0.010 0.15 0.011 0.18 0.006    (0.05) (0.05) (0.06) Walkability Index -0.19 0.018 -0.19 0.019 -0.20 0.034    (0.07) (0.07) (0.08) MAPS Grand Leisure × Walkability Index     0.01 0.829 -0.11 0.750        (0.02) (0.02) Parent Support         0.42 0.038            (0.18) MAPS Grand Leisure × Parent Support         -0.10 0.179            (0.06) Walkability Index × Parent Support         0.11 0.159            (0.07) MAPS Grand Leisure × Walkability Index ×          0.05 0.037 Parent Support         (0.02) Random Effects & Model Information    Intercept (SD) 3.16 3.17 3.28 Residual (SD) 0.43 0.42 0.41 AUC 99.99% 99.99% 99.89% Num. obs. 679 679 678 Num. groups 571 571 570 Note: All models adjusted for age, income, gender, ethnicity, neighbourhood preference, temperature, precipitation, and regional accessibility. †- p-values were adjusted for False Discovery Rates and rounded to 3 decimal places.  Acronym: MVPA: Moderate to Vigorous Physical Activity; SE: Standard Error; SD: Standard Deviation; AUC: Area Under the Curve. Refer Appendix B  for full model results; Appendix C  for VIF for final model. 135   Figure 6-4. Walkability index, MAPS leisure score, and parental support for physical activity explaining the likelihood of spending at least 60 minutes doing MVPA daily during non-school hours. High and Low refer to values 1 standard deviation above and below the mean.   Figure 6-5. Predicted probabilities of objective physical activity at different levels of walkability and MAPS leisure scores for high and low parental support. High and Low refer to values 1 standard deviation above and below the mean.   136  6.1.3.3 Supplementary Analysis Using Valence Scores and Subscales There were four significant interactions (p-value < 0.01) between neighbourhood walkability, MAPS grand score, and MAPS valence/subscales (Table 6-5) after adjusting for the false discovery rate. They include the interactions with the street characteristics related to aesthetics and social environment, overall route score, and overall positive scores on street design.  Figure 6-6 shows the graphical representation of the interactions.   Table 6-5. Results of the three-way interaction between walkability, parental support, and MAPS valence and subscales.      MAPS Scores (Valence scores and Subscales) ‡ Fixed Effects of Key Variables   Overall Positive  Route Overall Aesthetic/ Social Overall Positive Aesthetic/Social  Log Odds Log Odds Log Odds Log Odds  (SE) (SE) (SE) (SE) MAPS Score† 0.14* 0.34* 1.05* 1.53*   (0.04) (0.05) (0.12) (0.20) Walkability index -0.32* -0.45* -0.15 -0.13   (0.09) (0.09) (0.09) (0.09) Parent Support 0.13 0.05 0.47 0.46   (0.20) (0.20) (0.20) (0.19) MAPS score† × Walkability index 0.01 0.00 -0.07· -0.08   (0.01) (0.02) (0.04) (0.07) MAPS score† × Parent Support -0.02 -0.19* -0.78* -1.14*   (0.05) (0.05) (0.11) (0.21) Walkability index × Parent Support 0.06 0.13· 0.00 0.03   (0.08) (0.08) (0.07) (0.07) MAPS Score† × Walkability index ×  0.05* 0.10* 0.19* 0.45*  Parent Support (0.01) (0.02) (0.04) (0.07) Note: † MAPS Score refer to ‡ MAPS label in each column.  * p-value < 0.01. All p-values were adjusted for false discovery rate.  All models adjusted for age, income, gender, ethnicity, neighbourhood preference, temperature, precipitation, and regional accessibility. Refer Appendix B  for full model results; Appendix C  for VIF for final model. 137   Overall Positive Score  Route Overall Score Aesthetic & Social Characteristics Overall Score Positive Aesthetic and Social Score Figure 6-6. Graphical representation of the interactions between neighborhood walkability, parental support and MAPS valence and sub scores. High and Low refer to values 1 standard deviation above and below the mean.   138  6.2 Sample 2: Teen Environment and Neighborhood (TEAN)  6.2.1 Transport Walking or Biking No significant interactions between neighbourhood walkability and MAPS grand score or any of the demographic and psychosocial factors were found.  6.2.2 Neighbourhood Physical Activity  There were no significant interactions between neighbourhood walkability and MAPS grand score or any demographic and psychosocial factors.  6.2.3 Objective Physical Activity  6.2.3.1 MAPS Grand Score Results  There were no significant interactions for MAPS grand score.  6.2.3.2 Domain-Specific MAPS Scores There was a significant two-way interaction between neighbourhood walkability and MAPS grand leisure score. The results are summarized in Table 6-6. There were no significant main effects of neighbourhood walkability and MAPS leisure scores; however, there was a significant Walkability index × MAPS grand leisure score interaction (Log Odds = 0.03, SE = 0.01, p = 0.015). The interaction plot (Figure 6-7) and the predicted probability graph (Figure 6-8) both show a disordinal interaction between neighbourhood walkability and MAPS leisure score, which clarifies the non-significant main effect in the final model.       139  Table 6-6. TEAN: Results of the regression model for objective MVPA ≥ 30 min/day during non-school hour using MAPS leisure score. Fixed Effects Base Model Objective 1:  2-Way Interaction Log Odds p-value† Log Odds p-value† (SE)   (SE)   Walkability Index 0.07 0.154 0.06 0.266   (0.04) (0.04) MAPS Leisure 0.01 0.815 0.02 0.571   (0.03)   (0.04) Walkability Index ×     0.03 0.015  MAPS Leisure     (0.01) Random Effects & Model Information     Intercept (SD) 0.00 0.00 Residual (SD) 1.02 1.02 AUC 75% 76% Num. obs. 855 855 Num. groups 429 429 Note: All models adjusted for age, income, gender, ethnicity, neighbourhood preference, temperature, precipitation, and regional accessibility. †- p-values were adjusted for False Discovery Rates and rounded to 3 decimal places.  Acronym: MVPA: Moderate to Vigorous Physical Activity; SE: Standard Error; SE: Standard Deviation; AUC: Area Under the Curve. Refer Appendix B  for full model results; Appendix C  for VIF for final model.  140   Figure 6-7. Neighbourhood walkability and MAPS grand leisure score interaction explaining objective physical activity (MVPA≥  30mins/day during non-school hour). High and Low refer to values 1 standard deviation above and below the mean.   Figure 6-8. Predicted probabilities for objective physical activity at different levels of walkability and MAPS grand scores.  141  6.2.3.3 Supplementary Analysis Using Valence Scores and Subscales  None of the MAPS valence scores or subscales had a significant interaction with neighbourhood walkability.  6.3 Sample 3: Senior Quality of Life Study (SNQLS)  6.3.1 Transport Walking or Biking 6.3.1.1 MAPS Grand Score  Table 6-7 summarizes the results of the analysis. The main effects model shows a positive relationship between transport walking and neighbourhood walkability and MAPS grand scores. There was a significant negative Walkability x MAPS Grand score interaction (Log Odds = -0.02; SE = 0.01, p-value= 0.009). The final model shows a significant three-way interaction between Walkability, MAPS Grand score, and Gender (Log Odds = 0.06; SE = 0.02, p-value = 0.003). Neighbourhood income, safety perception (traffic safety, crime, personal safety), self-efficacy for physical activity and family support for physical activity did not have a significant three-way interaction with the Walkability index and MAPS grand score. Except for neighbourhood income, these variables were not retained in the final model.  Figure 6-10 shows the significant three-way interaction between Walkability, MAPS Grand score, and Gender. The figure shows that for both males and females, the effect of neighbourhood walkability on active transportation increases with an increase in MAPS scores. However, for males, the effect is stronger when neighbourhood walkability is low. This shows the traits of buffering interaction. For females, the overall impact is positive for both high- and low-walkability neighbourhoods, but it is stronger when neighbourhood walkability is high, which suggests the presence of synergistic interaction. Figure 6-9 shows the predicted probability of walking or biking as transportation for males and females at different levels of 142  neighbourhood walkability and MAPS grand scores. Results show a similar trend, as seen in Figure 6-9. The probability of walking or biking for transportation among males is the lowest when both walkability and MAPS scores are low (probability = 1.55%). As the MAPS score increases, the rate of increase in the likelihood of walking or biking for transportation increases sharply in a less walkable neighbourhood (from 1.55% to 36.79%) compared to a highly walkable neighbourhood (from 45.70% to 56.31%). However, in females, with the increase in MAPS scores, the change in probability is higher in a highly walkable neighbourhood than a less walkable neighbourhood. 143  Table 6-7. SNQLS: Results of the regression model for transport walking or biking using MAPS grand score.  Fixed Effects of Key Variables  Base Model Objective 1:  2-Way Interaction Objective 2 & Objective 3:  Final Model Log Odds p-value† Log Odds p-value† Log Odds p-value† (SE)  (SE)  (SE)  Gender (Female) 0.51 0.179 0.59 0.034  0.36 0.257   (0.33) (0.26) (0.28) Walkability Index 0.35 0.002 0.38 0.000  0.47 0.000   (0.09) (0.09) (0.12) MAPS Grand 0.08 0.003 0.10 0.000  0.13 0.000   (0.02) (0.02) (0.03) Walkability Index × MAPS Grand     -0.02 0.009 -0.04 0.000       (0.01) (0.01) Gender (Female) × Walkability Index         -0.05 0.733           (0.12) Gender (Female) × MAPS Grand         -0.08 0.094           (0.04) Gender (Female) × Walkability Index ×          0.06 0.003  MAPS Grand         (0.02) Random Effects & Model Information       Intercept (SD) 0.00 1.09 1.32 Residual (SD) 1.19 0.84 0.79 AUC 81% 91% 93% Num. obs. 359 359 359 Num. groups 11 116 116 Note: All models control for age, neighbourhood income, ethnicity, mobility status, and regional accessibility.  †- p-values were adjusted for False Discovery Rates and rounded to 3 decimal places.  Acronym: SE: Standard Error; SD: Standard Deviation; AUC: Area Under the Curve. Refer Appendix B  for full model results; Appendix C  for VIF for final model. 144   Figure 6-9. Predicted probabilities for transport walking or biking at different levels of walkability and MAPS grand scores for males and females.   Figure 6-10. Neighbourhood walkability, MAPS grand score, and gender interactions explaining transport walking or biking. High and Low refer to values 1 standard deviation above and below the mean.   145   6.3.1.2 Domain-Specific MAPS Score The results using MAPS score for active transport are shown in Table 6-8. Results showed a similar pattern seen in the MAPS active transport (MAPS AT) score results (Table 6-7). There was a significant Walkability × MAPS AT score interaction (objective 1). There was positive significant three-way interaction between neighbourhood walkability, MAPS active transport score and gender. Neighbourhood income did not have significant interaction with the walkability index and MAPS AT score. There were no significant interactions between any of the psychosocial factors, neighbourhood walkability and MAPS AT score. The graphical representation of the interaction using high and low values of walkability index for males and females is shown in Figure 6-11; Figure 6-12 shows graphs with predicted probabilities of walking or biking at different levels of walkability index and MAPS active transport score.            146  Table 6-8. SNQLS: Results of the regression model for transport walking or biking using MAPS active transport (MAPS AT) score.  Fixed Effects of Key Variables  Base Model Objective 1:  2-Way Interaction Objective 2 & Objective 3:  Final Model Log Odds p-value† Log Odds p-value† Log Odds p-value† (SE) (SE) (SE) Gender (Female) 0.47 0.244 0.54 0.059 0.27 0.456   (0.34) (0.26) (0.29) Walkability Index 0.31 0.005 0.37 0.000 0.44 0.001   (0.10) (0.09) (0.19) MAPS AT 0.08 0.003 0.09 0.000 0.12 0.000 (0.02) (0.02) (0.03) Walkability Index × MAPS AT     -0.02 0.005 -0.03 0.001     (0.01) (0.01) Gender (Female) × Walkability Index         -0.05 0.691       (0.14) Gender (Female) × MAPS AT         -0.05 0.222           (0.04) Gender (Female) × Walkability Index ×          0.03 0.030  MAPS AT         (0.01) Random Effects    Intercept (SD) 0.00 0.1 1.15 Residual (SD) 1.23 0.85 0.82 AUC 81.64% 90.7% 92.41% Num. obs. 359 359 359 Num. groups 116 116 116 Note: All models control for age, neighbourhood income, ethnicity, mobility status, and regional accessibility.  †- p-values were adjusted for False Discovery Rates and rounded to 3 decimal places.  Acronym: SE: Standard Error; SD: Standard Deviation; AUC: Area Under the Curve. Refer Appendix B  for full model results; Appendix C  for VIF for final model. 147     Figure 6-11. Neighbourhood walkability, MAPS active transport score, and gender interactions explaining transport walking or biking. High and Low refer to values 1 standard deviation above and below the mean.   Figure 6-12. Predicted probabilities for transport walking or biking at different levels of walkability and MAPS active transport scores for male and female. 148  6.3.1.3 Supplementary Analysis Using Valence Scores and Subscales   Table 6-9 summarizes the results of the interaction between neighbourhood walkability, gender, and MAPS subscales, which were significant at the 0.01 alpha level. There were six significant interactions. The interaction patterns showed a similar trend as the MAPS grand score interaction. Positive MAPS subscales had significant positive interaction, whereas negative MAPS subscales showed negative interaction (Figure 6-13 and Figure 6-14).   Table 6-9. SNQLS: Results of the three-way interaction between walkability, gender, and MAPS valence and subscales for transport walking or biking.  MAPS Scores (Valence scores and Subscales) ‡ Fixed Effects of Key Variables   Positive Streetscape Negative Streetscape Streetscape Overall Route Overall Crossing Impediment Crossing  Overall  Log Odds Log Odds Log Odds Log Odds Log Odds Log Odds  (SE) (SE) (SE) (SE) (SE) (SE) Gender (Female) 0.51 0.54 0.51 0.39 0.59 0.48   (0.27) (0.27) (0.27) (0.27) (0.32) (0.31) Walkability Index 0.47* 0.52* 0.51* 0.47* 0.61* 0.49*   (0.11) (0.12) (0.12) (0.11) (0.15) (0.14) MAPS Score† 0.22 -1.07* 0.25* 0.14* -1.02* 0.28*   (0.09) (0.29) (0.08) (0.04) (0.30) (0.09) Walkability Index ×  -0.12 0.29* -0.10* -0.05* 0.63* -0.11*  MAPS Score† (0.04) (0.10) (0.03) (0.01) (0.16) (0.04) Gender (Female) ×  0.02 -0.04 -0.01 -0.04 -0.05 0.05  Walkability Index (0.12) (0.13) (0.12) (0.12) (0.15) (0.14) Gender (Female) ×  -0.27 0.94 -0.27 -0.09· 0.40 -0.15  MAPS Score† (0.12) (0.37) (0.10) (0.05) (0.36) (0.11) Gender (Female) × Walkability Index ×  0.19* -0.49* 0.17* 0.06* -0.80* 0.15*  MAPS Score† (0.06) (0.14) (0.05) (0.02) (0.20) (0.05) Note: † MAPS Score refer to ‡ MAPS label in each column.  * p-value < 0.01. All p-values were adjusted for false discovery rate.  All models adjusted for age, income, ethnicity, mobility status, and regional accessibility. Refer Appendix B  for full model results; Appendix C  for VIF for final model.      149   Positive Streetscape Negative Streetscape Overall Streetscape Route Overall Figure 6-13. Graphical representation of the interactions between neighborhood walkability, gender and MAPS valence and sub scores for route. High and Low refer to values 1 standard deviation above and below the mean.   150    6.3.2 Leisure Walking  6.3.2.1 MAPS Grand Score  There was no significant two-way interaction between neighbourhood walkability and MAPS grand score. Likewise, none of the demographic and psychosocial factors had a significant three-way interaction with neighbourhood walkability and MAPS grand score.  6.3.2.2 Domain-Specific MAPS Scores MAPS leisure score also did not show any significant interaction.  6.3.2.3 Supplementary Analysis Using Valence Scores and Subscales  Since the MAPS grand score did not show significant interaction, no further analysis with and MAPS valence scores and subscales was conducted.  Crossing Impediments Crossing Overall Figure 6-14. Graphical representation of the interactions between neighborhood walkability, gender and MAPS valence and sub scores for crossing. High and Low refer to values 1 standard deviation above and below the mean.   151  6.3.3 Objective Physical Activity  6.3.3.1 MAPS Grand Score  The results of the objective physical activity (MVPA) models are summarized in Table 6-10. The base model showed that MVPA is significantly associated with MAPS grand score (Log Odds = 0.06, SE = 0.023, adjusted p = 0.037) but not with neighbourhood walkability (Log Odds = -0.05, SE = 0.11, adjusted p = 0.928) or income (Log Odds = 0.21, SE = 0.51, adjusted p = 0.928). A similar trend was seen in the model, which examined the interaction between neighbourhood walkability and MAPS grand score.   The final model showed a significant three-way interaction between neighbourhood walkability, MAPS grand score, and neighbourhood income (Log Odds = 0.18, SE = 0.05, adjusted p = 0.001). There were significant two-way interactions between neighbourhood walkability and MAPS grand score (Log Odds = −0.15, SE = 0.04, adjusted p = 0.002), and neighbourhood income and MAPS grand score (Log Odds = 0.18, SE = 0.07, adjusted p = 0.033). However, the main effects of income, walkability, and MAPS grand score were not significant in the final model. Gender did not show a significant three-way interaction with walkability index and MAPS Grand score. Likewise, none of the psychosocial factors had a significant three-way interaction with walkability and MAPS grand score.    152  Table 6-10. SNQLS: Results of the regression model for objective MVPA ≥ 30 min/day MAPS grand score. Fixed Effects of Key Variables  Base Model Objective 1:  2-Way Interaction Objective 2 & Objective 3:  Final Model   Log Odds p-value† Log Odds p-value† Log Odds p-value†   (SE)   (SE)   (SE)   Income (High) 0.21 0.928 0.15 0.835 −0.66 0.531 (0.51) (0.51) (0.66) Walkability Index −0.05 0.928 −0.07 0.802 −0.06 0.900   (0.11) (0.12) (0.28) MAPS Grand 0.06 0.037 0.07 0.019 −0.01 0.900   (0.03) (0.03) (0.06) Walkability Index × MAPS Grand     −0.02 0.167 −0.15 0.002     (0.01) (0.04) Income (High) × Walkability Index         0.04 0.900           (0.31) Income (High) × MAPS Grand         0.18 0.033           (0.07) Income (High) × Walkability Index ×         0.18 0.001  MAPS Grand         (0.05) Random Effects & Model Information  Intercept (SD) 1.72 1.73 2.23 Residual (SD) 0.67 0.66 0.61 AUC 96% 96% 98% Num. obs. 352 352 352 Num. groups 116 116 116 Note: All models control for age, gender, ethnicity, mobility status, regional accessibility, temperature, and precipitation.  †- p-values were adjusted for False Discovery Rates and rounded to 3 decimal places.  Acronym: SE: Standard Error; SD: Standard Deviation; AUC: Area Under the Curve. Refer Appendix B  for full model results; Appendix C  for VIF for final model.  The interaction between walkability, MAPS Grand score, and neighbourhood income is graphically shown in Figure 6-15. The figure shows a disordinal interaction between neighbourhood walkability and MAPS grand score in low-income neighbourhoods and synergistic interaction between the two variables in high-income neighbourhoods. A similar 153  trend is seen in Figure 6-16, which shows the predicted probability of spending at least 30 minutes daily doing MVPA at different neighbourhood income levels, walkability, and MAPS grand scores. The likelihood of spending at least 30 minutes a day doing MVPA in low-income neighbourhoods is high when walkability is low and MAPS grand score is high, and when walkability is high, and MAPS grand score is low. Likewise, in high-income neighbourhoods, the probability of spending 30 minutes or more each day doing MVPA is high when both neighbourhood walkability and MAPS grand scores are high. However, the effects are very low at the selected levels of neighbourhood walkability and MAPS grand scores.   Figure 6-15. Interaction between neighbourhood walkability, MAPS grand score, and income for MVPA. High and Low walkability refer to values 1 standard deviation above and below the mean walkbility index.   154     6.3.3.2 Domain-Specific MAPS Scores MAPS leisure score and MAPS active transport scores were used to test hypotheses. However, only MAPS active transport score showed a significant Income (High) × Walkability Index × MAPS active transport score (Table 6-11). The interaction was similar to the MAPS grand score interaction seen in Table 6-10. Graphical representation of the interaction is shown in Figure 6-17 and Figure 6-18.     Figure 6-16. Predicted probabilities for MVPA ≥ 30min/day at different levels of walkability and MAPS grand scores for high- and low-income neigbourhoods  155  Table 6-11. SNQLS: Results of the regression model for objective MVPA ≥ 30 min/day MAPS active transport score.  Fixed Effects of Key Variables  Base Model Objective 1:  2-Way Interaction Objective 2 & Objective 3:  Final Model Log Odds p-value Log Odds p-value Log Odds p-value (SE)   (SE)   (SE)   Income (High) 0.24 0.866 0.20 0.890 -0.40 0.772   (0.49) (0.50) (0.61) Walkability Index -0.06 0.866 -0.07 0.817 -0.02 0.931   (0.11) (0.12) (0.26) MAPS AT 0.05 0.136 0.05 0.092 -0.04 0.636   (0.02) (0.03) (0.05) Walkability Index × MAPS AT     -0.02 0.225 -0.08 0.027       (0.01) (0.03) Income (High) × Walkability Index         -0.06 0.903           (0.29) Income (High) × MAPS AT         0.16 0.018           (0.06) Income (High) × Walkability Index ×         0.09 0.027  MAPS AT         (0.04) Random Effects & Model Information    Intercept (SD) 1.63 1.66 1.93 Residual (SD) 0.69 0.67 0.65 AUC 96.18% 96.29% 97.18% Num. obs. 352 352 352 Num. groups 116 116 116 Note: All models control for age, gender, ethnicity, mobility status, regional accessibility, temperature, and precipitation.  †- p-values were adjusted for False Discovery Rates and rounded to 3 decimal places.  Acronym: SE: Standard Error; SD: Standard Deviation; AUC: Area Under the Curve. Refer Appendix B  for full model results; Appendix C  for VIF for final model. 156    Figure 6-17. Interaction between neighbourhood walkability, MAPS active transport score, and income for MVPA. High and Low walkability refer to values 1 standard deviation above and below the mean walkbility index. Figure 6-18. Predicted probabilities for MVPA ≥ 30min/day at different levels of walkability and MAPS grand scores for high- and low-income neigbourhoods 157  6.3.3.3 Supplementary Analysis Using Valence Scores and Subscales Table 6-12 summarizes the results of the interactions between neighbourhood income, walkability, and MAPS valence scores and subscales, which were significant at the 0.01 alpha level. Overall crossing score and curb quality showed a significant interaction with walkability index and gender. A graphical representation of the interactions is shown in Figure 6-19.   Table 6-12. SNQLS: Results of the three-way interaction between walkability, neighbourhood income, and MAPS valence and subscales for objective physical activity.  MAPS Scores (Valence scores and Subscales) ‡ Fixed Effects of Key Variables   Overall Crossing Score Curb Quality  Log Odds Log Odds  (SE) (SE) Income (High) 0.29 -0.07   (0.66) (0.72) Walkability Index -0.35 -0.32   (0.24) (0.25) MAPS Score† 0.07 0.12   (0.12) (0.45) Walkability Index × MAPS Score† -0.14 -0.45   (0.06) (0.21) Income (High) × Walkability Index 0.28 0.05   (0.29) (0.30) Income (High) × MAPS Score† 0.21 1.29   (0.16) (0.60) Income (High) × Walkability Index × 0.25* 0.92*  MAPS Score† (0.08) (0.28) † MAPS Score refers to ‡ MAPS label in each column.  * p-value < 0.01. All p-values were adjusted for false discovery rate.  All models adjusted for age, gender, ethnicity, mobility status, and regional accessibility. Refer Appendix B  for full model results; Appendix C  for VIF for final model.       158   6.4 Summary of Results Table 6-13 shows a final summary of the results for MAPS grand score and domain-specific scores. There were eight significant interactions in total. For transport walking, data from older adults showed a significant interaction between gender, walkability index and MAPS grand score, and the MAPS active transport score. There was no significant interaction for leisure walking in older adults and neighbourhood physical activity in children and teens. In the case of objective physical activity, the interaction between parental support, the walkability index, and MAPS grand score was significant in children, and the interaction between neighbourhood income, the walkability index, and MAPS grand score was significant in older adults. The MAPS active transport also showed a significant interaction with neighbourhood Curb Quality Crossing Overall Figure 6-19. Graphical representation of the interactions between neighbourhood walkability, income and MAPS valence and sub scores. High and Low walkability refer to values 1 standard deviation above and below the mean walkability index.   159  walkability and income to explain objective physical activity in older adults. Teen data showed a significant two-way interaction between neighbourhood walkability and MAPS leisure score for objective physical activity.  Table 6-13. Summary of results. Behaviour Significant Interactions Children (NIK) Teens (TEAN) Older adults (SNQLS) Walking or Biking for Transport None None Gender × Walkability Index × MAPS Grand  Gender × Walkability Index × MAPS Active Transport Leisure Walking NA NA None Neighbourhood Physical Activity None None NA Objective MVPA Parent Support × Walkability Index × MAPS Grand  Parent Support × Walkability Index × MAPS Leisure  Walkability Index × MAPS Grand Leisure Income × Walkability Index × MAPS Grand  Income × Walkability Index × MAPS Active Transport          160  CHAPTER 7: DISCUSSION AND CONCLUSION 7.1  Introduction This concluding chapter discusses the results within the context of the broader literature and provides policy recommendations. It also reviews the limitations of this study and suggestions for future research. There are five sections in the chapter. The first section discusses the results for each study sample, followed by a section that discusses the study's limitations. The third section discusses the policy implication of the study. The fourth and fifth sections discuss implications for future research and overall conclusion, respectively.  7.2 Discussion of Results This study examined whether the health benefits of neighbourhood walkability in children, teens, and older adults are higher in neighbourhoods with a supportive pedestrian environment. There were three objectives set to achieve the aim of this study. Within each objective were sets of hypotheses that were tested using the regression models. The objectives and corresponding hypotheses are listed below: Objective 1: To examine whether there is an interaction between neighbourhood walkability and the pedestrian environment in explaining walking and physical activity in children, teens, and older adults. Hypothesis: Street design qualities and neighbourhood walkability interact to affect physical activity. The interaction effect is most synergistic when both street design qualities and neighbourhood walkability are high.  Objective 2: To examine whether the interaction between neighbourhood walkability and the pedestrian environment varies by gender and income across the three age groups. 161  Hypothesis: The interaction between neighbourhood walkability and the pedestrian environment is synergistic in males compared to females and is synergistic in high-income neighbourhoods compared to low-income neighbourhoods Objective 3: To examine whether the interaction between neighbourhood walkability and the pedestrian environment varies by parental/family support, self-efficacy, and safety perception.  Hypothesis 1: In children and teens, the interaction between street design qualities and neighbourhood walkability is synergistic when their parents support them to do physical activity, whereas, in older adults, interaction is synergistic when their self-efficacy and social support are high. Hypothesis 2: In children and teens, the interaction between street design qualities and neighbourhood walkability is synergistic when their parents have a high perception of neighbourhood safety, whereas, in older adults, the interaction is synergistic when their perception of neighbourhood safety is high.  The remaining portion of this section discusses the results of the final models for each study sample. Only significant results are discussed.  7.2.1 Sample 1: Neighbourhood Impacts of Kids (NIK) The NIK sample results showed one significant interaction for MAPS grand score and one for MAPS leisure score, both related to non-school hours MVPA (objective 3, hypothesis 1). The remaining portion of this section discusses the result related to objective physical activity.  Objective Physical Activity: Children’s moderate to vigorous physical activity showed a positive relationship with the pedestrian environment in high and low walkability neighbourhoods when parental support for physical activity was high. The association was stronger in high walkability neighbourhoods as 162  shown by the slope of the line in Figure 6-2 and change in predicted probability from 4.28% to 33.22% as MAPS score increased from low to high as shown in Figure 6-3. However, the results also showed a higher likelihood of meeting daily MVPA guidelines during non-school hours among children in low walkability neighbourhoods (29.61% to 43.79% from low to high MAPS score; Figure 6-3). The high probability in less walkable neighbourhoods can be attributed to high parental support for physical activity. Studies have shown that high parental support for physical activity positively correlates with children’s physical activity (Langer et al., 2014; Loprinzi & Trost, 2010; Rhodes et al., 2019). Furthermore, since MVPA captures a broader range of physical activity, children living in low walkability neighbourhoods may be accumulating MVPA through structured physical activity supported by their parents. Parents can support their kids’ physical activity through encouragement, logistical support and co-activity (Rhodes et al., 2015; Rhodes, Perdew, et al., 2020). This phenomenon may be at play in explaining the higher probability of non-school MVPA engagement among children in low walkability neighbourhoods.  Similar to the results for high parental support, children living in both high- and low-walkability neighbourhoods showed a positive response to street design qualities with regard to their physical activity even with low parental support for physical activity (Figure 6-2 and Figure 6-3). Interestingly, the relationship was more pronounced in children living in low walkability neighbourhoods. The results showed that for children with low parental support and living in less walkable neighbourhoods, a higher MAPS score was associated with a higher probability of engaging in MVPA during non-school hours. While this may look counterintuitive initially, studies have shown that children living in less walkable neighbourhoods may benefit from having pedestrian infrastructures. A study by Mecredy, Pickett and Janssen (2011) found that 163  children living in neighbourhoods with less connected street patterns, a typical feature of a less walkable neighbourhood, were more likely to be physically active outside their school. Poor street connectivity with cul-de-sacs and less traffic may act as outdoor play-spaces for children to be physically active (Ding et al., 2011; Handy et al., 2008). Additionally, low walkability sub-urban neighbourhoods with pedestrian amenities tend to be new suburban neighbourhoods that are generally popular among families with kids.  This may create a social fabric in sub-urban neighbourhoods leading to more peer support or sibling support associated with physical activity in children (Blazo & Smith, 2018; M. J. Edwards et al., 2015). This factor may be acting as a confounder of the relationship observed in this study and warrants further investigation.  The results from the MAPS leisure score, MAPS valence score and subscale scores shed light on some of the pedestrian environment's details that may help target MVPA in children. The MAPS leisure score includes factors related to recreational land uses (private and public), positive aesthetics and social environment, sidewalk characteristics, buffers in the street, bike infrastructures, trees in the street, building aesthetics and design. Access to recreational space is found to be associated with physical activity in children. Similarly, sidewalk buffers, trees along the street, building aesthetics, and design help create safe play spaces in the neighbourhood for children to participate in non-structured physical activity such as free play (Lambert et al., 2019; R. E. Lee et al., 2016; Tranter, 2015).  Urban planners can target these aspects of the pedestrian environment to increase MVPA in children.  7.2.2 Sample 2: Teen Environment and Neighborhood (TEAN) The results of the TEAN data showed one significant interaction for objective physical activity. There was a significant two-way interaction between walkability index and MAPS 164  leisure score (objective 1). The following section discuss the results on objective physical activity.  Objective Physical Activity: The results showed that teens are more likely to spend at least 30 minutes daily doing moderate to vigorous physical activity during non-school hours when both neighbourhood walkability and the pedestrian environment as measured by the MAPS leisure score are high (Figure 6-7 & Figure 6-8). The predicted probability of doing at least 30 minutes of daily non-school MVPA among teens living in highly walkable neighbourhoods increased from 18.64% to 30.17% with increased MAPS leisure scores (Figure 6-7). The MAPS leisure score includes factors related to recreational land uses (private and public), positive aesthetics and social environment, sidewalk characteristics, buffers in the street, bike infrastructures, trees along the street, building aesthetics and design. This shows that physical activity in teens living in highly walkable neighbourhoods is associated positively with the pedestrian environment that has recreational land uses, sidewalk amenities, bike infrastructures and appealing building aesthetics. Based on prior literature, such a combination of the built environment can influence physical activity in two ways.  First, studies have shown that having recreational facilities in close proximity is positively associated with physical activity (Frank et al., 2007; Norman et al., 2006).Similarly, a study by Thornton et al. (2017) used the same dataset used in this dissertation and found that teens were more likely to report (self-reported measure) participating 60 min/day if they had access to recreational environment in their neighbourhoods. Second, having recreation facilities within walking distance can reduce the need to drive to these locations, which increases the amount of physical activity accumulated by walking or biking to these destinations (Sallis et al., 165  2004). Though a synergistic interaction between walkability and pedestrian environment is seen in highly walkable neighbourhoods, an inverse relationship was observed in the low walkability neighbourhoods.  Teens living in less walkable neighbourhoods showed an inverse relationship between the pedestrian environment measured by the MAPS leisure score and physical activity (Figure 6-7 and Figure 6-8). This suggests that in some cases having a supportive pedestrian environment may not be enough to offset the low neighbourhood walkability. One possible explanation could be related to the macro level neighbourhood environment characteristics related to walkability. Less walkable neighbourhoods tend to lack the appropriate infrastructure, such as street connectivity, mix and density of different land uses with various destinations, which may be required for the benefits of a supportive pedestrian environment to be effective. As mentioned before, Frank et al. (2007) and others (Norman et al., 2006) found that teens walked more if there were various destinations in their neighbourhood. Additionally, since MVPA captures a broad range of activities, other factors may also be acting and confounders of this finding. Further investigation is needed to understand the underlying mechanism of the interaction effect.  Nevertheless, this study's finding supports that investing in a pedestrian environment may be effective in areas where macro walkability-related infrastructures are already in place.  7.2.3 Sample 3: Senior Neighbourhood Quality of Life Study (SNQLS) The results of the SNQLS sample showed significant interactions for transport walking and objective physical activity. There was a significant three-way interaction between neighbourhood walkability, MAPS grand score and gender, neighbourhood walkability, MAPS active transport score and gender for transport walking. Similarly, neighbourhood walkability, MAPS grand score, and neighbourhood income showed a significant three-way interaction for 166  objective physical activity. MAPS active transport score also showed a significant three-way interaction with neighbourhood income and walkability. The results are discussed in the following sections.  Transport walking: The final model for transport walking showed a positive interaction between neighbourhood walkability, MAPS grand score and gender (female) (Figure 6-10). The interaction between neighbourhood walkability and the pedestrian environment was synergistic in females. The probability of transport walking in females increased from 39.19% to 72.50% with increased MAPS grand scores in high walkability neighbourhoods (Figure 6-9). A similar pattern was seen in the MAPS active transport score as well (Figure 6-12). Though these results are counterintuitive to the hypothesis that males show more synergy than females, it makes sense because of the gender-specific response to urban design features. Urban theorists like Jane Jacobs and Willian Whyte have argued that a better pedestrian environment is crucial for neighbourhood livability. “Eyes on the streets,” one of the key features of urban design for visual connection between a street with the building, is essential for women and children as it creates a sense of safety (Jacobs, 1961).   Similarly, urban spaces with better design tend to attract more women (Whyte, 1980). Studies have also shown that women are more sensitive to smaller details of the built environment, including the pedestrian environment's features (Golan et al., 2019; Jensen et al., 2017; L. Luo et al., 2008; Pelclová et al., 2013). Jensen et al. (2017) examined how females responded to the various street environment features and found a higher proportion of females using streets with a good pedestrian environment. Streets with better pedestrian amenities make the walking experience positive among women (Golan et al., 2019). Having more urban design 167  features and amenities that make the pedestrian environment safer and more friendly could increase the likelihood of walking or biking for transportation among women.  The results for males showed a different pattern compared to females. With the increase in the MAPS grand score, the probability of walking or biking for transport increased from 1.55% to 36.79% in the less walkable neighbourhood than from 45.70% to 56.31% in highly walkable neighbourhood (Figure 6-9). In low walkability neighbourhoods, the lack of destinations may be offset by some of the positive characteristics of the MAPS grand score resulting in a higher level of transport walking or biking. MAPS grand score also includes some of the land use and destination characteristics that are essential for transport walking in low walkability neighbourhoods. Additionally, less stress of high density in less walkable neighbourhoods may further facilitate transport walking among older adults living in less walkable neighbourhoods with a good pedestrian environment (Curl et al., 2020; Van Cauwenberg et al., 2012). However, this result presents a significant challenge to policy considerations as the less walkable neighbourhood sprawl over a large area may not be cost-effective (Gunn et al., 2014; Veerman et al., 2016). Suburban retrofitting to increase density and mix of different land uses may be crucial to harnessing the synergy between neighbourhood walkability and pedestrian environment to improve active transport in less walkable neighbourhoods.  Though the change in predicted probability for transport walking or biking in a less walkable neighbourhood was high compared to a high walkability neighbourhood, the probability of transport walking or biking in a highly walkable neighbourhood was still higher (Figure 6-9). This pattern is consistent with the existing literature that has shown a positive association of walkability and street design features with active transport in older adults 168  (Alidoust et al., 2018; Koohsari et al., 2020; Marquet & Miralles-Guasch, 2015). The positive interaction, i.e., synergy, between MAPS grand score and transport walking in a highly walkable neighbourhood also has important population health implications. Due to the high population density in walkable neighbourhoods, many people stand to benefit from improvements in the pedestrian environment (Gunn et al., 2014; Veerman et al., 2016).   The supplementary analysis results using MAPS valence scores and subscales showed a similar pattern seen in the MAPS grand score. For females living in a highly walkable neighbourhood, positive streetscape, overall streetscape score, route overall score, and crossing overall score showed a positive relationship with transport walking or biking (Figure 6-13 and Figure 6-14). In low walkability neighbourhoods, negative aspects of the pedestrian environment reflected by a negative streetscape score and crossing impediments showed inverse relationships with active transport in females. The results for males did not reveal a clear pattern in highly walkable neighbourhoods (Figure 6-13 and Figure 6-14). For example, in highly walkable neighbourhoods, male active transport showed a downward trend with increased positive and negative streetscape scores and a flat trend for overall streetscape scores. However, the probability of active transport in a highly walkable neighbourhood was still higher in a highly walkable neighbourhood than a low walkable neighbourhood regardless of the MAPS valence and subscale scores. These patterns in males need to be further investigated.   Objective physical activity: A significant interaction between neighbourhood walkability, MAPS grand score and neighbourhood income was observed in models examining the likelihood of older adults spending at least 30 minutes a day doing MVPA. Results showed that older adults living in high and low-walkability neighbourhoods benefit from better street design qualities when 169  neighbourhood income is high. In such a case, older adults in the highly walkable neighbourhoods had a slightly higher advantage over low-walkability neighbourhood residents. However, the association was small as there was only about a 2% difference in the predicted probability of spending at least 30 minutes doing MVPA daily between older adults living in high walkability with low MAPS grand score and those living in high walkability and high MAPS grand score neighbourhoods (Figure 6-18). The results for low-income neighbourhoods showed a complicated pattern (Figure 6-18).  For low-income neighbourhoods that are less walkable, a supportive pedestrian environment was more strongly associated with higher MVPA. However, older adults living in low-income areas that were highly walkable showed a decrease in the likelihood of spending time in MVPA, even with the improvement in the pedestrian environment. This inverse relationship in high walkability neighbourhoods may have to do with the perception of the built environment in low-income neighbourhoods. Studies have shown that people living in a low socio-economic status neighbourhood tend to perceive their built environment as less favourable than their counterparts living in a neighbourhood with high socioeconomic status (Kamphuis et al., 2010; Sallis et al., 2011). Another possible reason for this counterintuitive result is related to the social environment in the low-income high walkability neighbourhoods (Kamphuis et al., 2010). Low-income high walkability neighbourhoods may not have destinations appropriate for older adults or enough social support for physical activity compared to high-walkability high-income neighbourhoods. The lack of favourable physical and social resources may deter spending time doing MVPA (Van Holle et al., 2016). The positive effect of a better pedestrian environment in a high walkability neighbourhood may be neutralized by an unfavourable social environment in low-income neighbourhoods. It may also be possible such a phenomenon may 170  not be in play in less walkable neighbourhoods resulting in a positive association between pedestrian environment and MVPA. Additionally, MVPA captures multitudes of physical activity that may be influenced by factors other than the built environment. Further study is required to disentangle the three-way interaction between neighbourhood income, walkability and the pedestrian environment. (Kolbe-Alexander et al., 2015; R. E. Lee et al., 2007; Mooney et al., 2017).  A similar pattern was seen in the interaction observed for MAPS active transport score and some of the MAPS valence and subscale scores. For high-income neighbourhoods, living in both high and low walkability neighbourhoods, there was a positive association between MAPS active transport score and MVPA (Figure 6-17). An inverse association was observed in low-income neighbourhoods. The supplementary analysis showed overall crossing and curb quality scores having significant three-way interactions with neighbourhood walkability and income. The patterns of interactions (Figure 6-19) were similar to MAPS grand score results (Figure 6-15). The significant interaction with crossing scores may be related to transport walking since similar interaction was observed between neighbourhood walkability, gender and crossing score (Table 6-7). However, this relationship can only be validated using a causal model such as mediation analysis.  7.3 Limitations  This research used data from three different studies, which were based on robust study designs and validated instruments to measure the built environment and various behavioural outcomes. There are, however, several limitations to this study.  First, the data used in this study are cross-sectional, which prevents us from making any causal interpretations of the findings. Though some studies have used cross-sectional data and 171  performed multivariate analysis using path models to establish a causal inference, such an approach does not allow us to make a correct causal inference. Second, the neighbourhood selection technique used in all three studies was purposeful, and some selection bias is inherent in the research design. Despite the research bias, the sampling strategy also allowed researchers to capture the maximum variation in the built environment to examine its effect on behaviour. Because of the funding and logistical challenges, purposeful sampling is an efficient way to capture maximum information on the built environment and behaviour (Frank, Sallis, et al., 2010). The third limitation is the differences in the study design and data collection tools used in the three studies. This limitation did not allow the use of a pooled analysis that would have given higher power and reduce the chances of type I errors. However, similarities in sampling strategy to select neighbourhoods and the use of same built environment metrics and similar outcome variables allowed us to make some cross-comparison across the age groups. Having a larger sample size and using more advanced statistical methods like Bayesian modelling techniques may result in a more accurate prediction of the interaction effects.  The fourth limitation is the use of a self-reported measure of active transport and neighbourhood physical activity for children, which was reported by their parents. These reports may not reflect the participants’ actual behaviour, but the tools used to capture this information have been validated in previous research. The fifth limitation is related to the measurement of the pedestrian environment. The pedestrian environment data were collected only for a 1/4-mile route from each participant’s home to predetermined locations using a short path. These data may not represent the actual street-built environment in the neighbourhoods. It was challenging to measure all streets in a neighbourhood since most data collection was done through in-field visits. Recently, studies have started to collect street data via online mapping services like 172  Google Street View in conjunction with artificial learning tools for image processing (Badland et al., 2010; Ben-Joseph et al., 2013; Clarke et al., 2010). Future studies can use these approaches to collect street data for all streets in the neighbourhood.  There was a temporal gap in the measurement of physical activity patterns and micro-scale built environment data. This gap may lead to changes in the pedestrian environment between physical activity surveys and measures of the street environment. Likewise, the geographic context within which the studies were conducted and the study sample's demographic composition may not allow generalizability other urban settings. Lastly, using different distance thresholds for the street network buffer to calculate neighbourhood walkability could shed more insight into how the interaction effect change in different spatial scales. Despite these limitations, the use of validated survey questionnaires and accelerometers to measure physical activity allowed us to make a strong inference. Likewise, the use of neighbourhood walkability and MAPS street audit tools to measure the built environment has been validated by various studies, and the scoring system can be simplified to help make policy decisions.  Lastly, interaction effects are highly complex phenomena and are inherently complicated to identify and interpret. For ease of interpretation, sub-group analysis is commonly used to test interactions (R. Wang & Ware, 2013). Subgroup analysis is done by splitting the data into groups based on a predetermined cut-point for continuous data or using the grouping factor for categorical data. However, such an approach has several limitations (Dawson & Richter, 2006). First, splitting the sample into small groups decreases power, thereby increasing Type I error chances. This is problematic in cases where the sample size is already small. Second, the risk of Type I error further increases because of multiple tests in the sub-groups. Third, splitting data can also change the distribution variables in the sub-groups leading to biased estimates. Fourth, 173  sub-group analysis does not allow comparison across groups, which restricts interpretation within the sub-group. Because of these limitations, sub-group analysis was not conducted in this dissertation.  7.4 Importance of Studying Interactions for Planning and Policy Formulation Studying interactions is important for planning and policy formulation for various reasons. Studying interactions is at the very heart of theory testing in the social sciences (Cohen, Cohen, West, & Aiken, 2003) including disciplines such as urban planning. Interaction tests help enrich the understanding of the built environment and physical activity relationships by understanding how such relationships vary under different conditions. These conditions can be demographic, such as gender and income, as well as psychosocial, such as social support, perception of safety, etc. Various studies have shown that the relationship between macro built environments, such as neighbourhood walkability, and physical activity related health outcomes is disproportionate across demographic and psychosocial factors (Adkins et al., 2017; Gullon et al., 2020). Understanding the variation in the relation between neighbourhood built environment and physical activity under different conditions helps identify subgroups that may benefit most from an intervention. Therefore, by identifying significant interactions, planners and policymakers can formulate policies targeted towards high-risk sub-group such as children, teens and seniors.   Interaction effects are essential from a planning and policy perspective (VanderWeele & Knol, 2014) because interactions show an intervention's multiplicative effect. This was also discussed in chapter (Section 1.2). Another demonstration, using the results on the interaction between MAPS grand score, neighbourhood walkability and gender for transport walking in the SQNLS dataset, is shown in Figure 7-1 and Figure 7-2. Figure 7-1 shows the result of the main 174  effect model reported in Table 6-7. Figure 7-2 shows the results of the final model in the same table (Table 6-7) and reported in Figure 6-10. When the interaction effect is not examined, the relationship between the MAPS grand score and transport walking/biking shows a similar trend for both males and females in high and low walkability neighbourhoods. There is a constant difference in the likelihood of walking for males(ΔM) and females (ΔF) living in high and low walkability neighbourhoods with increased MAPS grand score. This difference is constant both at high and low MAPS scores, shown by the parallel lines. This constant difference reflects the additive effect.  However, when there is an interaction between neighbourhood walkability, MAPS grand score and gender, the relationship between MAPS grand score and transport walking/biking in older adults differs by levels of walkability (low vs high) and by gender (male vs female).  This difference is shown by the intersecting lines shown in Figure 7-2 as opposed to the constant difference (additive effect) in the likelihood of walking/biking (Figure 7-1), the difference in the likelihood of walking/biking with increasing MAPS grand score is different at different levels of walkability (high vs low) and gender (male vs female). This phenomenon reflects the multiplicative effect through the three-way interaction.  For example, the difference in the likelihood of walking/biking for transportation among older females living in high vs low walkability neighbourhoods is ΔF1 for low MAPS score and ΔF2 for high MAPS grand score. ΔF1 is less than ΔF2, which shows that investing in pedestrian environment improvement may have a higher effect in highly walkable neighbourhoods than low walkable neighbourhoods. Therefore, by examining interactions, planners identify different policy needs basing on the different patterns of relationships between neighbourhood walkability and pedestrian environment and active transportation in various demographic groups.  175       Figure 7-2. Interaction between neighbourhood walkability, MAPS grand score and gender for transport walking/biking. A case of interaction between neighbourhood walkability, MAPS grand score and gender.  Figure 7-1. Relation between neighbourhood walkability, MAPS grand score in males and females for transport walking/biking. A case of no interaction between neighbourhood walkability, MAPS grand score and gender.  Δ M Δ M Δ F Δ F Low MAPS High MAPS Low MAPS High MAPS Δ M1 Δ M2 Δ F1 Δ F2 Low MAPS High MAPS Low MAPS High MAPS   176 7.5 Implications for Planning and Policymaking City planners and policymakers have started to recognize the effects of their decisions on active transportation, physical activity and related health outcomes. The results of this study have implications for cities and municipalities. The pedestrian level built environment features fall within the jurisdiction of local municipalities (Steinmetz-Wood et al., 2020) (citation). Cities use codes and ordinances to regulate the form and characteristics of neighbourhoods. Zoning code is the most common tool that provides cities with a mechanism for regulating development in communities in their jurisdiction. Zoning codes look at the macro-level development pattern such as density, land use mix and street connectivity. Besides, cities also use various guidelines to regulate the neighbourhoods' form and characteristics not covered by the zoning codes (Duerksen et al., 2017). These guidelines may include urban design guidelines, complete street guidelines, and form-based codes. These guidelines and codes provide detailed information about the pedestrian environment design features and are intended to fill the gaps not covered by regulatory Zoning codes (Duerksen et al., 2017). By examining the interactions between neighbourhood walkability and pedestrian environment, this study showed that a better pedestrian environment could harness neighbourhood walkability's health potential.  7.5.1 Implications for Urban Design Guidelines Urban design guidelines set urban design and development recommendations for public and city builders (City of Edmonton, 2020). Cities use urban design guidelines and regulations to shape urban spaces' physical and visual quality and the built environment (Kumar, 2002). The city of Seattle, King County USA, lists the purpose of their design guidelines as follows: “define the qualities of architecture, urban design, and public space that make for successful projects and communities, and to serve as a tool for guiding individual projects to meet those expectations through the City’s Design   177 Review Program. Urban design guidelines are usually targeted for certain neighbourhoods in a city and have specific guidelines about design standards” (City of Seattle, 2013, iv).  Seattle’s design guideline also recommends design approaches and solutions at the pedestrian level to enhance neighbourhood walkability. These include considering ‘eyes on the street,’ lighting for safety and street-level transparency (Figure 7-3). Similarly, building design guidelines recommend architectural and façade composition, building materials, signage, etc. These features are captured by the MAPS tool used in this study—streetscape, aesthetics and social aspect, route characteristics. The supplementary analyses showed that MAPS valence scores and subscales related to route characteristics, aesthetics, and social aspect of streetscape play a significant role in active transport in seniors and physical activity in seniors and children.  Figure 7-3. Eyes on the street urban design guidelines used by Seattle. (Source: City of Seattle, 2013)   178   Urban design guidelines can be implemented citywide (such as in the City of Seattle) or be targeted to a specific neighbourhood in a city. The most common areas are downtown and historic districts. For example, San Diego, one of the study areas, has urban design guidelines for its downtown area (Perkins+Will et al., 2011). Figure 7-4 shows examples of the guideline for street design, building and block design, and building materials used in the City of San Diego. Neighbourhood-specific design guidelines are particularly important as they can also target certain groups of populations based on their socio-demographic status at the neighbourhood level. Studies from public health have looked at neighbourhood factors and their relationship with health outcomes (Halonen et al., 2012; Pedigo et al., 2011) by identifying neighbourhoods at higher risk; cities can implement area-specific urban design guideline to create a pedestrian-friendly street environment associated with increased physical activity.    The qualities of the pedestrian realm may vary, but all should exhibit certain characteristics, including an Edge Zone, Furnishings Zone, Throughway Zone, and Frontage Zone (where possible). Vertical + Horizontal Modulation Vertical Plane Modulation Horizontal Plane Modulation High standard materials to avoid weathering and staining and minimize deterioration. A well-defined building base and glazed facades and balconies, to create a people-oriented development.  Durable upgraded material for building base. Street Design Block and Building Design Building Materials Figure 7-4. Downtown design guidelines examples used by City of San Diego. (Source: City of San Diego)   179 7.5.2 Implications for Form-Based Code The results of this study can also be incorporated as evidence to formulate form-based codes. Unlike urban design guidelines, a form-based code is a regulatory code that gives cities power like zoning regulation to control development patterns. Form-based codes go beyond conventional zoning and incorporate details of the built environment by addressing the relationship between building facades and the public realm, the form and mass of buildings in relation to one another, and the scale and types of streets and blocks (Form-Based Codes Institute, 2020) (Figure 7-5). This study's results have implications for public space standards, building standards, and architecture standards, which are commonly used in form-based codes.  Figure 7-5. Difference between how conventional zoning and form-based codes regulate urban form (Source: Form-Based Codes Institute, 2020).    The public space standards specify design codes for the public realm; building standards regulate the features, combination and functions of the buildings; and architecture standards set design   180 standards that characterize the public realm (Sitkowski & Russell, 2007). The public realm includes various features and amenities of the pedestrian environment, such as sidewalks, street lighting, sitting, bike lanes, etc. Similarly, building standards regulate the building height, coverage and building use standards. Architecture standards govern the permitted building styles, details, and materials. Some of the results of this dissertation speak to these standards. For example, the results on children and older adults showed that streetscape, aesthetics and social environment play a significant role in physical activity and active transport. Streetscape characteristics measured by the MAPS tool used in this dissertation covers streetlight, sitting arrangements, buildings overhangs, traffic calming measures, etc. Similarly, the aesthetic and social environment covers urban design features such as fountains, sculptures or public/private art, and soft landscapes. These aspects can be incorporated into the public realm, building standard as well as building design standards.  Like urban design guidelines, form-based codes can be formulated at the city level or specific to certain neighbourhoods or street corridors [example:(City of Palm Desert, 2017), Figure 7-6]. Form-based codes can be transect-based, building type-based or street-based codes. Transect based codes are based on the rural-urban continuum; building type-based codes are organized to regulate development based on building types such as townhouses, apartments, etc.; street-based codes are based on street types boulevards, arterial, and collectors (Chicago Metropolitan Agency, 2013).  Regardless of the types of form-based codes, the overarching goal of form-based codes is to create compact, mixed-use, pedestrian-friendly neighbourhoods that are livable, vibrant and healthy. The findings of this dissertation lend supportive evidence for this overarching goal.     181  7.5.3 Implications for Complete Street Guidelines  Another implication of this study is related to complete street guidelines. A Complete Streets approach creates an integrated transportation system that supports safe travel for people of all ages and abilities (Smart Growth America, 2016). Complete streets are associated with higher levels of walking and physical activity (B. B. Brown et al., 2016; S. A. Carlson et al., 2017). This study's specific implications can be on the street's physical features and elements, accommodating the needs for different user groups, and street design based on the context, i.e., level of urbanization. As discussed by Gregg and Hess (Gregg & Hess, 2019), these aspects are usually incorporated in street design policies by various municipalities in the US.  For example, on their official website, the City of San Diego mentioned that “complete streets focuses on enabling safe, attractive and comfortable access so that pedestrians, bicyclists, motorists Figure 7-6. Example of form based code to regulate public space in commercial side street. (Source: City of Palm Desert, 2017).   182 and transit users of all ages and abilities can safely travel within the public right-of-way” (City of San Diego, 2020). Accordingly, they have adopted complete guidelines for their downtown area. Complete street guidelines are important for downtown areas since downtown areas have higher walkability levels and tend to have heavy traffic. This can put pedestrians and bicyclists at higher crash risk (Osama et al., 2020). Such a situation can put vulnerable age groups like children, teens and older adults at even higher risks. Complete street guidelines can also be formulated based on street types. The results of this study showed street crossing characteristics to be important for active transport and physical activity in older adults. Crossing characteristics include features related to crosswalk amenities, curb quality, intersection control, which fall under complete street design guidelines [for example, see Boston Complete Street Design Guidelines (City of Boston, 2013), Figure 7-7].   Figure 7-7. Example of complete street design guideline with focus on intersection design.  (Source: City of Boston, 2013)   183 7.6 Implications for Canadian City Planners Though this study's results are specific to the study area and have the inherent limitation of generalizability, Canadian city planners can use the findings of this study to inform their city planning and neighbourhood design strategies. Like the United States, Canadian cities are facing the problem of high auto dependence, lower levels of active transportation and a higher proportion of the population that does not meet recommended levels of physical activity. In Canada, 47% of children age 5-11 years, 31% of youth age 12-17 years meet 60 minutes/day guideline for MVPA (Statistics Canada, 2019), and 40.8% of Canadians 65 years and over meet the weekly 150 minutes of MVPA (Statistics Canada, 2020). City planners can use tools like urban design guidelines, form-based codes and complete street guidelines to enhance walkability in their jurisdiction.  7.7 The Rationale for Investing in the Pedestrian Environment to Optimize Macro-level Walkability The rationale for investing in the pedestrian environment to optimize macro-level walkability is the ease of changing the pedestrian level built environment features. Making changes in the pedestrian environment falls within the municipal government's budget and jurisdiction (Steinmetz-Wood et al., 2020) and is a part of regular street maintenance works (D. R. Young et al., 2020). This allows cities to improve the pedestrian environment in an expedited manner through their local infrastructure upgrade projects. Local infrastructure upgrade projects are usually completed within a short time; for example, see the City of Vancouver BC street improvement projects (City of Vancouver, 2020).  In addition, cities can use an experimental approach to test whether making changes in the pedestrian environment can be useful or not. Such an experimental approach can be done by making temporary installation of pedestrian infrastructures. The City of New Westminster, BC Canada used a similar approach by installing a temporary greenway in one of the streets in the   184 city (City of New Westminster, 2020). This approach allows users to experience the pedestrian environment and provides feedback to the city on what works and what does not, which gives city information to make the pedestrian environment better for its users. These flexibilities which are not available when making macro-level changes such as new road construction, neighbourhood densification, and transit line construction make pedestrian environment changes an effective tool to harness macro level walkability’s health potential. 7.8  Future Research  Future studies can include intervention research to study how a modification to the pedestrian environment can impact physical activity—using cross-sectional data in the current dissertation does now allow making any causal inferences. Conducting natural experiments evaluating temporary design modifications in pedestrian environments to study how it affects pedestrian behaviour could be one intervention approach. Some examples of this approach include popular events like tactical urbanism movement (Lydon & Garcia, 2015), PARK(ing) Day (Littke, 2016), Pop-Up city events (Greco, 2012), which involve temporary intervention in the street environment. Conducting these events in high and low walkability neighbourhoods using a case and control study design approach could be an effective way to establish causal evidence on the role of the pedestrian environment and neighbourhood walkability on physical activity.  The small sample size was another limitation of this study. Using a larger sample size that offers enough power to examine the interactions between neighbourhood walkability and street design qualities can address some of the limitations inherent to small sample size. Though there were significant interaction effects even after controlling for FDRs, a larger sample size would provide better results with higher accuracy. Using the other statistical method based on the Bayesian approach could also shed some more insights into the interaction effects observed in this study.    185 Likewise, collecting more data on the street environment covering a larger geography of the neighbourhood could help researchers better understand the street design qualities' role on neighbourhood walkability and physical activity. In addition to measuring the street design qualities quantitatively, studies can also look at the association of the pedestrian environment and physical activity using qualitative methods. Measuring the aesthetic quality of the pedestrian environment quantitatively may not be enough.  7.9 Conclusion  This study shows that the pedestrian environment has the potential to enhance the association between neighbourhood walkability and physical activity in children, teens, and older adults. Because the built environment is a modifiable correlate of physical activity, investing in improving the pedestrian environment can be a practical approach for planners and policymakers to create active communities (Sallis et al., 2012). The current urban form is a result of the large investments made by cities and regional agencies for many decades. The maintenance and modification of the infrastructure in urban areas are among the major challenges for cities and regional planning agencies (Talen, 2011). Making small-scale changes to improve the pedestrian environment may be an effective policy approach for local and regional planning agencies to adopt to create active communities.  Physical activity is a significant part of a healthy life. It is one of the human behaviours essential for individual and community survival and a flourishing life (Mutrie & Faulkner, 2012). Regular physical activity is associated with a decreased risk of coronary heart disease, type 2 diabetes, some forms of cancers, hypertension, obesity, clinical depression, and other chronic disorders (Heath et al., 2012). Despite this evidence, a significant portion of the world’s population still does not meet the recommended daily or weekly physical activity levels (World Health Organization, 2014).    186 Planners and policymakers have started to recognize the effects of their decisions on active transportation, physical activity and related health outcomes. However, the century-long land use and transportation planning approaches have created urban forms that will require significant efforts to become favourable for active transportation and physical activity (American Planning Association, 2015). Making small-scale changes at the street level may be an effective approach to fine-tune some of the existing infrastructures that favour active transport. This needs confirmation in future experimental research.  The evidence provided in this study, though not fully generalizable, provides a framework to formulate plans and policies to design better pedestrian environments and retrofit the existing urban infrastructure to create better environments for physical activity.                  187 REFERENCES Active Living Research. (2015). Tools and measures. http://activelivingresearch.org/toolsandresources/tools Adams, M. A., Sallis, J. F., Conway, T. L., Frank, L. D., Saelens, B. E., Kerr, J., Cain, K. L., & King, A. C. (2012). Neighborhood environment profiles for physical activity among older adults. American Journal of Health Behavior, 36(6), 757–769. https://doi.org/10.5993/AJHB.36.6.4 Adkins, A., Dill, J., Luhr, G., & Neal, M. (2012). Unpacking Walkability: Testing the Influence of Urban Design Features on Perceptions of Walking Environment Attractiveness. Journal of Urban Design, 17(4), 499–510. https://doi.org/10.1080/13574809.2012.706365 Adkins, A., Makarewicz, C., Scanze, M., Ingram, M., & Luhr, G. (2017). Contextualizing Walkability: Do Relationships Between Built Environments and Walking Vary by Socioeconomic Context? Journal of the American Planning Association, 83(3), 296–314. https://doi.org/10.1080/01944363.2017.1322527 Ahlbom, A., & Alfredsson, L. (2005). Interaction: A word with two meanings creates confusion. European Journal of Epidemiology, 20(7), 563–564. https://doi.org/10.1007/s10654-005-4410-4 Ahlport, K. N., Linnan, L., Vaughn, A., Evenson, K. R., & Ward, D. S. (2006). Barriers to and Facilitators of Walking and Bicycling to School: Formative Results From the Non-Motorized Travel Study. Health Education & Behavior, 35(2), 221–244. https://doi.org/10.1177/1090198106288794 Alexander, C. (1977). A Pattern Language: Towns, Buildings, Construction. In Leonardo. Oxford University Press. Alexander, C. (1979). The Timeless Way of Building. In New York Oxford University Press. Oxford University Press. Alidoust, S., Bosman, C., & Holden, G. (2018). Talking while walking: an investigation of perceived neighbourhood walkability and its implications for the social life of older people. Journal of Housing and the Built Environment, 33(1), 133–150. https://doi.org/10.1007/s10901-017-9558-1 Allison, P. (2012). When Can You Safely Ignore Multicollinearity? | Statistical Horizons. https://statisticalhorizons.com/multicollinearity Alparone, F. R., & Pacilli, M. G. (2012). On children’s independent mobility: The interplay of demographic, environmental, and psychosocial factors. Children’s Geographies, 10(1), 109–122. https://doi.org/10.1080/14733285.2011.638173 Alves, F., Cruz, S., Ribeiro, A., Silva, A. B., Martins, J., & Cunha, I. (2020). Walkability index for elderly health: A proposal. Sustainability (Switzerland), 12(18). https://doi.org/10.3390/SU12187360   188 Amagasa, S., Fukushima, N., Kikuchi, H., Takamiya, T., Oka, K., & Inoue, S. (2017). Light and sporadic physical activity overlooked by current guidelines makes older women more active than older men. International Journal of Behavioral Nutrition and Physical Activity, 14(1), 1–7. https://doi.org/10.1186/s12966-017-0519-6 Ameli, S. H., Hamidi, S., Garfinkel-Castro, A., & Ewing, R. (2015). Do Better Urban Design Qualities Lead to More Walking in Salt Lake City, Utah? Journal of Urban Design, 20(3), 393–410. https://doi.org/10.1080/13574809.2015.1041894 American Planning Association. (2015). The Benefits of Street-Scale Features for Walking and Biking the Benefits of Street-Scale Features for Walking and Biking. https://planning-org-uploaded-media.s3.amazonaws.com/legacy_resources/nationalcenters/health/streetscale/pdf/walkingbikingfinalreport.pdf Armstrong, N., & Welsman, J. (1997). Young people and physical activity. Oxford University Press. Austad, S. N., & Fischer, K. E. (2016). Sex Differences in Lifespan. Cell Metabolism, 23(6), 1022–1033. https://doi.org/https://doi.org/10.1016/j.cmet.2016.05.019 Badland, H. M., Opit, S., Witten, K., Kearns, R. A., & Mavoa, S. (2010). Can virtual streetscape audits reliably replace physical streetscape audits? Journal of Urban Health, 87(6), 1007–1016. https://doi.org/10.1007/s11524-010-9505-x Bandura, A. (1971). Social learning theory. General Learning Press. http://www.asecib.ase.ro/mps/Bandura_SocialLearningTheory.pdf Barnett, D. W., Barnett, A., Nathan, A., Van Cauwenberg, J., & Cerin, E. (2017). Built environmental correlates of older adults’ total physical activity and walking: A systematic review and meta-analysis. International Journal of Behavioral Nutrition and Physical Activity, 14(1), 1–24. https://doi.org/10.1186/s12966-017-0558-z Baslington, H. (2008). Travel Socialization: A Social Theory of Travel Mode Behavior. International Journal of Sustainable Transportation, 2(2), 91–114. https://doi.org/10.1080/15568310601187193 Bauer, K. W., Laska, M. N., Fulkerson, J. A., & Neumark-Sztainer, D. (2011). Longitudinal and secular trends in parental encouragement for healthy eating, physical activity, and dieting throughout the adolescent years. Journal of Adolescent Health, 49(3), 306–311. https://doi.org/10.1016/j.jadohealth.2010.12.023 Bauer, K. W., Neumark-Sztainer, D., Fulkerson, J. A., Hannan, P. J., & Story, M. (2011). Familial correlates of adolescent girls’ physical activity, television use, dietary intake, weight, and body composition. International Journal of Behavioral Nutrition and Physical Activity, 8, 1–10. https://doi.org/10.1186/1479-5868-8-25 Bauman, A. E., Reis, R. S., Sallis, J. F., Wells, J. C., Loos, R. J. F., & Martin, B. W. (2012). Correlates of physical activity: Why are some people physically active and others not? The Lancet, 380(9838), 258–271. https://doi.org/10.1016/S0140-6736(12)60735-1   189 Beenackers, M. A., Kamphuis, C. B. M., Mackenbach, J. P., Burdorf, A., & van Lenthe, F. J. (2013). Why some walk and others don’t: exploring interactions of perceived safety and social neighborhood factors with psychosocial cognitions. Health Education Research, 28(2), 220–233. https://doi.org/10.1093/her/cyt002 Beets, M. W., Cardinal, B. J., & Alderman, B. L. (2010). Parental social support and the physical activity-related behaviors of youth: A review. Health Education and Behavior, 37(5), 621–644. https://doi.org/10.1177/1090198110363884 Ben-Elia, E., & Shiftan, Y. (2010). Which road do I take? A learning-based model of route-choice behavior with real-time information. Transportation Research Part A: Policy and Practice, 44(4), 249–264. https://doi.org/https://doi.org/10.1016/j.tra.2010.01.007 Ben-Joseph, E., Lee, J. S., Cromley, E. K., Laden, F., & Troped, P. J. (2013). Virtual and actual: Relative accuracy of on-site and web-based instruments in auditing the environment for physical activity. Health and Place, 19(1), 138–150. https://doi.org/10.1016/j.healthplace.2012.11.001 Benjamini, Y., & Hochberg, Y. (1995). Controlling the False Discovery Rate: A Practical and Powerful Approach to Multiple Testing. Journal of the Royal Statistical Society. Series B (Methodological), 57(1), 289–300. http://www.jstor.org/stable/2346101 Bennetts, S. K., Cooklin, A. R., Crawford, S., D’Esposito, F., Hackworth, N. J., Green, J., Matthews, J., Strazdins, L., Zubrick, S. R., & Nicholson, J. M. (2018). What Influences Parents’ Fear about Children’s Independent Mobility? Evidence from a State-Wide Survey of Australian Parents. American Journal of Health Promotion, 32(3), 667–676. https://doi.org/10.1177/0890117117740442 Berger, U., Der, G., Mutrie, N., & Hannah, M. K. (2005). The impact of retirement on physical activity. Ageing and Society, 25(2), 181–195. https://doi.org/10.1017/S0144686X04002739 Besser, L. M., & Dannenberg, A. L. (2005). Walking to Public Transit: Steps to Help Meet Physical Activity Recommendations. American Journal of Preventive Medicine, 29(4), 273–280. https://doi.org/https://doi.org/10.1016/j.amepre.2005.06.010 Bingham, D. D., Costa, S., Hinkley, T., Shire, K. A., Clemes, S. A., & Barber, S. E. (2016). Physical Activity During the Early Years: A Systematic Review of Correlates and Determinants. American Journal of Preventive Medicine, 51(3), 384–402. https://doi.org/10.1016/j.amepre.2016.04.022 Black, S. V., Cooper, R., Martin, K. R., Brage, S., Kuh, D., & Stafford, M. (2015). Physical Activity and Mental Well-being in a Cohort Aged 60-64 Years. American Journal of Preventive Medicine, 49(2), 172–180. https://doi.org/10.1016/j.amepre.2015.03.009 Blazo, J. A., & Smith, A. L. (2018). A systematic review of siblings and physical activity experiences. International Review of Sport and Exercise Psychology, 11(1), 122–159. https://doi.org/10.1080/1750984X.2016.1229355 Boarnet, M. G., Forsyth,  a., Day, K., & Oakes, J. M. (2011). The Street Level Built Environment and   190 Physical Activity and Walking: Results of a Predictive Validity Study for the Irvine Minnesota Inventory. Environment and Behavior, 43(6), 735–775. https://doi.org/10.1177/0013916510379760 Boarnet, Marlon G., Day, K., Alfonzo, M., Forsyth, A., & Oakes, M. (2006). The Irvine-Minnesota inventory to measure built environments: Reliability tests. American Journal of Preventive Medicine, 30(2), 153–159. https://doi.org/10.1016/j.amepre.2005.09.018 Bolívar, J., Daponte, A., Rodríguez, M., & Sánchez, J. J. (2010). The influence of individual, social and physical environment factors on physical activity in the adult population in Andalusia, Spain. International Journal of Environmental Research and Public Health, 07(01), 60–77. https://doi.org/10.3390/ijerph7010060 Bolker, B. M., Brooks, M. E., Clark, C. J., Geange, S. W., Poulsen, J. R., Stevens, M. H. H., & White, J.-S. S. (2009). Generalized linear mixed models: a practical guide for ecology and evolution. Trends in Ecology & Evolution, 24(3), 127–135. https://doi.org/https://doi.org/10.1016/j.tree.2008.10.008 Bopp, M., Wilcox, S., Oberrecht, L., Kammermann, S., & McElmurray, C. T. (2004). Correlates of strength training in older rural African American and Caucasian women. Women and Health, 40(1), 1–20. https://doi.org/10.1300/J013v40n01_01 Borst, H. C., Miedema, H. M. E., de Vries, S. I., Graham, J. M. A., & van Dongen, J. E. F. (2008). Relationships between street characteristics and perceived attractiveness for walking reported by elderly people. Journal of Environmental Psychology, 28(4), 353–361. https://doi.org/10.1016/j.jenvp.2008.02.010 Bostock, L. (2001). Pathways of disadvantage? Walking as a mode of transport among low-income mothers. Health and Social Care in the Community, 9(1), 11–18. https://doi.org/10.1046/j.1365-2524.2001.00275.x Boyle, A., Barrilleaux, C., & Scheller, D. (2014). Does Walkability Influence Housing Prices? Social Science Quarterly, 95(3), 852–867. https://doi.org/10.1111/ssqu.12065 Bracy, N. L., Millstein, R. A., Carlson, J. A., Conway, T. L., Sallis, J. F., Saelens, B. E., Kerr, J., Cain, K. L., Frank, L. D., & King, A. C. (2014). Is the relationship between the built environment and physical activity moderated by perceptions of crime and safety? International Journal of Behavioral Nutrition and Physical Activity, 11(1), 24. https://doi.org/10.1186/1479-5868-11-24 Bronfenbrenner, U. (1993). Ecological models of human development. In M. Gauvain & M. Cole (Eds.), Readings on the development of children (02 ed., Vol. 3, pp. 37–43). Elsevier. Brown, B. B., Smith, K. R., Tharp, D., Werner, C. M., Tribby, C. P., Miller, H. J., & Jensen, W. (2016). A complete street intervention for walking to transit, nontransit walking, and bicycling: A quasi-experimental demonstration of increased use. Journal of Physical Activity and Health, 13(11), 1210–1219. https://doi.org/10.1123/jpah.2016-0066   191 Brown, S. C., Lombard, J., Toro, M., Huang, S., Perrino, T., Perez-Gomez, G., Plater-Zyberk, E., Pantin, H., Affuso, O., Kumar, N., Wang, K., & Szapocznik, J. (2014). Walking and proximity to the urban growth boundary and central business district. American Journal of Preventive Medicine, 47(4), 481–486. https://doi.org/10.1016/j.amepre.2014.05.008 Brownson, R. C., Hoehner, C. M., Brennan, L. K., Cook, R. A., Elliott, M. B., & McMullen, K. M. (2004). Reliability of 2 instruments for auditing the environment for physical activity. Journal of Physical Activity and Health. https://doi.org/10.1123/jpah.1.3.191 Brownson, R. C., Hoehner, C. M., Day, K., Forsyth, A., & Sallis, J. F. (2009). Measuring the Built Environment for Physical Activity: State of the Science. American Journal of Preventive Medicine, 36(4, Supplement), S99-S123.e12. https://doi.org/https://doi.org/10.1016/j.amepre.2009.01.005 Buman, M. P., Winter, S. J., Sheats, J. L., Hekler, E. B., Otten, J. J., Grieco, L. A., & King, A. C. (2013). The Stanford Healthy Neighborhood Discovery Tool. American Journal of Preventive Medicine, 44(4), e41–e47. https://doi.org/10.1016/j.amepre.2012.11.028 Burns, R. D., Colotti, T. E., Pfledderer, C. D., Fu, Y., Bai, Y., & Byun, W. (2020). Familial Factors Associating with Youth Physical Activity Using a National Sample. Children, 7(7), 79. https://doi.org/10.3390/children7070079 Butcher, K., Sallis, J. F., Mayer, J. A., & Woodruff, S. (2008). Correlates of Physical Activity Guideline Compliance for Adolescents in 100 U.S. Cities. Journal of Adolescent Health, 42(4), 360–368. https://doi.org/10.1016/j.jadohealth.2007.09.025 Cain, K. L., Geremia, C. M., Conway, T. L., Frank, L. D., Chapman, J. E., Fox, E. H., Timperio, A., Veitch, J., Van Dyck, D., Verhoeven, H., Reis, R., Augusto, A., Cerin, E., Mellecker, R. R., Queralt, A., Molina-García, J., & Sallis, J. F. (2018). Development and reliability of a streetscape observation instrument for international use: MAPS-global. International Journal of Behavioral Nutrition and Physical Activity, 15(1), 1–11. https://doi.org/10.1186/s12966-018-0650-z Cain, K. L., Millstein, R. A., Sallis, J. F., Conway, T. L., Gavand, K. A., Frank, L. D., Saelens, B. E., Geremia, C. M., Chapman, J., Adams, M. A., Glanz, K., & King, A. C. (2014). Contribution of streetscape audits to explanation of physical activity in four age groups based on the Microscale Audit of Pedestrian Streetscapes (MAPS). Social Science and Medicine, 116, 82–92. https://doi.org/10.1016/j.socscimed.2014.06.042 Cao, X., Handy, S. L., & Mokhtarian, P. L. (2006). The influences of the built environment and residential self-selection on pedestrian behavior: Evidence from Austin, TX. Transportation, 33(1), 1–20. https://doi.org/10.1007/s11116-005-7027-2 Carlson, J. A., Sallis, J. F., Conway, T. L., Saelens, B. E., Frank, L. D., Kerr, J., Cain, K. L., & King, A. C. (2012). Interactions between psychosocial and built environment factors in explaining older adults’ physical activity. Preventive Medicine, 54(1), 68–73. https://doi.org/10.1016/j.ypmed.2011.10.004 Carlson, J. A., Sallis, J. F., Kerr, J., Conway, T. L., Cain, K., Frank, L. D., & Saelens, B. E. (2014).   192 Built environment characteristics and parent active transportation are associated with active travel to school in youth age 12-15. British Journal of Sports Medicine, April 2016, 1–7. https://doi.org/10.1136/bjsports-2013-093101 Carlson, J. A., Schipperijn, J., Kerr, J., Saelens, B. E., Natarajan, L., Frank, L. D., Glanz, K., Conway, T. L., Chapman, J. E., Cain, K. L., & Sallis, J. F. (2016). Locations of Physical Activity as Assessed by GPS in Young Adolescents. Pediatrics, 137(01). https://doi.org/10.1542/peds.2015-2430 Carlson, S. A., Paul, P., Kumar, G., Watson, K. B., Atherton, E., & Fulton, J. E. (2017). Prevalence of Complete Streets policies in U.S. municipalities. Journal of Transport and Health, 05, 142–150. https://doi.org/10.1016/j.jth.2016.11.003 Carr, L. J., Dunsiger, S. I., & Marcus, B. H. (2011). Validation of Walk Score for estimating access to walkable amenities. British Journal of Sports Medicine, 45(14), 1144–1148. https://doi.org/10.1136/bjsm.2009.069609 Carver, A., Timperio, A., & Crawford, D. (2008). Playing it safe: The influence of neighbourhood safety on children’s physical activity-A review. Health and Place, 14(2), 217–227. https://doi.org/10.1016/j.healthplace.2007.06.004 Carver, A., Timperio, A., Hesketh, K., & Crawford, D. (2010). Are children and adolescents less active if parents restrict their physical activity and active transport due to perceived risk? Social Science and Medicine, 70(11), 1799–1805. https://doi.org/10.1016/j.socscimed.2010.02.010 Caspersen, C. J., Powell, K. E., & Christenson, G. M. (1985). Physical activity, exercise, and physical fitness: definitions and distinctions for health-related research. Public Health Reports (Washington, D.C. : 1974), 100(2), 126–131. https://doi.org/10.2307/20056429 Cerin, E., Leslie, E., Owen, N., & Bauman, A. (2007). Applying GIS in Physical Activity Research: Community `Walkability’ and Walking Behaviors. In P. C. Lai & A. S. H. Mak (Eds.), GIS for Health and the Environment: Development in the Asia-Pacific Region With 110 Figures (pp. 72–89). Springer Berlin Heidelberg. https://doi.org/10.1007/978-3-540-71318-0_6 Cerin, E., Saelens, B. E., Sallis, J. F., & Frank, L. D. (2006). Neighborhood environment walkability scale: Validity and development of a short form. Medicine and Science in Sports and Exercise, 38, 1682–1691. https://doi.org/10.1249/01.mss.0000227639.83607.4d Cerin, E., Sit, C. H. P., Barnett, A., Huang, W. Y. J., Gao, G. Y., Wong, S. H. S., & Sallis, J. F. (2017). Reliability of self-report measures of correlates of obesity-related behaviours in Hong Kong adolescents for the iHealt(H) and IPEN adolescent studies. Archives of Public Health, 75(1), 38. https://doi.org/10.1186/s13690-017-0209-5 Cerin, E., Sit, C. H. P., Huang, Y.-J., Barnett, A., Macfarlane, D. J., & Wong, S. S. H. (2014). Repeatability of self-report measures of physical activity, sedentary and travel behaviour in Hong Kong adolescents for the iHealt(H) and IPEN – Adolescent studies. BMC Pediatrics, 14(1), 142. https://doi.org/10.1186/1471-2431-14-142   193 Cervero, R., & Kockelman, K. (1997). Travel demand and the 3Ds: Density, diversity, and design. Transportation Research Part D: Transport and Environment, 2(3), 199–219. https://doi.org/10.1016/S1361-9209(97)00009-6 Chang, S. H., & Kim, K. (2017). A review of factors limiting physical activity among young children from low-income families. Journal of Exercise Rehabilitation, 13(4), 375–377. https://doi.org/10.12965/jer.1735060.350 Chapin, F. S. (1974). Human Activity Patterns in the City: Things People Do in Time and in Space. John Wiley & Sons Inc. Chicago Metropolitan Agency. (2013). Form-Based Codes: A Step-by-Step Guide for Communities. www.formbasedcodes.org Cho, G.-H., & Rodríguez, D. A. (2015). Location or design? Associations between neighbourhood location, built environment and walking. Urban Studies, 52(8), 1434–1453. https://doi.org/10.1177/0042098014537691 Choi, E. (2012). Walkability as an urban design problem. Understanding the activity of walking in the urban environment. Christian, H. E., Klinker, C. D., Villanueva, K., Knuiman, M. W., Foster, S. A., Zubrick, S. R., Divitini, M., Wood, L., & Giles-Corti, B. (2015). The Effect of the Social and Physical Environment on Children’s Independent Mobility to Neighborhood Destinations. Journal of Physical Activity & Health, 12(Suppl 1), S84–S93. https://doi.org/10.1123/jpah.2014-0271 Chudyk, A. M., McKay, H. A., Winters, M., Sims-Gould, J., & Ashe, M. C. (2017). Neighborhood walkability, physical activity, and walking for transportation: A cross-sectional study of older adults living on low income. BMC Geriatrics, 17(1), 1–14. https://doi.org/10.1186/s12877-017-0469-5 Chudyk, A. M., Winters, M., Moniruzzaman, M., Ashe, M. C., Gould, J. S., & McKay, H. (2015). Destinations matter: The association between where older adults live and their travel behavior. Journal of Transport and Health, 02(01), 50–57. https://doi.org/10.1016/j.jth.2014.09.008 City of Boston. (2013). Boston Complete Streets. Boston Complete Streets. https://www.boston.gov/departments/transportation/boston-complete-streets City of Edmonton. (2020). Urban Design Guidelines. The City of Edmonton. https://www.edmonton.ca/city_government/urban_planning_and_design/urban-design-guidelines.aspx City of New Westminster. (2020). Agnes Street Greenway - Installation of temporary treatment | City of New Westminster. https://www.newwestcity.ca/2020/12/14/agnes-street-greenway-installation-of-temporary-treatment.php City of Palm Desert. (2017). University neighborhood specific plan (Issue January). https://formbasedcodes.org/content/uploads/2017/05/2_Palm-Desert-University-Neighborhoods-  194 SP_300dpi_STP.pdf City of San Diego. (2020). Complete Streets | Planning Department | City of San Diego Official Website. https://www.sandiego.gov/planning/programs/transportation/complete-streets City of Seattle. (2013). Seattle Design Guidelines. http://www.seattle.gov/Documents/Departments/OPCD/Vault/CitywideDesignGuidelinesUpdate/SeattleDesignGuidelines.pdf City of Vancouver. (2020). Quebec Street Improvements | City of Vancouver. https://vancouver.ca/streets-transportation/quebec-street-improvements.aspx# Clarke, P., Ailshire, J., Melendez, R., Bader, M., & Morenoff, J. (2010). Using Google Earth to conduct a neighborhood audit: Reliability of a virtual audit instrument. Health and Place, 16(6), 1224–1229. https://doi.org/10.1016/j.healthplace.2010.08.007 Cleland, V. J., Ball, K., Salmon, J., Timperio, A. F., & Crawford, D. A. (2010). Personal, social and environmental correlates of resilience to physical inactivity among women from socio-economically disadvantaged backgrounds. Health Education Research, 25(2), 268–281. https://doi.org/10.1093/her/cyn054  Climate Data Online. (2020). Climate Data Online (CDO) - The National Climatic Data Center's (NCDC) Climate Data Online (CDO. Ncdc.noaa.gov. Retrieved 22 July 2019, from https://www.ncdc.noaa.gov/cdo-web/. Cohen, J., Cohen, P., West, S. G., & Aiken, L. S. (2013). Applied Multiple Regression/Correlation Analysis for the Behavioral Sciences. Routledge. Colbert, L. H., Hootman, J. M., & Macera, C. A. (2000). Physical Activity-Related Injuries in Walkers and Runners in the Aerobics Center Longitudinal Study. Clinical Journal of Sport Medicine, 10(4). https://journals.lww.com/cjsportsmed/Fulltext/2000/10000/Physical_Activity_Related_Injuries_in_Walkers_and.6.aspx Cousins, S. O. B. (1995). Social support for exercise among elderly women in Canada. Health Promotion International, 10(4), 273–282. Craggs, C., Corder, K., Van Sluijs, E. M. F., & Griffin, S. J. (2011). Determinants of change in physical activity in children and adolescents: A systematic review. American Journal of Preventive Medicine, 40(6), 645–658. https://doi.org/10.1016/j.amepre.2011.02.025 Curl, A., Fitt, H., & Tomintz, M. (2020). Experiences of the built environment, falls and fear of falling outdoors among older adults: An exploratory study and future directions. International Journal of Environmental Research and Public Health, 17(4). https://doi.org/10.3390/ijerph17041224 Dagkas, S., & Stathi, A. (2007). Exploring Social and Environmental Factors Affecting Adolescents’ Participation in Physical Activity. European Physical Education Review, 13(3), 369–384.   195 https://doi.org/10.1177/1356336X07081800 Davison, K. K., & Jago, R. (2009). Change in Parent and Peer Support across Ages 9 to 15 yr and Adolescent Girls’ Physical Activity. Medicine and Science in Sports and Exercise, 41(09), 1816–1825. https://doi.org/:10.1249/MSS.0b013e3181a278e2 Dawson, J. F., & Richter, A. W. (2006). Probing three-way interactions in moderated multiple regression: Development and application of a slope difference test. Journal of Applied Psychology, 91(4), 917–926. https://doi.org/10.1037/0021-9010.91.4.917 De Bourdeaudhuij, I., & Sallis, J. (2002). Relative contribution of psychosocial variables to the explanation of physical activity in three population-based adult samples. Preventive Medicine, 34(2), 279–288. https://doi.org/10.1006/pmed.2001.0979 De la Torre-Cruz, M. J., López-Serrano, S., Ruiz-Ariza, A., & Martínez-López, E. J. (2019). Perceived parental support toward physical activity positively predicts physical self-concept in young adolescents. Educational Psychology, 39(7), 941–959. https://doi.org/10.1080/01443410.2019.1620921 Dilorenzo, T. M., Stucky-Ropp, R. C., Vander Wal, J. S., & Gotham, H. J. (1998). Determinants of exercise among children. II. A longitudinal analysis. Preventive Medicine, 27(3), 470–477. https://doi.org/10.1006/pmed.1998.0307 Ding, D., Bracy, N. L., Sallis, J. F., Saelens, B. E., Norman, G. J., Harris, S. K., Durant, N., Rosenberg, D., & Kerr, J. (2012). Is fear of strangers related to physical activity among youth? American Journal of Health Promotion, 26(3), 189–195. https://doi.org/10.4278/ajhp.100701-QUAN-224 Ding, D., Sallis, J. F., Conway, T. L., Saelens, B. E., Frank, L. D., Cain, K. L., & Slymen, D. J. (2012). Interactive effects of built environment and psychosocial attributes on physical activity: A test of ecological models. Annals of Behavioral Medicine, 44(3), 365–374. https://doi.org/10.1007/s12160-012-9394-1 Ding, D., Sallis, J. F., Kerr, J., Lee, S., & Rosenberg, D. E. (2011). Neighborhood environment and physical activity among youth: A review. American Journal of Preventive Medicine, 41(4), 442–455. https://doi.org/10.1016/j.amepre.2011.06.036 Dlugonski, D., DuBose, K. D., Habeeb, C. M., & Rider, P. (2020). Physical activity coparticipation among parent–young-child dyads. Pediatric Exercise Science, 32(3), 132–139. https://doi.org/10.1123/pes.2019-0213 Dormann, C. F., Elith, J., Bacher, S., Buchmann, C., Carl, G., Carré, G., Marquéz, J. R. G., Gruber, B., Lafourcade, B., Leitão, P. J., Münkemüller, T., McClean, C., Osborne, P. E., Reineking, B., Schröder, B., Skidmore, A. K., Zurell, D., & Lautenbach, S. (2013). Collinearity: a review of methods to deal with it and a simulation study evaluating their performance. Ecography, 36(1), 27–46. https://doi.org/https://doi.org/10.1111/j.1600-0587.2012.07348.x Dörrzapf, L., Kovács-Győri, A., Resch, B., & Zeile, P. (2019). Defining and assessing walkability: a   196 concept for an integrated approach using surveys, biosensors and geospatial analysis. Urban Development Issues, 62(1), 5–15. https://doi.org/10.2478/udi-2019-0008 Drenowatz, C., Eisenmann, J. C., Pfeiffer, K. A., Welk, G., Heelan, K., Gentile, D., & Walsh, D. (2010). Influence of socio-economic status on habitual physical activity and sedentary behavior in 8- to 11-year old children. BMC Public Health, 10. https://doi.org/10.1186/1471-2458-10-214 Dubbert, P. M., King, A. C., Marcus, B. H., & Sallis, J. F. (2004). Promotion of physical activity through the life span. Handbook of Clinical Health Psychology:  Volume 2. Disorders of Behavior and Health., 2, 147–181. https://doi.org/10.1037/11589-005 Duerksen, C. J., Dale, C. G., & Elliot, D. L. (2017). The Citizen’s Guide to Planning 4th Edition. In The Citizen’s Guide to Planning 4th Edition. American Planning Association: Planners Press. https://doi.org/10.4324/9781351179669 Duncan, D. T., Aldstadt, J., Whalen, J., Melly, S. J., & Gortmaker, S. L. (2011). Validation of Walk Score® for Estimating Neighborhood Walkability: An Analysis of Four US Metropolitan Areas. In International Journal of Environmental Research and Public Health  (Vol. 8, Issue 11). https://doi.org/10.3390/ijerph8114160 Economou, A., & Theodossiou, I. (2011). Poor and sick: Estimating the relationship between household income and health. Review of Income and Wealth, 57(3), 395–411. https://doi.org/10.1111/j.1475-4991.2010.00416.x Edwards, M. J., Jago, R., Sebire, S. J., Kesten, J. M., Pool, L., & Thompson, J. L. (2015). The influence of friends and siblings on the physical activity and screen viewing behaviours of children aged 5-6 years: A qualitative analysis of parent interviews. BMJ Open, 5(5), 5–10. https://doi.org/10.1136/bmjopen-2014-006593 Edwards, N., Hooper, P., Trapp, G. S. A., Bull, F., Boruff, B., & Giles-Corti, B. (2013). Development of a Public Open Space Desktop Auditing Tool (POSDAT): A remote sensing approach. Applied Geography, 38(1), 22–30. https://doi.org/10.1016/j.apgeog.2012.11.010 Edwardson, C. L., & Gorely, T. (2010). Parental influences on different types and intensities of physical activity in youth: A systematic review. Psychology of Sport and Exercise, 11(6), 522–535. https://doi.org/10.1016/j.psychsport.2010.05.001 Ermagun, A., Hossein Rashidi, T., & Samimi, A. (2015). A joint model for mode choice and escort decisions of school trips. Transportmetrica A: Transport Science, 11(3), 270–289. https://doi.org/10.1080/23249935.2014.968654 Ewing, R. (1997). Is Los Angeles-Style Sprawl Desirable? Journal of the American Planning Association, 63(1), 107–126. https://doi.org/10.1080/01944369708975728 Ewing, R., & Cervero, R. (2001). Travel and the Built Environment: A Synthesis. Transportation Research Record, 1780(1), 87–114. https://doi.org/10.3141/1780-10 Ewing, R., & Cervero, R. (2010). Travel and the built environment: A meta-analysis. Journal of the   197 American Planning Association, 76(03), 265–294. https://doi.org/10.1080/01944361003766766 Ewing, R., Clemente, O., Handy, S. L., Brownson, R. C., & Winston, E. (2005). Identifying and Measuring Urban Design Qualities Related to Walkability. Ewing, R., Pendall, R., & Chen, D. (2003). Measuring Sprawl and Its Transportation Impacts. Transportation Research Record, 1831(1), 175–183. https://doi.org/10.3141/1831-20 Farooq, M. A., Parkinson, K. N., Adamson, A. J., Pearce, M. S., Reilly, J. K., Hughes, A. R., Janssen, X., Basterfield, L., & Reilly, J. J. (2018). Timing of the decline in physical activity in childhood and adolescence: Gateshead Millennium Cohort Study. British Journal of Sports Medicine, 52(15), 1002–1006. https://doi.org/10.1136/bjsports-2016-096933 Faulkner, G. E. J., Buliung, R. N., Flora, P. K., & Fusco, C. (2009). Active school transport, physical activity levels and body weight of children and youth: A systematic review. Preventive Medicine, 48(1), 3–8. https://doi.org/10.1016/j.ypmed.2008.10.017 Faulkner, G. E. J., Richichi, V., Buliung, R. N., Fusco, C., & Moola, F. (2010). What’s quickest and easiest Parental decision making about school trip mode. International Journal of Behavioral Nutrition and Physical Activity, 7, 1–11. https://doi.org/10.1186/1479-5868-7-62 Ferreira, M. T., Matsudo, S. M. M., Ribeiro, M. C. S. A., & Ramos, L. R. (2010). Health-related factors correlate with behavior trends in physical activity level in old age: Longitudinal results from a population in São Paulo, Brazil. BMC Public Health, 10. https://doi.org/10.1186/1471-2458-10-690 Finkelstein, D. M., Petersen, D. M., & Schottenfeld, L. S. (2017). Promoting children’s physical activity in low-incomecommunities in Colorado: What are the barriers and opportunities? Preventing Chronic Disease, 14(12), 1–10. https://doi.org/10.5888/pcd14.170111 Fishbein, M., & Ajzen, I. (1975). Belief, attitude, intention, and behavior : an introduction to theory and research. Addison-Wesley Pub. Co. Flay, B. R., & Petraitis, J. (1994). The Theory of Triadic Influence: A New Theory of Health Behavior With Implications for Preventive Interventions. Advances in Medical Sociology, 4, 19–44. Foerster, J. F. (1979). Mode choice decision process models: A comparison of compensatory and non-compensatory structures. Transportation Research Part A: General, 13(1), 17–28. https://doi.org/10.1016/0191-2607(79)90083-9 Foltête, J. C., & Piombini, A. (2007). Urban layout, landscape features and pedestrian usage. Landscape and Urban Planning, 81(3), 225–234. https://doi.org/10.1016/j.landurbplan.2006.12.001 Form-Based Codes Institute. (2020). Form-Based Codes Defined - Form-Based Codes Institute at Smart Growth America : Form-Based Codes Institute at Smart Growth America. https://formbasedcodes.org/definition/   198 Forsyth, A. (2015). What is a walkable place? The walkability debate in urban design. Urban Design International, 20(04), 274–292. https://doi.org/10.1057/udi.2015.22 Forsyth, A., & Oakes, J. M. (2014). Workplace Neighborhoods, Walking, Physical Activity, Weight Status, and Perceived Health: Assessing the Built Environment. Transportation Research Record, 2452(1), 98–104. https://doi.org/10.3141/2452-12 Foster, S., Villanueva, K., Wood, L., Christian, H., & Giles-Corti, B. (2014). The impact of parents’ fear of strangers and perceptions of informal social control on children’s independent mobility. Health and Place, 26, 60–68. https://doi.org/10.1016/j.healthplace.2013.11.006 Fotheringham, A. S., & Wong, D. W. S. (1991). The Modifiable Areal Unit Problem in Multivariate Statistical Analysis. Environment and Planning A: Economy and Space, 23(7), 1025–1044. https://doi.org/10.1068/a231025 Francis, J., Martin, K., Wood, L., & Foster, S. (2017). ‘I’ll be driving you to school for the rest of your life’: A qualitative study of parents’ fear of stranger danger. Journal of Environmental Psychology, 53, 112–120. https://doi.org/10.1016/j.jenvp.2017.07.004 Frank, L. D., Bradley, M., Kavage, S., Chapman, J., & Lawton, T. K. (2008). Urban form, travel time, and cost relationships with tour complexity and mode choice. Transportation, 35(1), 37–54. https://doi.org/10.1007/s11116-007-9136-6 Frank, L. D., Greenwald, M. J., Winkelman, S., Chapman, J., & Kavage, S. (2010). Carbonless footprints: Promoting health and climate stabilization through active transportation. Preventive Medicine, 50(SUPPL.), S99–S105. https://doi.org/10.1016/j.ypmed.2009.09.025 Frank, L. D., Hong, A., & Ngo, V. D. (2019). Causal evaluation of urban greenway retrofit: A longitudinal study on physical activity and sedentary behavior. Preventive Medicine, 123(January), 109–116. https://doi.org/10.1016/j.ypmed.2019.01.011 Frank, L. D., Kerr, J., Chapman, J., & Sallis, J. F. (2007). Urban form relationships with walk trip frequency and distance among youth. American Journal of Health Promotion, 21(04-Suppl), 305–311. https://doi.org/10.4278/0890-1171-21.4s.305 Frank, L. D., & Pivo, G. (1994). Impacts of mixed use and density on utilization of three modes of travel: Single-occupant vehicle, transit, and walking. Transportation Research Record, 1466, 44–52. Frank, L. D., Saelens, B. E., Chapman, J., Sallis, J. F., Kerr, J., Glanz, K., Couch, S. C., Learnihan, V., Zhou, C., Colburn, T., & Cain, K. L. (2012). Objective assessment of obesogenic environments in youth: Geographic information system methods and spatial findings from the neighborhood impact on kids study. American Journal of Preventive Medicine, 42(5), e47–e55. https://doi.org/10.1016/j.amepre.2012.02.006 Frank, L. D., Sallis, J. F., Saelens, B. E., Leary, L., Cain, L., Conway, T. L., & Hess, P. M. (2010). The development of a walkability index: Application to the neighborhood quality of life study. In British Journal of Sports Medicine. https://doi.org/10.1136/bjsm.2009.058701   199 Frank, L. D., Schmid, T. L., Sallis, J. F., Chapman, J., & Saelens, B. E. (2005). Linking objectively measured physical activity with objectively measured urban form: Findings from SMARTRAQ. American Journal of Preventive Medicine, 28(2 SUPPL. 2), 117–125. https://doi.org/10.1016/j.amepre.2004.11.001 Frazer, A., Voss, C., Winters, M., Naylor, P.-J., Higgins, J. W., & McKay, H. (2015). Differences in adolescents’ physical activity from school-travel between urban and suburban neighbourhoods in Metro Vancouver, Canada. Preventive Medicine Reports, 2, 170–173. https://doi.org/https://doi.org/10.1016/j.pmedr.2015.02.008 Freedson, P. S., Melanson, E., & Sirard, J. (1998). Calibration of the Computer Science and Applications, Inc. accelerometer. Medicine & Science in Sports & Exercise, 30(5). https://journals.lww.com/acsm-msse/Fulltext/1998/05000/Calibration_of_the_Computer_Science_and.21.aspx Freeland, A. L., Banerjee, S. N., Dannenberg, A. L., & Wendel, A. M. (2013). Walking Associated With Public Transit: Moving Toward Increased Physical Activity in the United States. American Journal of Public Health, 103(3), 536–542. https://doi.org/10.2105/AJPH.2012.300912 Freeman, L., Neckerman, K., Schwartz-Soicher, O., Quinn, J., Richards, C., Bader, M. D. M., Lovasi, G., Jack, D., Weiss, C., Konty, K., Arno, P., Viola, D., Kerker, B., & Rundle, A. G. (2013). Neighborhood walkability and active travel (walking and cycling) in New York City. Journal of Urban Health, 90(4), 575–585. https://doi.org/10.1007/s11524-012-9758-7 Gallagher, N. A., Clarke, P. J., Loveland-Cherry, C., Ronis, D. L., & Gretebeck, K. A. (2015). Self-efficacy, neighborhood walking, and fall history in older adults. Journal of Aging and Physical Activity, 23(1), 64–71. https://doi.org/10.1123/JAPA.2012-0287 Gallagher, N. A., Clarke, P. J., Ronis, D. L., Loveland-Cherry, C., Nyquist, L., & Gretebeck, K. A. (2012). Influences on Neighborhood Walking in Older Adults. Research in Gerontological Nursing, 5(4), 238–250. https://doi.org/10.3928/19404921-20120906-05.Influences Gallimore, J. M., Brown, B. B., & Werner, C. M. (2011). Walking routes to school in new urban and suburban neighborhoods : An environmental walkability analysis of blocks and routes. Journal of Environmental Psychology, 31(2), 184–191. https://doi.org/10.1016/j.jenvp.2011.01.001 Galster, G., Hanson, R., Ratcliffe, M. R., Wolman, H., Coleman, S., & Freihage, J. (2001). Wrestling Sprawl to the Ground: Defining and measuring an elusive concept. Housing Policy Debate, 12(4), 681–717. https://doi.org/10.1080/10511482.2001.9521426 García-Hermoso, A., Saavedra, J. M., Ramírez-Vélez, R., Ekelund, U., & del Pozo-Cruz, B. (2017). Reallocating sedentary time to moderate-to-vigorous physical activity but not to light-intensity physical activity is effective to reduce adiposity among youths: a systematic review and meta-analysis. Obesity Reviews, 18(9), 1088–1095. https://doi.org/https://doi.org/10.1111/obr.12552 Garcia, A. W., & King, A. C. (1991). Predicting Long-term Adherence to Aerobic Exercise: A Comparison of Two Models. Journal of Sport and Exercise Psychology, 13(4), 394–410. https://doi.org/10.1123/jsep.13.4.394   200 Garden, F. L., & Jalaludin, B. B. (2009). Impact of urban sprawl on overweight, obesity, and physical activity in Sydney, Australia. Journal of Urban Health : Bulletin of the New York Academy of Medicine, 86(1), 19–30. https://doi.org/10.1007/s11524-008-9332-5 Ghani, F., Rachele, J. N., Washington, S., & Turrell, G. (2016). Gender and age differences in walking for transport and recreation: Are the relationships the same in all neighborhoods? Preventive Medicine Reports, 4, 75–80. https://doi.org/https://doi.org/10.1016/j.pmedr.2016.05.001 Ghekiere, A., Deforche, B., Carver, A., Mertens, L., de Geus, B., Clarys, P., Cardon, G., De Bourdeaudhuij, I., & Van Cauwenberg, J. (2017). Insights into children’s independent mobility for transportation cycling—Which socio-ecological factors matter? Journal of Science and Medicine in Sport, 20(3), 267–272. https://doi.org/10.1016/j.jsams.2016.08.002 Gielen, A. C., DeFrancesco, S., Bishai, D., Mahoney, P., Ho, S., & Guyer, B. (2004). Child pedestrians: The role of parental beliefs and practices in promoting safe walking in urban neighborhoods. Journal of Urban Health, 81(4), 545–555. https://doi.org/10.1093/jurban/jth139 Gilderbloom, J. I., Riggs, W. W., & Meares, W. L. (2015). Does walkability matter? An examination of walkability’s impact on housing values, foreclosures and crime. Cities, 42(PA), 13–24. https://doi.org/10.1016/j.cities.2014.08.001 Giles-Corti, B., & Donovan, R. J. (2002). The relative influence of individual, social and physical environment determinants of physical activity. Social Science and Medicine, 54(12), 1793–1812. https://doi.org/10.1016/S0277-9536(01)00150-2 Giles-Corti, B., Kelty, S. F., Zubrick, S. R., & Villanueva, K. P. (2009). Encouraging walking for transport and physical activity in children and adolescents: How important is the built environment? Sports Medicine, 39(12), 995–1009. https://doi.org/10.2165/11319620-000000000-00000 Giles-corti, B., Timperio, A., Bull, F., & Pikora, T. (2005). Correlates : Increased Specificity for Ecological. Exercise and Sport Sciences Reviews, 33(04), 175–181. Gill, M., Chan-Golston, A. M., Rice, L. N., Roth, S. E., Crespi, C. M., Cole, B. L., Koniak-Griffin, D., & Prelip, M. L. (2018). Correlates of Social Support and its Association With Physical Activity Among Young Adolescents. Health Education and Behavior, 45(2), 207–216. https://doi.org/10.1177/1090198117714826 Golan, Y., Wilkinson, N., Henderson, J. M., & Weverka, A. (2019). Gendered walkability: Building a daytime walkability index for women. Journal of Transport and Land Use, 12(1), 501–526. https://doi.org/10.5198/jtlu.2019.1472 Gómez, J. E., Johnson, B. A., Selva, M., & Sallis, J. F. (2004). Violent crime and outdoor physical activity among inner-city youth. Preventive Medicine, 39(5), 876–881. https://doi.org/10.1016/j.ypmed.2004.03.019 Gordon-Larsen, P., McMurray, R. G., & Popkin, B. M. (2000). Determinants of adolescent physical   201 activity and inactivity patterns. Pediatrics, 89(01), 19–26. https://doi.org/10.1542/peds.105.6.e83 Gordon-Larsen, Penny, Hou, N., Sidney, S., Sternfeld, B., Lewis, C. E., Jacobs David R, J., & Popkin, B. M. (2008). Fifteen-year longitudinal trends in walking patterns and their impact on weight change. The American Journal of Clinical Nutrition, 89(1), 19–26. https://doi.org/10.3945/ajcn.2008.26147 Grant, T. L., Edwards, N., Sveistrup, H., Andrew, C., & Egan, M. (2010). Inequitable walking conditions among older people: Examining the interrelationship of neighbourhood socio-economic status and urban form using a comparative case study. BMC Public Health, 10(1), 677. https://doi.org/10.1186/1471-2458-10-677 Greco, J. (2012). From pop-up to permanent. Planning, 78(9), 14–18. Gregg, K., & Hess, P. (2019). Complete streets at the municipal level: A review of American municipal Complete Street Policy. International Journal of Sustainable Transportation, 13(06), 407–418. https://doi.org/10.1080/15568318.2018.1476995 Grow, H. M., Saelens, B. E., Kerr, J., Durant, N. H., Norman, G. J., & Sallis, J. F. (2008). Where Are Youth Active? Roles of Proximity, Active Transport, and Built Environment. Medicine & Science in Sports & Exercise, 40(12), 2071–2079. https://journals.lww.com/acsm-msse/Fulltext/2008/12000/Where_Are_Youth_Active__Roles_of_Proximity,_Active.9.aspx  GTFS Data Exchange. Gtfs-data-exchange.com. (2016). Retrieved 11 November 2019, from http://www.gtfs-data-exchange.com. Gu, P., Han, Z., Cao, Z., Chen, Y., & Jiang, Y. (2018). Using Open Source Data to Measure Street Walkability and Bikeability in China: A Case of Four Cities. Transportation Research Record: Journal of the Transportation Research Board, 2672(31), 63–75. https://doi.org/10.1177/0361198118758652 Gullon, P., Bilal, U., Hirsch, J. A., Rundle, A. G., Judd, S., Safford, M. M., & Lovasi, G. S. (2020). Does a physical activity supportive environment ameliorate or exacerbate socioeconomic inequities in incident coronary heart disease? Journal of Epidemiology and Community Health, 1–6. https://doi.org/10.1136/jech-2020-215239 Gunn, L. D., Lee, Y., Geelhoed, E., Shiell, A., & Giles-Corti, B. (2014). The cost-effectiveness of installing sidewalks to increase levels of transport-walking and health. Preventive Medicine, 67, 322–329. https://doi.org/10.1016/j.ypmed.2014.07.041 Guo, Z., & Loo, B. P. Y. (2013). Pedestrian environment and route choice: Evidence from New York City and Hong Kong. Journal of Transport Geography, 28, 124–136. https://doi.org/10.1016/j.jtrangeo.2012.11.013 Gustafson, S. L., & Rhodes, R. E. (2006). Parental correlates of physical activity in children and early adolescents. Sports Medicine, 36(1), 79–97. https://doi.org/10.2165/00007256-200636010-00006 Hägerstrand, T. (1970). What about people in Regional Science? In Papers of the Regional Science   202 Association (Vol. 24, Issue 1, pp. 6–21). https://doi.org/10.1007/BF01936872 Halonen, J. I., Kivimäki, M., Pentti, J., Kawachi, I., Virtanen, M., Martikainen, P., Subramanian, S. V., & Vahtera, J. (2012). Quantifying Neighbourhood Socioeconomic Effects in Clustering of Behaviour-Related Risk Factors: A Multilevel Analysis. PLoS ONE, 7(3), e32937. https://doi.org/10.1371/journal.pone.0032937 Hamidi, S., Ewing, R., Preuss, I., & Dodds, A. (2015). Measuring Sprawl and Its Impacts: An Update. Journal of Planning Education and Research, 35(1), 35–50. https://doi.org/10.1177/0739456X14565247 Handy, S. L. (1992). Regional versus local acessibility: Implications for nonwork travel. Transportation Research Record, 18(04), 253–267. Handy, S. L., Boarnet, M. G., Ewing, R., & Killingsworth, R. E. (2002). How the built environment affects physical activity: Views from urban planning. American Journal of Preventive Medicine. https://doi.org/10.1016/S0749-3797(02)00475-0 Handy, S. L., Cao, X., & Mokhtarian, P. (2008). Neighborhood Design and Children’s Outdoor Play: Evidence from Northern California. Children, Youth and Environments, 18(02), 160–179. https://www.jstor.org/stable/10.7721/chilyoutenvi.18.2.0160#metadata_info_tab_contents Handy, S. L., & Niemeier, D. A. (1997). Measuring Accessibility: An Exploration of Issues and Alternatives. Environment and Planning A: Economy and Space, 29(7), 1175–1194. https://doi.org/10.1068/a291175 Hanley, J. A., & McNeil, B. J. (1982). The meaning and use of the area under a receiver operating characteristic (ROC) curve. Radiology, 143(1), 29–36. https://doi.org/10.1148/radiology.143.1.7063747 Harris, T. J., Owen, C. G., Victor, C. R., Adams, R., & Cook, D. G. (2009). What factors are associated with physical activity in older people, assessed objectively by accelerometry? British Journal of Sports Medicine, 43(6), 442–450. https://doi.org/10.1136/bjsm.2008.048033 Hart, J. (2009). The Health Benefits of Walking. Alternative and Complementary Therapies, 15(1), 7–10. https://doi.org/10.1089/act.2009.15105 Heath, G. W., Brownson, R. C., Kruger, J., Miles, R., Powell, K. E., Ramsey, L. T., & __. (2006). The Effectiveness of Urban Design and Land Use and Transport Policies and Practices to Increase Physical Activity: A Systematic Review. Journal of Physical Activity and Health, 03(S01), S55–S76. https://doi.org/10.1123/jpah.3.s1.s55 Heath, G. W., Parra, D. C., Sarmiento, O. L., Andersen, L. B., Owen, N., Goenka, S., Montes, F., Brownson, R. C., Alkandari, J. R., Bauman, A. E., Blair, S. N., Bull, F. C., Craig, C. L., Ekelund, U., Guthold, R., Hallal, P. C., Haskell, W. L., Inoue, S., Kahlmeier, S., … Wells, J. C. (2012). Evidence-based intervention in physical activity: Lessons from around the world. The Lancet, 380(9838), 272–281. https://doi.org/10.1016/S0140-6736(12)60816-2   203 Hillman, M., Adams, J., & Whitelegg, J. (1990). One False Move. In One False Move … A Study of Children’s Independent Mobility. PSI Publishing. Hobbs, M., Griffiths, C., Green, M. A., Jordan, H., Saunders, J., & McKenna, J. (2018). Neighbourhood typologies and associations with body mass index and obesity: A cross-sectional study. Preventive Medicine, 111, 351–357. https://doi.org/https://doi.org/10.1016/j.ypmed.2017.11.024 Huang, R., Moudon, A. V, Zhou, C., Stewart, O. T., & Saelens, B. E. (2017). Light Rail Leads to More Walking Around Station Areas. Journal of Transport & Health, 6, 201–208. https://doi.org/10.1016/j.jth.2017.02.002 Ikeda, E., Hinckson, E., Witten, K., & Smith, M. (2019). Assessment of direct and indirect associations between children active school travel and environmental, household and child factors using structural equation modelling. International Journal of Behavioral Nutrition and Physical Activity, 16(1), 1–17. https://doi.org/10.1186/s12966-019-0794-5 Jacobs, J. (1961). The Death and Life of Great American Cities. Random House. Jago, R., Thompson, J. L., Page, A. S., Brockman, R., Cartwright, K., & Fox, K. R. (2009). Licence to be active: Parental concerns and 10-11-year-old children’s ability to be independently physically active. Journal of Public Health, 31(4), 472–477. https://doi.org/10.1093/pubmed/fdp053 James, P., Troped, P. J., Hart, J. E., Joshu, C. E., Colditz, G. A., Brownson, R. C., Ewing, R., & Laden, F. (2013). Urban sprawl, physical activity, and body mass index: Nurses’ health study and nurses’ health study II. American Journal of Public Health, 103(2), 369–375. https://doi.org/10.2105/AJPH.2011.300449 Jelleyman, C., McPhee, J., Brussoni, M., Bundy, A., & Duncan, S. (2019). A cross-sectional description of parental perceptions and practices related to risky play and independent mobility in children: The New Zealand state of play survey. International Journal of Environmental Research and Public Health, 16(2), 1–19. https://doi.org/10.3390/ijerph16020262 Jensen, W. A., Stump, T. K., Brown, B. B., Werner, C. M., & Smith, K. R. (2017). Walkability, complete streets, and gender: Who benefits most? Helath and Place, 48, 80–89. https://doi.org/10.1016/j.physbeh.2017.03.040 Joe, L., Carlson, J. A., & Sallis, J. F. (2012a). Active Where? Individual Item Reliability Statistics Parent/Adolescent Survey. http://www.drjamessallis.sdsu.edu/Documents/AW_item_reliability_ParentAdo.pdf Joe, L., Carlson, J. A., & Sallis, J. F. (2012b). Active Where? Individual Item Reliability Statistics Parent/Child Survey. http://www.drjamessallis.sdsu.edu/Documents/AW_item_reliability_ParentChild.pdf Jones, A. P., Coombes, E. G., Griffin, S. J., & van Sluijs, E. M. F. (2009). Environmental supportiveness for physical activity in English schoolchildren: a study using Global Positioning Systems. International Journal of Behavioral Nutrition and Physical Activity, 6(1), 42.   204 https://doi.org/10.1186/1479-5868-6-42 Jou, R.-C., Kitamura, R., Weng, M.-C., & Chen, C.-C. (2008). Dynamic commuter departure time choice under uncertainty. Transportation Research Part A: Policy and Practice, 42(5), 774–783. https://doi.org/https://doi.org/10.1016/j.tra.2008.01.017 Kamargianni, M., & Polydoropoulou, A. (2013). Hybrid choice model to investigate effects of teenagers’ attitudes toward walking and cycling on mode choice behavior. Transportation Research Record, 2382, 151–161. https://doi.org/10.3141/2382-17 Kamphuis, C. B. M. M., Mackenbach, J. P., Giskes, K., Huisman, M., Brug, J., & van Lenthe, F. J. (2010). Why do poor people perceive poor neighbourhoods? The role of objective neighbourhood features and psychosocial factors. Health and Place, 16(4), 744–754. https://doi.org/10.1016/j.healthplace.2010.03.006 Kärmeniemi, M., Lankila, T., Ikäheimo, T., Koivumaa-Honkanen, H., & Korpelainen, R. (2018). The Built Environment as a Determinant of Physical Activity: A Systematic Review of Longitudinal Studies and Natural Experiments. Annals of Behavioral Medicine, February, 239–251. https://doi.org/10.1093/abm/kax043 Katzmarzyk, P. T., Denstel, K. D., Kim Beals, Jordan Carlson, S. E. C., McKenzie, T. L., Pate, R. R., Sisson, S. B., Staiano, A. E., Stanish, H., Ward, D. S., Whitt-Glover, M., & Wright, C. (2018). Results from the United States 2018 report card on physical activity for children and youth. Journal of Physical Activity and Health, 15(Suppl 2), S422–S424. https://doi.org/https://doi.org/10.1123/jpah.2018-0476 Kelly, C. M., Wilson, J. S., Baker, E. A., Miller, D. K., & Schootman, M. (2013). Using Google Street View to audit the built environment: Inter-rater reliability results. Annals of Behavioral Medicine, 45(SUPPL.1), 108–112. https://doi.org/10.1007/s12160-012-9419-9 Kerr, J., Rosenberg, D., & Frank, L. (2012). The Role of the Built Environment in Healthy Aging: Community Design, Physical Activity, and Health among Older Adults. Journal of Planning Literature, 27(1), 43–60. https://doi.org/10.1177/0885412211415283 Kim, H., & Yang, S. (2017). Neighborhood Walking and Social Capital: The Correlation between Walking Experience and Individual Perception of Social Capital. Sustainability, 9(5). https://doi.org/10.3390/su9050680 King, A. C., Sallis, J. F., Frank, L. D., Saelens, B. E., Cain, K., Conway, T. L., Chapman, J. E., Ahn, D. K., & Kerr, J. (2011). Aging in neighborhoods differing in walkability and income: Associations with physical activity and obesity in older adults. Social Science and Medicine, 73(10), 1525–1533. https://doi.org/10.1016/j.socscimed.2011.08.032 King, K. A., Tergerson, J. L., & Wilson, B. R. (2008). Effect of social support on adolescents’ perceptions of and engagement in physical activity. Journal of Physical Activity and Health, 5(3), 374–384. https://doi.org/10.1123/jpah.5.3.374 Kneeshaw-Price, S. H., Saelens, B. E., Sallis, J. F., Frank, L. D., Grembowski, D. E., Hannon, P. A.,   205 Smith, N. L., & Chan, K. C. G. (2015). Neighborhood Crime-Related Safety and Its Relation to Children’s Physical Activity. Journal of Urban Health, 92(3), 472–489. https://doi.org/10.1007/s11524-015-9949-0 Kneeshaw-Price, S., Saelens, B. E., Sallis, J. F., Glanz, K., Frank, L. D., Kerr, J., Hannon, P. a, Grembowski, D. E., Chan, K. C. G., & Cain, K. L. (2013). Children’s objective physical activity by location: Why the neighborhood matters. Pediatric Exercise Science, 25(3), 468–486. https://doi.org/2012-0127 [pii] Kockelman, K. (1997). Travel Behavior as Function of Accessibility, Land Use Mixing, and Land Use Balance: Evidence from San Francisco Bay Area. Transportation Research Record: Journal of the Transportation Research Board, 1607(970048), 116–125. https://doi.org/10.3141/1607-16 Kolbe-Alexander, T. L., Pacheco, K., Tomaz, S. A., Karpul, D., & Lambert, E. V. (2015). The relationship between the built environment and habitual levels of physical activity in South African older adults: a pilot study. BMC Public Health, 15(1), 518. https://doi.org/10.1186/s12889-015-1853-8 Koohsari, M. J., McCormack, G. R., Nakaya, T., Shibata, A., Ishii, K., Yasunaga, A., Liao, Y., & Oka, K. (2020). Walking-friendly built environments and objectively measured physical function in older adults. Journal of Sport and Health Science, 9(6), 651–656. https://doi.org/10.1016/j.jshs.2020.02.002 Koohsari, M. J., Sugiyama, T., Lamb, K. E., Villanueva, K., & Owen, N. (2014). Street connectivity and walking for transport: Role of neighborhood destinations. Preventive Medicine, 66, 118–122. https://doi.org/10.1016/j.ypmed.2014.06.019 Korpela, K., Kyttä, M., & Hartig, T. (2002). RESTORATIVE EXPERIENCE, SELF-REGULATION, AND CHILDREN’s PLACE PREFERENCES. Journal of Environmental Psychology, 22(4), 387–398. https://doi.org/https://doi.org/10.1006/jevp.2002.0277 Kumar, S. (2002). Canadian urban design practice: A review of urban design regulations. Canadian Journal of Urban Research, 11(02), 239–263. https://www.jstor.org/stable/44320719 Kutner, M. H., Nachtsheim, C. J., Neter, J., & Li, W. (2005). Applied Linear Statistical Models (5th Editio). McGraw-Hillf. Kuzmyak, J. R., Baber, C., & Savory, D. (2006). Use of Walk Opportunities Index to Quantify Local Accessibility. Transportation Research Record: Journal of the Transportation Research Board, 1977(1977), 145–153. https://doi.org/10.3141/1977-19 Kwan, M. Y., Cairney, J., Faulkner, G. E., & Pullenayegum, E. E. (2012). Physical activity and other health-risk behaviors during the transition into early adulthood: A longitudinal cohort study. American Journal of Preventive Medicine, 42(1), 14–20. https://doi.org/10.1016/j.amepre.2011.08.026 Kyttä, M., Hirvonen, J., Rudner, J., Pirjola, I., & Laatikainen, T. (2015). The last free-range children? Children’s independent mobility in Finland in the 1990s and 2010s. Journal of Transport   206 Geography, 47, 1–12. https://doi.org/10.1016/j.jtrangeo.2015.07.004 Lachapelle, U., Frank, L. D., Saelens, B. E., Sallis, J. F., & Conway, T. L. (2011). Commuting by public transit and physical activity: where you live, where you work, and how you get there. Journal of Physical Activity & Health, 8 Suppl 1(Suppl 1), S72–S82. Laird, Y., Fawkner, S., & Niven, A. (2018). A grounded theory of how social support influences physical activity in adolescent girls. International Journal of Qualitative Studies on Health and Well-Being, 13(1). https://doi.org/10.1080/17482631.2018.1435099 Lambert, A., Vlaar, J., Herrington, S., & Brussoni, M. (2019). What is the relationship between the neighbourhood built environment and time spent in outdoor play? A systematic review. International Journal of Environmental Research and Public Health, 16(20), 3840. https://doi.org/10.3390/ijerph16203840 Lämmle, L., Worth, A., & Bös, K. (2012). Socio-demographic correlates of physical activity and physical fitness in German children and adolescents. European Journal of Public Health, 22(6), 880–884. https://doi.org/10.1093/eurpub/ckr191 Lampinen, E. K., Eloranta, A. M., Haapala, E. A., Lindi, V., Väistö, J., Lintu, N., Karjalainen, P., Kukkonen-Harjula, K., Laaksonen, D., & Lakka, T. A. (2017). Physical activity, sedentary behaviour, and socioeconomic status among Finnish girls and boys aged 6–8 years. European Journal of Sport Science, 17(4), 462–472. https://doi.org/10.1080/17461391.2017.1294619 Langer, S. L., Crain, A. L., Senso, M. M., Levy, R. L., & Sherwood, N. E. (2014). Predicting child physical activity and screen time: Parental support for physical activity and general parenting styles. Journal of Pediatric Psychology, 39(6), 633–642. https://doi.org/10.1093/jpepsy/jsu021 Lanza, S. T., & Rhoades, B. L. (2013). Latent class analysis: an alternative perspective on subgroup analysis in prevention and treatment. Prevention Science : The Official Journal of the Society for Prevention Research, 14(2), 157–168. https://doi.org/10.1007/s11121-011-0201-1 Lau, E. Y., Faulkner, G., Qian, W., & Leatherdale, S. T. (2016). Longitudinal associations of parental and peer influences with physical activity during adolescence: findings from the COMPASS study. Health Promotion and Chronic Disease Prevention in Canada, 36(11), 235–242. https://doi.org/10.24095/hpcdp.36.11.01 Lawman, H. G., & Wilson, D. K. (2014). Associations of social and environmental supports with sedentary behavior, light and moderate-to-vigorous physical activity in obese underserved adolescents. The International Journal of Behavioral Nutrition and Physical Activity, 11, 92. https://doi.org/10.1186/s12966-014-0092-1 Lee, R. E., Cubbin, C., & Winkleby, M. (2007). Contribution of neighbourhood socioeconomic status and physical activity resources to physical activity among women. Journal of Epidemiology and Community Health. https://doi.org/10.1136/jech.2006.054098 Lee, R. E., Soltero, E. G., Jáuregui, A., Mama, S. K., Barquera, S., Jauregui, E., Lopez y Taylor, J., Ortiz-Hernández, L., & Lévesque, L. (2016). Disentangling Associations of Neighborhood Street   207 Scale Elements With Physical Activity in Mexican School Children. Environment and Behavior, 48(01), 150–171. https://doi.org/10.1177/0013916515615389 Lee, S., Lee, C., Nam, J. W., Abbey-Lambertz, M., & Mendoza, J. A. (2020). School walkability index: Application of environmental audit tool and GIS. Journal of Transport and Health, 18(July 2019), 100880. https://doi.org/10.1016/j.jth.2020.100880 Lee, Y. S. (2005). Gender differences in physical activity and walking among older adults. Journal of Women and Aging, 17(1–2), 55–70. https://doi.org/10.1300/J074v17n01_05 Li, F., Fisher, K. J., Brownson, R. C., & Bosworth, M. (2005). Multilevel modelling of built environment characteristics related to neighbourhood walking activity in older adults. Journal of Epidemiology & Community Health, 59(7), 558–564. https://doi.org/10.1136/jech.2004.028399 Liao, B., van den Berg, P. E. W., van Wesemael, P. J. V., & Arentze, T. A. (2020). How does walkability change behavior? A comparison between different age groups in the Netherlands. International Journal of Environmental Research and Public Health, 17(2), 9–11. https://doi.org/10.3390/ijerph17020540 Lin, E. Y., Witten, K., Smith, M., Carroll, P., Asiasiga, L., Badland, H., & Parker, K. (2017). Social and built-environment factors related to children’s independent mobility: The importance of neighbourhood cohesion and connectedness. Health and Place, 46(November 2016), 107–113. https://doi.org/10.1016/j.healthplace.2017.05.002 Lindsay Smith, G., Banting, L., Eime, R., O’Sullivan, G., & van Uffelen, J. G. Z. (2017). The association between social support and physical activity in older adults: A systematic review. International Journal of Behavioral Nutrition and Physical Activity, 14(1), 1–21. https://doi.org/10.1186/s12966-017-0509-8 Littke, H. (2016). Revisiting the San Francisco parklets problematizing publicness, parks, and transferability. Urban Forestry & Urban Greening, 15, 165–173. https://doi.org/https://doi.org/10.1016/j.ufug.2015.12.010 Liu, C., Bardaka, E., Palakurthy, R., & Tung, L. W. (2020). Analysis of travel characteristics and access mode choice of elderly urban rail riders in Denver, Colorado. Travel Behaviour and Society, 19(October 2019), 194–206. https://doi.org/10.1016/j.tbs.2019.11.004 Lloyd, A. B., Lubans, D. R., Plotnikoff, R. C., Collins, C. E., & Morgan, P. J. (2014). Maternal and paternal parenting practices and their influence on children’s adiposity, screen-time, diet and physical activity. Appetite, 79, 149–157. https://doi.org/10.1016/j.appet.2014.04.010 Loprinzi, P. D. (2013). Objectively measured light and moderate-to-vigorous physical activity is associated with lower depression levels among older US adults. Aging & Mental Health, 17(7), 801–805. https://doi.org/10.1080/13607863.2013.801066 Loprinzi, P. D., & Trost, S. G. (2010). Parental influences on physical activity behavior in preschool children. Preventive Medicine, 50(3), 129–133. https://doi.org/10.1016/j.ypmed.2009.11.010   208 Lovasi, G. S., Neckerman, K. M., Quinn, J. W., Weiss, C. C., & Rundle, A. (2009). Effect of individual or neighborhood disadvantage on the association between neighborhood walkability and body mass index. American Journal of Public Health, 99(2), 279–284. https://doi.org/10.2105/AJPH.2008.138230 Luo, L., Zhou, S., Cai, W., Yoke, M., Low, H., Tian, F., Wang, Y., Xiao, X., & Chen, D. (2008). Agent-based human behavior modeling. Computer Animation And Virtual Worlds, August, 271–281. https://doi.org/10.1002/cav Luo, M., Li, H., Pan, X., Fei, T., Dai, S., Qiu, G., Zou, Y., Vos, H., Luo, J., & Jia, P. (2020). Neighbourhood speed limit and childhood obesity. Obesity Reviews, May, 1–11. https://doi.org/10.1111/obr.13052 Lydon, M., & Garcia, A. (2015). A Tactical Urbanism How-To. In Tactical Urbanism (pp. 171–208). https://doi.org/10.5822/978-1-61091-567-0_5 Lynch, K. (1960). The image of the city. The Technology Press and Harvard University Press. Määttä, S., Ray, C., Vepsäläinen, H., Lehto, E., Kaukonen, R., Ylönen, A., & Roos, E. (2018). Parental education and pre-School children’s objectively measured sedentary time: The role of co-participation in physical activity. International Journal of Environmental Research and Public Health, 15(2), 1–14. https://doi.org/10.3390/ijerph15020366 MacDonald, J. M., Stokes, R. J., Cohen, D. A., Kofner, A., & Ridgeway, G. K. (2010). The Effect of Light Rail Transit on Body Mass Index and Physical Activity. American Journal of Preventive Medicine, 39(2), 105–112. https://doi.org/10.1016/j.amepre.2010.03.016 Maghelal, P. K., & Capp, C. J. (2011). Walkability: A review of existing pedestrian indices. URISA Journal, 23(02). Maisel, J. L. (2016). Impact of older adults’ neighborhood perceptions on walking behavior. Journal of Aging and Physical Activity, 24(2), 247–255. https://doi.org/10.1123/japa.2014-0278 Malina, R. M. (2001). Physical activity and fitness: Pathways from childhood to adulthood. American Journal of Human Biology, 13, 162–172. https://doi.org/10.1002/1520-6300(200102/03)13:2<162::AID-AJHB1025>3.0.CO;2-T Malone, K., & Rudner, J. (2011). Global Perspectives on Children’s Independent Mobility: A Socio-Cultural Comparison and Theoretical Discussion of Children’s Lives in Four Countries in Asia and Africa. Global Studies of Childhood, 1(3), 243–259. https://doi.org/10.2304/gsch.2011.1.3.243 Maly, M. R., Costigan, P. A., & Olney, S. J. (2007). Self-efficacy mediates walking performance in older adults with knee osteoarthritis. Journals of Gerontology - Series A Biological Sciences and Medical Sciences, 62(10), 1142–1146. https://doi.org/10.1093/gerona/62.10.1142 Manaugh, K., & El-Geneidy, A. (2011). Validating walkability indices: How do different households respond to the walkability of their neighborhood? Transportation Research Part D: Transport   209 and Environment, 16(4), 309–315. https://doi.org/10.1016/j.trd.2011.01.009 Mandic, S., Williams, J., Moore, A., Hopkins, D., Flaherty, C., Wilson, G., García Bengoechea, E., & Spence, J. C. (2016). Built Environment and Active Transport to School (BEATS) Study: protocol for a cross-sectional study. BMJ Open, 6(5), e011196. https://doi.org/10.1136/bmjopen-2016-011196 Marquet, O., & Miralles-Guasch, C. (2015). Neighbourhood vitality and physical activity among the elderly: THE role of walkable environments on active ageing in Barcelona, Spain. Social Science and Medicine, 135, 24–30. https://doi.org/10.1016/j.socscimed.2015.04.016 Marselle, M. R., Irvine, K. N., & Warber, S. L. (2013). Walking for Well-Being: Are Group Walks in Certain Types of Natural Environments Better for Well-Being than Group Walks in Urban Environments? International Journal of Environmental Research and Public Health, 10(11), 5603–5628. https://doi.org/10.3390/ijerph10115603 Matthews, C. E., Jurj, A. L., Shu, X. O., Li, H. L., Yang, G., Li, Q., Gao, Y. T., & Zheng, W. (2007). Influence of exercise, walking, cycling, and overall nonexercise physical activity on mortality in Chinese women. American Journal of Epidemiology, 165(12), 1343–1350. https://doi.org/10.1093/aje/kwm088 Mcauley, E., Szabo, A., Gothe, N., & Olson, E. A. (2011). Self-Efficacy: Implications for Physical Activity, Function, and Functional Limitations in Older Adults. American Journal of Lifestyle Medicine, 5(4), 361–369. https://doi.org/10.1177/1559827610392704 McCormack, G., Giles-Corti, B., Lange, A., Smith, T., Martin, K., & Pikora, T. J. (2004). An update of recent evidence of the relationship between objective and self-report measures of the physical environment and physical activity behaviours. In Journal of science and medicine in sport / Sports Medicine Australia. https://doi.org/10.1016/s1440-2440(04)80282-2 McDonald, N. C. (2012). Is there a gender gap in school travel? An examination of US children and adolescents. Journal of Transport Geography, 20(1), 80–86. https://doi.org/10.1016/j.jtrangeo.2011.07.005 McFadden, D. (1973). Conditional logit analysis of qualitative choice behavior. In P. Zarembka (Ed.), Frontiers in Econometrics (pp. 105–142). Academic Press. McLeroy, K. R., Bibeau, D., Steckler, A., & Glanz, K. (1988). Ecological Perspective on Promotion Programs. Health Education Quarterly, 15(4), 351–377. https://doi.org/10.1177/109019818801500401 Mechanic, D., & Tanner, J. (2007). Vulnerable People, Groups, And Populations: Societal View. Health Affairs, 26(5), 1220–1230. https://doi.org/10.1377/hlthaff.26.5.1220 Mecredy, G., Pickett, W., & Janssen, I. (2011). Street connectivity is negatively associated with physical activity in Canadian youth. International Journal of Environmental Research and Public Health, 8(8), 3333–3350. https://doi.org/10.3390/ijerph8083333   210 Mendes de Leon, C. F., Cagney, K. A., Bienias, J. L., Barnes, L. L., Skarupski, K. a, Scherr, P. A., & Evans, D. A. (2009). Neighborhood Social Cohesion and Disorder in Relation to Walking in Community-Dwelling Older Adults. Journal of Aging and Health, 21(1), 155–171. https://doi.org/10.1177/0898264308328650 Mendonça, G., Cheng, L. A., Mélo, E. N., & de Farias Júnior, J. C. (2014). Physical activity and social support in adolescents: a systematic review. Health Education Research, 29(5), 822–839. https://doi.org/10.1093/her/cyu017 Millstein, R. A., Cain, K. L., Sallis, J. F., Conway, T. L., Geremia, C., Frank, L. D., Chapman, J., Van Dyck, D., Dipzinski, L. R., Kerr, J., Glanz, K., & Saelens, B. E. (2013). Development, scoring, and reliability of the Microscale Audit of Pedestrian Streetscapes (MAPS). BMC Public Health, 13(1), 403. https://doi.org/10.1186/1471-2458-13-403 Millward, H., Spinney, J., & Scott, D. (2013). Active-transport walking behavior: Destinations, durations, distances. Journal of Transport Geography, 28, 101–110. https://doi.org/10.1016/j.jtrangeo.2012.11.012 Mishler, A. D., & Neider, M. B. (2017). Improving Wayfinding for Older Users With Selective Attention Deficits. Ergon Des., 25(1), 11–16. https://doi.org/10.1177/1064804616659992 Mitra, R., Faulkner, G. E., Buliung, R. N., & Stone, M. R. (2014). Do parental perceptions of the neighbourhood environment influence children’s independent mobility? Evidence from Toronto, Canada. Urban Studies, 51(16), 3401–3419. https://doi.org/10.1177/0042098013519140 Molnar, B. E., Gortmaker, S. L., Bull, F. C., & Buka, S. L. (2004). Unsafe to Play? Neighborhood Disorder and Lack of Safety Predict Reduced Physical Activity among Urban Children and Adolescents. American Journal of Health Promotion, 18(5), 378–386. https://doi.org/10.4278/0890-1171-18.5.378 Monsalves, M. J., Bangdiwala, A. S., Thabane, A., & Bangdiwala, S. I. (2020). LEVEL (Logical Explanations & Visualizations of Estimates in Linear mixed models): recommendations for reporting multilevel data and analyses. BMC Medical Research Methodology, 20(1), 3. https://doi.org/10.1186/s12874-019-0876-8 Mooney, S. J., Joshi, S., Cerdá, M., Kennedy, G. J., Beard, J. R., & Rundle, A. G. (2017). Contextual Correlates of Physical Activity among Older Adults: A Neighborhood Environment-Wide Association Study (NE-WAS). Cancer Epidemiology Biomarkers & Prevention, 26(4), 495–504. https://doi.org/10.1158/1055-9965.EPI-16-0827 Moschny, A., Platen, P., Klaaßen-Mielke, R., Trampisch, U., & Hinrichs, T. (2011). Barriers to physical activity in older adults in Germany: a cross-sectional study. International Journal of Behavioral Nutrition and Physical Activity, 8(1), 121. https://doi.org/10.1186/1479-5868-8-121 Motl, R. W., Dishman, R. K., Saunders, R. P., Dowda, M., Felton, G., Ward, D. S., & Pate, R. R. (2002). Examining social-cognitive determinants of intention and physical activity among black and white adolescent girls using structural equation modeling. Health Psychology, 21(5), 459–467. https://doi.org/10.1037/0278-6133.21.5.459   211 Mottram, S., Peat, G., Thomas, E., Wilkie, R., & Croft, P. (2008). Patterns of pain and mobility limitation in older people: Cross-sectional findings from a population survey of 18,497 adults aged 50 years and over. Quality of Life Research, 17(4), 529–539. https://doi.org/10.1007/s11136-008-9324-7 Moudon, A. V., Lee, C., Cheadle, A. D., Garvin, C., Johnson, D. B., Schmid, T. L., & Weathers, R. D. (2007). Attributes of environments supporting walking. American Journal of Health Promotion, 21(5), 448–459. https://doi.org/10.4278/0890-1171-21.5.448 Moudon, A. V., Lee, C., Cheadle, A. D., Garvin, C., Johnson, D., Schmid, T. L., Weathers, R. D., & Lin, L. (2006). Operational Definitions of Walkable Neighborhood: Theoretical and Empirical Insights. Journal of Physical Activity and Health, 3(s1), S99–S117. https://doi.org/10.1123/jpah.3.s1.s99 Mullen, S. P., Palac, D. E., & Bryant, L. L. (2016). Maps to apps: Evaluating wayfinding technology. In Community Wayfinding: Pathways to Understanding. https://doi.org/10.1007/978-3-319-31072-5_8 Musselman, K., & Brouwer, B. (2005). Gender-related differences in physical performance among seniors. Journal of Aging and Physical Activity, 13(3), 239–253. https://doi.org/10.1123/japa.13.3.239 Mutrie, N., & Faulkner, G. (2012). Physical Activity: Positive Psychology in Motion. In Positive Psychology in Practice (pp. 146–164). John Wiley & Sons, Inc. https://doi.org/10.1002/9780470939338.ch9 Myers, N. (1989). Synergistic interactions and environment. BioScience, 39(08), 506. https://doi.org/10.1093/bioscience/39.8.506 Næss, P. (2011). ‘New urbanism’ or metropolitan-level centralization? A comparison of the influences of metropolitan-level and neighborhood-level urban form characteristics on travel behavior. Journal of Transport and Land Use, 4(1), 25–44. https://doi.org/10.5198/jtlu.v4i1.170 Nasar, J. L., Holloman, C., & Abdulkarim, D. (2015). Street characteristics to encourage children to walk. Transportation Research Part A: Policy and Practice, 72, 62–70. https://doi.org/10.1016/j.tra.2014.12.004 Nasri, A., & Zhang, L. (2014). Assessing the Impact of Metropolitan- , County- , and Local-Level Built Environment on Travel Behavior : Evidence from 19 U . S . Urban Areas. Wstlur 2014, 1(301), 1–17. https://doi.org/10.1061/(ASCE)UP.1943-5444.0000226. Nechyba, T. J., & Walsh, R. P. (2004). URBAN sprawl. British Medical Journal, 2(4794), 1142–1143. https://doi.org/10.1257/0895330042632681 Neckerman, K. M., Lovasi, G. S., Davies, S., Purciel, M., Quinn, J., Feder, E., Raghunath, N., Wasserman, B., & Rundle, A. (2009). Disparities in Urban Neighborhood Conditions: Evidence from GIS Measures and Field Observation in New York City. Journal of Public Health Policy, 30(S1), S264–S285. https://doi.org/10.1057/jphp.2008.47   212 Neckerman, K. M., Purciel-Hill, M., Quinn, J. W., & Rundle, A. (2013). Urban Design Qualities for New York City BT  - Measuring Urban Design: Metrics for Livable Places (R. Ewing, O. Clemente, K. M. Neckerman, M. Purciel-Hill, J. W. Quinn, & A. Rundle (eds.); pp. 63–82). Island Press/Center for Resource Economics. https://doi.org/10.5822/978-1-61091-209-9_4 Nelson, M. E., Rejeski, W. J., Blair, S. N., Duncan, P. W., Judge, J. O., King, A. C., Macera, C. A., & Castaneda-Sceppa, C. (2007). Physical activity and public health in older adults: Recommendation from the American College of Sports Medicine and the American Heart Association. Circulation, 116(9), 1094–1105. https://doi.org/10.1161/CIRCULATIONAHA.107.185650 Nordstrom, C. K., Diez Roux, A. V., Schulz, R., Haan, M. N., Jackson, S. A., & Balfour, J. L. (2007). Socioeconomic position and incident mobility impairment in the Cardiovascular Health Study. BMC Geriatrics, 7, 1–8. https://doi.org/10.1186/1471-2318-7-11 Norman, G. J., Nutter, S. K., Ryan, S., Sallis, J. F., Calfas, K. J., & Patrick, K. (2006). Community Design and Access to Recreational Facilities as Correlates of Adolescent Physical Activity and Body-Mass Index. Journal of Physical Activity and Health, 03(S1), S118–S128. https://doi.org/10.1123/jpah.3.s1.s118 Notthoff, N., Reisch, P., & Gerstorf, D. (2017). Individual Characteristics and Physical Activity in Older Adults: A Systematic Review. Gerontology, 63(5), 443–459. https://doi.org/10.1159/000475558 O’Brien, R. M. (2007). A caution regarding rules of thumb for variance inflation factors. Quality and Quantity, 41(5), 673–690. https://doi.org/10.1007/s11135-006-9018-6 Orsega-Smith, E. M., Payne, L. L., Mowen, A. J., Ho, C. H., & Godbey, G. C. (2007). The role of social support and self-efficacy in shaping the leisure time physical activity of older adults. Journal of Leisure Research, 39(4), 705–727. https://doi.org/10.1080/00222216.2007.11950129 Ory, M. G., Towne, S. D., Won, J., Forjuoh, S. N., & Lee, C. (2016). Social and environmental predictors of walking among older adults. BMC Geriatrics, 16(1), 155. https://doi.org/10.1186/s12877-016-0327-x Osama, A., Albitar, M., Sayed, T., & Bigazzi, A. (2020). Determining if walkability and bikeability indices reflect pedestrian and cyclist safety. Transportation Research Record, 2674(09). https://doi.org/10.1177/0361198120931844 Owen, N., Cerin, E., Leslie, E., duToit, L., Coffee, N., Frank, L. D., Bauman, A. E., Hugo, G., Saelens, B. E., & Sallis, J. F. (2007). Neighborhood Walkability and the Walking Behavior of Australian Adults. American Journal of Preventive Medicine, 33(5), 387–395. https://doi.org/10.1016/j.amepre.2007.07.025 Pabayo, R. A., Gauvin, L., Barnett, T. A., Morency, P., Nikiéma, B., & Séguin, L. (2012). Understanding the determinants of active transportation to school among children: Evidence of environmental injustice from the Quebec longitudinal study of child development. Health and Place, 18(2), 163–171. https://doi.org/10.1016/j.healthplace.2011.08.017   213 Pacilli, M. G., Giovannelli, I., Prezza, M., & Augimeri, M. L. (2013). Children and the public realm: antecedents and consequences of independent mobility in a group of 11-13-year-old Italian children. Children’s Geographies, 11(4), 377–393. https://doi.org/10.1080/14733285.2013.812277 Pacilli, M. G., Giovannelli, I., & Spaccatini, F. (2016). Children’s Independent Mobility: Antecedents and Consequences at Macro- and Microlevels. In C. Ni Laoire, A. White, & T. Skelton (Eds.), Movement, Mobilities, and Journeys, Geographies of Children and Young People (pp. 1–21). Springer Singapore. https://doi.org/10.1007/978-981-4585-93-4_6-2 Page, A. S., Cooper, A. R., Griew, P., & Jago, R. (2010). Independent mobility, perceptions of the built environment and children’s participation in play, active travel and structured exercise and sport: the PEACH Project. International Journal of Behavioral Nutrition and Physical Activity, 7(1), 17. https://doi.org/10.1186/1479-5868-7-17 Pang, B., Kubacki, K., & Rundle-Thiele, S. (2017). Promoting active travel to school: a systematic review (2010–2016). BMC Public Health, 17(1), 638. https://doi.org/10.1186/s12889-017-4648-2 Panter, J. R., Jones, A. P., & van Sluijs, E. M. F. (2008). Environmental determinants of active travel in youth: A review and framework for future research. International Journal of Behavioral Nutrition and Physical Activity, 5, 1–14. https://doi.org/10.1186/1479-5868-5-34 Park, S., Choi, K., & Lee, J. S. (2015). To Walk or Not to Walk: Testing the Effect of Path Walkability on Transit Users’ Access Mode Choices to the Station. International Journal of Sustainable Transportation, 9(8), 529–541. https://doi.org/10.1080/15568318.2013.825036 Park, S., Choi, K., & Lee, J. S. (2017). Operationalization of path walkability for sustainable transportation. International Journal of Sustainable Transportation, 11(7), 471–485. https://doi.org/10.1080/15568318.2016.1226996 Park, S., Deakin, E., & Jang, K. (2015). Can Good Walkability Expand the Size of Transit-Oriented Developments? Transportation Research Record: Journal of the Transportation Research Board, 2519(1), 157–164. https://doi.org/10.3141/2519-17 Peck, J. (2012). Austerity urbanism: American cities under extreme economy. City, 16(6), 626–655. https://doi.org/10.1080/13604813.2012.734071 Pedigo, A., Seaver, W., & Odoi, A. (2011). Identifying Unique Neighborhood Characteristics to Guide Health Planning for Stroke and Heart Attack: Fuzzy Cluster and Discriminant Analyses Approaches. PLoS ONE, 6(7), e22693. https://doi.org/10.1371/journal.pone.0022693 Pelclová, J., Frömel, K., & Cuberek, R. (2013). Gender-Specific associations between perceived neighbourhood walkability and meeting walking recommendations when walking for transport and recreation for Czech inhabitants over 50 years of age. International Journal of Environmental Research and Public Health, 11(1), 527–536. https://doi.org/10.3390/ijerph110100527   214 Pels, F., & Kleinert, J. (2016). Loneliness and physical activity: A systematic review. International Review of Sport and Exercise Psychology, 9(1), 231–260. https://doi.org/10.1080/1750984X.2016.1177849 Perkins+Will, Gail M. Goldman Associates, Olin Partnership, & Public Solutions. (2011). San Diego Downtown Design Guidelines. https://www.sandiego.gov/sites/default/files/downtown-design-guidelines.pdf Perkins, J. M., Multhaup, K. S., Perkins, H. W., & Barton, C. (2008). Self-Efficacy and Participation in Physical and Social Activity Among Older Adults in Spain and the United States. The Gerontologist, 48(1), 51–58. https://doi.org/10.1093/geront/48.1.51 Pikora, T. J., Bull, F. C. ., Jamrozik, K., Knuiman, M., Giles-Corti, B., & Donovan, R. J. (2002). Developing a reliable audit instrument to measure the physical environment for physical activity. American Journal of Preventive Medicine, 23(3), 187–194. https://doi.org/10.1016/S0749-3797(02)00498-1 Pikora, T. J., Giles-Corti, B., Knuiman, M. W., Bull, F. C., Jamrozik, K., & Donovan, R. J. (2006). Neighborhood environmental factors correlated with walking near home: Using SPACES. Medicine and Science in Sports and Exercise, 38(4), 708–714. https://doi.org/10.1249/01.mss.0000210189.64458.f3 Piro, F. N., Nœss, Ø., & Claussen, B. (2006). Physical activity among elderly people in a city population: The influence of neighbourhood level violence and self perceived safety. Journal of Epidemiology and Community Health, 60(7), 626–632. https://doi.org/10.1136/jech.2005.042697 Poitras, V. J., Gray, C. E., Borghese, M. M., Carson, V., Chaput, J.-P., Janssen, I., Katzmarzyk, P. T., Pate, R. R., Connor Gorber, S., Kho, M. E., Sampson, M., & Tremblay, M. S. (2016). Systematic review of the relationships between objectively measured physical activity and health indicators in school-aged children and youth. Applied Physiology, Nutrition, and Metabolism, 41(6 (Suppl. 3)), S197–S239. https://doi.org/10.1139/apnm-2015-0663 Pollard, T. M., & Wagnild, J. M. (2017). Gender differences in walking (for leisure, transport and in total) across adult life: A systematic review. BMC Public Health, 17(1), 1–11. https://doi.org/10.1186/s12889-017-4253-4 Pyper, E., Harrington, D., & Manson, H. (2016). The impact of different types of parental support behaviours on child physical activity, healthy eating, and screen time: A cross-sectional study. BMC Public Health, 16(1), 1–15. https://doi.org/10.1186/s12889-016-3245-0 Rajamani, J., Bhat, C. R., Handy, S. L., Knaap, G., & Song, Y. (2003). Assessing Impact of Urban Form Measures on Nonwork Trip Mode Choice After Controlling for Demographic and Level-of-Service Effects. Transportation Research Record, 1831(1), 158–165. https://doi.org/10.3141/1831-18 Ramakreshnan, L., Aghamohammadi, N., Fong, C. S., & Sulaiman, N. M. (2021). A comprehensive bibliometrics of ‘walkability’ research landscape: visualization of the scientific progress and future prospects. Environmental Science and Pollution Research, 28(2), 1357–1369.   215 https://doi.org/10.1007/s11356-020-11305-x Rantakokko, M., Mänty, M., Iwarsson, S., Törmäkangas, T., Leinonen, R., Heikkinen, E., & Rantanen, T. (2009). Fear of Moving Outdoors and Development of Outdoor Walking Difficulty in Older People. Journal of the American Geriatrics Society, 57(4), 634–640. https://doi.org/10.1111/j.1532-5415.2009.02180.x Raudsepp, L., & Viira, R. (2000). Sociocultural correlates of physical activity in adolescents. Pediatric Exercise Science, 12(1), 51–60. https://doi.org/10.1123/pes.12.1.51 Reis, J. P., Macera, C. A., Ainsworth, B. E., & Hipp, D. A. (2008). Prevalence of total daily walking among US adults, 2002-2003. Journal of Physical Activity and Health, 5(3), 337–346. https://doi.org/10.1123/jpah.5.3.337 Reyer, M., Fina, S., Siedentop, S., & Schlicht, W. (2014). Walkability is only part of the story: walking for transportation in Stuttgart, Germany. International Journal of Environmental Research and Public Health, 11(6), 5849–5865. https://doi.org/10.3390/ijerph110605849 Rhodes, R. E., Berry, T., Craig, C. L., Faulkner, G., Latimer-Cheung, A., Spence, J. C., & Tremblay, M. S. (2013). Understanding parental support of child physical activity behavior. American Journal of Health Behavior, 37(4), 469–477. https://doi.org/10.5993/AJHB.37.4.5 Rhodes, R. E., Guerrero, M. D., Vanderloo, L. M., Barbeau, K., Birken, C. S., Chaput, J. P., Faulkner, G., Janssen, I., Madigan, S., Mâsse, L. C., McHugh, T. L., Perdew, M., Stone, K., Shelley, J., Spinks, N., Tamminen, K. A., Tomasone, J. R., Ward, H., Welsh, F., & Tremblay, M. S. (2020). Development of a consensus statement on the role of the family in the physical activity, sedentary, and sleep behaviours of children and youth. International Journal of Behavioral Nutrition and Physical Activity, 17(1), 1–31. https://doi.org/10.1186/s12966-020-00973-0 Rhodes, R. E., Perdew, M., & Malli, S. (2020). Correlates of Parental Support of Child and Youth Physical Activity: a Systematic Review. International Journal of Behavioral Medicine, 27(6), 636–646. https://doi.org/10.1007/s12529-020-09909-1 Rhodes, R. E., Spence, J. C., Berry, T., Deshpande, S., Faulkner, G., Latimer-Cheung, A. E., O’Reilly, N., & Tremblay, M. S. (2015). Predicting Changes Across 12 Months in Three Types of Parental Support Behaviors and Mothers’ Perceptions of Child Physical Activity. Annals of Behavioral Medicine, 49(6), 853–864. https://doi.org/10.1007/s12160-015-9721-4 Rhodes, R. E., Stearns, J., Berry, T., Faulkner, G., Latimer-Cheung, A. E., O’Reilly, N., Tremblay, M. S., Vanderloo, L., & Spence, J. C. (2019). Predicting parental support and parental perceptions of child and youth movement behaviors. Psychology of Sport and Exercise, 41(November 2018), 80–90. https://doi.org/10.1016/j.psychsport.2018.11.016 Riazi, N. A., Blanchette, S., Trudeau, F., Larouche, R., Tremblay, M. S., & Faulkner, G. (2019). Correlates of children’s independent mobility in canada: A multi-site study. International Journal of Environmental Research and Public Health, 16(16), 1–14. https://doi.org/10.3390/ijerph16162862   216 Richardson, A. S., Troxel, W. M., Ghosh-Dastidar, M. B., Beckman, R., Hunter, G. P., DeSantis, A. S., Colabianchi, N., & Dubowitz, T. (2017). One size doesn’t fit all: cross-sectional associations between neighborhood walkability, crime and physical activity depends on age and sex of residents. BMC Public Health, 17(1), 1–11. https://doi.org/10.1186/s12889-016-3959-z Rissel, C., Curac, N., Greenaway, M., & Bauman, A. (2012). Physical Activity Associated with Public Transport Use—A Review and Modelling of Potential Benefits. International Journal of Environmental Research and Public Health, 9(7), 2454–2478. https://doi.org/10.3390/ijerph9072454 Rodríguez, D. A., & Joo, J. (2004). The relationship between non-motorized mode choice and the local physical environment. Transportation Research Part D: Transport and Environment, 9(2), 151–173. https://doi.org/10.1016/j.trd.2003.11.001 Rodríguez, D. A., Merlin, L., Prato, C. G., Conway, T. L., Cohen, D., Elder, J. P., Evenson, K. R., McKenzie, T. L., Pickrel, J. L., & Veblen-Mortenson, S. (2014). Influence of the Built Environment on Pedestrian Route Choices of Adolescent Girls. Environment and Behavior, 47(4), 359–394. https://doi.org/10.1177/0013916513520004 Roe, J., & Aspinall, P. (2011). The restorative benefits of walking in urban and rural settings in adults with good and poor mental health. Health and Place, 17(1), 103–113. https://doi.org/10.1016/j.healthplace.2010.09.003 Romero, A. J. (2005). Low-income neighborhood barriers and resources for adolescents’ physical activity. Journal of Adolescent Health, 36(3), 253–259. https://doi.org/10.1016/j.jadohealth.2004.02.027 Rosenberg, D., Ding, D., Sallis, J. F., Kerr, J., Norman, G. J., Durant, N., Harris, S. K., & Saelens, B. E. (2009). Neighborhood Environment Walkability Scale for Youth (NEWS-Y): Reliability and relationship with physical activity. Preventive Medicine, 49(2–3), 213–218. https://doi.org/10.1016/j.ypmed.2009.07.011 Rothman, K. J., Greenland, S., & Walker, A. M. (1974). Concepts of interaction. Modern Epidemiology: Third Edition, 99(06), 385–388. Rothman, L., Macpherson, A. K., Ross, T., & Buliung, R. N. (2018). The decline in active school transportation (AST): A systematic review of the factors related to AST and changes in school transport over time in North America. Preventive Medicine, 111(July 2017), 314–322. https://doi.org/10.1016/j.ypmed.2017.11.018 Saelens, B. E., Frank, L. D., Auffrey, C., Whitaker, R. C., Burdette, H. L., & Colabianchi, N. (2006). Measuring Physical Environments of Parks and Playgrounds: EAPRS Instrument Development and Inter-Rater Reliability. Journal of Physical Activity and Health, 3(s1), S190–S207. https://doi.org/10.1123/jpah.3.s1.s190 Saelens, B. E., & Handy, S. L. (2008). Built environment correlates of walking: a review. Medicine and Science in Sports and Exercise, 40(7 Suppl), S550–S566. https://doi.org/10.1249/MSS.0b013e31817c67a4   217 Saelens, B. E., Sallis, J. F., Black, J. B., & Chen, D. (2003). Neighborhood-Based Differences in Physical Activity: An Environmental Scale Evaluation. Research and Practice, 93(9), 1552–1558. https://doi.org/10.2105/ajph.93.9.1552 Saelens, B. E., Sallis, J. F., & Frank, L. D. (2003). Environmental correlates of walking and cycling: findings from the transportation, urban design, and planning literatures. Annals of Behavioral Medicine : A Publication of the Society of Behavioral Medicine, 25(c), 80–91. https://doi.org/10.1207/S15324796ABM2502_03 Saelens, B. E., Vernez Moudon, A., Kang, B., Hurvitz, P. M., & Zhou, C. (2014). Relation Between Higher Physical Activity and Public Transit Use. American Journal of Public Health, 104(5), 854–859. https://doi.org/10.2105/AJPH.2013.301696 Sallis, J. F., Cain, K. L., Conway, T. L., Gavand, K. A., Millstein, R. A., Geremia, C. M., Frank, L. D., Saelens, B. E., Glanz, K., & King, A. C. (2015). Is Your Neighborhood Designed to Support Physical Activity ? A Brief Streetscape Audit Tool. Preventing Chronic Disease, 12(9), 1–11. https://doi.org/10.5888/pcd12.150098 Sallis, J. F., Cervero, R. B., Ascher, W., Henderson, K. A., Kraft, M. K., & Kerr, J. (2006). AN ECOLOGICAL APPROACH TO CREATING ACTIVE LIVING COMMUNITIES. Annual Review of Public Health, 27(1), 297–322. https://doi.org/10.1146/annurev.publhealth.27.021405.102100 Sallis, J. F., Floyd, M. F., Rodr??guez, D. A., Saelens, B. E., Rodriguez, D. A., & Saelens, B. E. (2012). The Role of Built Environments in Physical Activity, Obesity, and CVD. Circulation, 125(5), 729–737. https://doi.org/10.1161/CIRCULATIONAHA.110.969022 Sallis, J. F., Frank, L. D., Saelens, B. E., & Kraft, M. K. (2004). Active transportation and physical activity: Opportunities for collaboration on transportation and public health research. Transportation Research Part A: Policy and Practice, 38(4), 249–268. https://doi.org/10.1016/j.tra.2003.11.003 Sallis, J. F., & Glanz, K. (2006). The role of built environments in physical activity, eating, and obesity in childhood. Future of Children, 16(01), 89–108. https://doi.org/10.1353/foc.2006.0009 Sallis, J. F., Grossman, R. M., Pinski, R. B., Patterson, T. L., & Nader, P. R. (1987). The development of scales to measure social support for diet and exercise behaviors. Preventive Medicine, 16(6), 825–836. https://doi.org/10.1016/0091-7435(87)90022-3 Sallis, J. F., & Owen, N. (1998). Physical activity & behavioural medicine. Sage Publications Inc. Sallis, J. F., Owen, N., & Fisher, E. B. (2008). Ecological models of health behavior. In Health Behavior and Health Education: Theory, Research, and Practice, 4th edition. https://doi.org/10.7326/0003-4819-116-4-350_1 Sallis, J. F., Prochaska, J. J., & Taylor, W. C. (2000). A review of correlates of physical activity of children and adolescents. Medicine and Science in Sports and Exercise, 32(5), 963–975. https://doi.org/10.1097/00005768-200005000-00014   218 Sallis, J. F., Saelens, B. E., Frank, L. D., Conway, T. L., Slymen, D. J., Cain, K. L., Chapman, J. E., & Kerr, J. (2009). Neighborhood built environment and income: Examining multiple health outcomes. Social Science and Medicine, 68(7), 1285–1293. https://doi.org/10.1016/j.socscimed.2009.01.017 Sallis, J. F., Slymen, D. J., Conway, T. L., Frank, L. D., Saelens, B. E., Cain, K., & Chapman, J. E. (2011). Income disparities in perceived neighborhood built and social environment attributes. Health and Place, 17(6), 1274–1283. https://doi.org/10.1016/j.healthplace.2011.02.006 Sallis, J. F., Taylor, W. C., Dowda, M., Freedson, P. S., & Pate, R. R. (2002). Correlates of Vigorous Physical Activity for Children in Grades 1 through 12: Comparing Parent-Reported and Objectively Measured Physical Activity. Pediatric Exercise Science, 14(1), 30–44. https://doi.org/10.1123/pes.14.1.30 10.1123/pes.14.1.30 10.1123/pes.14.1.30 10.1123/pes.14.1.30 Salomon, I., & Ben-Akiva, M. (1983). The Use of the Life-Style Concept in Travel Demand Models. Environment and Planning A: Economy and Space, 15(5), 623–638. https://doi.org/10.1068/a150623 Satariano, W. A., Guralnik, J. M., Jackson, R. J., Marottoli, R. A., Phelan, E. A., & Prohaska, T. R. (2012). Mobility and aging: New directions for public health action. American Journal of Public Health, 102(8), 1508–1515. https://doi.org/10.2105/AJPH.2011.300631 Sayers, S. P., Jette, A. M., Haley, S. M., Heeren, T. C., Guralnik, J. M., & Fielding, R. A. (2004). Validation of the Late-Life Function and Disability Instrument. Journal of the American Geriatrics Society, 52(9), 1554–1559. https://doi.org/10.1111/j.1532-5415.2004.52422.x Scheiner, J. (2007). Mobility Biographies : Elements of a Biographical Theory of Travel Demand. Erdkunde, 61(02), 161–173. Schwartz, M. A., Aytur, S. A., Evenson, K. R., & Rodríguez, D. A. (2009). Are Perceptions about Worksite Neighborhoods and Policies Associated with Walking? American Journal of Health Promotion, 24(2), 146–151. https://doi.org/10.4278/ajhp.071217134 Seabra, A. C., Seabra, A. F., Mendonça, D. M., Brustad, R., Maia, J. A., Fonseca, A. M., & Malina, R. M. (2013). Psychosocial correlates of physical activity in school children aged 8-10 years. European Journal of Public Health, 23(5), 794–798. https://doi.org/10.1093/eurpub/cks149 Shankar, A., McMunn, A., Banks, J., & Steptoe, A. (2011). Loneliness, Social Isolation, and Behavioral and Biological Health Indicators in Older Adults. Health Psychology, 30(4), 377–385. https://doi.org/10.1037/a0022826 Shashank, A., & Schuurman, N. (2019). Unpacking walkability indices and their inherent assumptions. Health and Place, 55, 145–154. https://doi.org/10.1016/j.healthplace.2018.12.005 Shaw, B., Bicket, M., Elliott, B., Fagan-Watson, B., Mocca, E., & Hillman, M. (2015). Children’s Independent Mobility: an international comparison and recommendations for action. In Policy Studies Institute (Issue July). http://www.psi.org.uk/docs/7350_PSI_Report_CIM_final.pdf   219 Shi, L., & Stevens, G. (2010). A general framework to study vulnerable populations. In L. Shi & G. Stevens (Eds.), Vulnerable populations in the United States (pp. 1–34). A Wiley Imprint. Shigematsu, R., Sallis, J. F., Conway, T. L., Saelens, B. E., Frank, L. D., Cain, K. L., Chapman, J. E., & King, A. C. (2009). Age differences in the relation of perceived neighborhood environment to walking. Medicine and Science in Sports and Exercise, 41(7), 314–321. https://doi.org/10.1249/MSS.0b013e318185496c Siegel, P. Z., Brackbill, R. M., & Heath, G. W. (1995). The epidemiology of walking for exercise: implications for promoting activity among sedentary groups. American Journal of Public Health, 85(5), 706–710. https://doi.org/10.2105/AJPH.85.5.706 Simon, H. A. (1959). Theories of Decision-Making in Economics and Behavioral Science. American Economic Review, 49(3), 253–283. https://doi.org/10.1257/aer.99.1.i Sitkowski, R. J., & Russell, J. (2007). NEW YORK ZONING LAW AND PRACTICE REPORT Editor-in-Chief. https://www.joelrussell.com/articles/Form-based Codes NY article (final).pdf Smart Growth America. (2016). Best Complete Streets Policies of 2015. https://smartgrowthamerica.org/resources/best-complete-streets-policies-of-2015/ Stamatakis, E., Hamer, M., & Murphy, M. H. (2018). What Hippocrates called ‘Man’s best medicine’: walking is humanity’s path to a better world. British Journal of Sports Medicine, 52(12), 753–754. https://doi.org/10.1136/bjsports-2018-099371 Stathopoulos, A., & Hess, S. (2012). Revisiting reference point formation, gains–losses asymmetry and non-linear sensitivities with an emphasis on attribute specific treatment. Transportation Research Part A: Policy and Practice, 46(10), 1673–1689. https://doi.org/https://doi.org/10.1016/j.tra.2012.08.005 Statistics Canada. (2019). Physical activity and screen time among Canadian children and youth, 2016 and 2017. https://www150.statcan.gc.ca/n1/pub/82-625-x/2019001/article/00003-eng.htm Statistics Canada. (2020). Table 13-10-0096-13 Physical activity, self reported, adult, by age group. https://doi.org/https://doi.org/10.25318/1310009601-eng Steinmetz-Wood, M., El-Geneidy, A., & Ross, N. A. (2020). Moving to policy-amenable options for built environment research: The role of micro-scale neighborhood environment in promoting walking. Health and Place. https://doi.org/10.1016/j.healthplace.2020.102462 Stephan, Y., Boiche, J., Trouilloud, D., Deroche, T., & Sarrazin, P. (2011). The relation between risk perceptions and physical activity among older adults: A prospective study. Psychology and Health, 26(7), 887–897. https://doi.org/10.1080/08870446.2010.509798 Sterdt, E., Liersch, S., & Walter, U. (2014). Correlates of physical activity of children and adolescents: A systematic review of reviews. Health Education Journal, 73(1), 72–89. https://doi.org/10.1177/0017896912469578   220 STEWART, A. L., MILLS, K. M., KING, A. C., HASKELL, W. L., GILLIS, D., & RITTER, P. L. (2001). CHAMPS Physical Activity Questionnaire for Older Adults: outcomes for interventions. Medicine and Science in Sports and Exercise, 1126–1141. https://doi.org/10.1097/00005768-200107000-00010 Stokols, D. (1992). Establishing and maintaining healthy environments. Toward a social ecology of health promotion. The American Psychologist, 47(1), 6–22. https://doi.org/10.1037/0003-066X.47.1.6 Stokols, D., Grzywacz, J. G., Mcmahan, S., & Phillips, K. (2003). Increasing the Health Promotive Capacity of Human Environments. American Journal of Health Promotion, 18(1), 4–14. Sugiyama, T., & Thompson, C. W. (2008). Attributes and Quality of Life for Older People. Environment And Behavior, 1–19. Sun, F., Norman, I. J., & While, A. E. (2013). Physical activity in older people: A systematic review. BMC Public Health, 13(1). https://doi.org/10.1186/1471-2458-13-449 Talen, E. (2011). Sprawl retrofit: Sustainable urban form in unsustainable places. Environment and Planning B: Planning and Design, 38(6), 952–978. https://doi.org/10.1068/b37048 Tandon, P. S., Zhou, C., Sallis, J. F., Cain, K. L., Frank, L. D., & Saelens, B. E. (2012). Home environment relationships with children’s physical activity, sedentary time, and screen time by socioeconomic status. International Journal of Behavioral Nutrition and Physical Activity, 9(1), 88. https://doi.org/10.1186/1479-5868-9-88 Telford, R. M., Telford, R. D., Olive, L. S., Cochrane, T., & Davey, R. (2016). Why are girls less physically active than boys? Findings from the LOOK longitudinal study. PLoS ONE, 11(3), 1–11. https://doi.org/10.1371/journal.pone.0150041 The Lancet Public Health. (2019). Time to tackle the physical activity gender gap. The Lancet Public Health, 4(8), e360. https://doi.org/10.1016/S2468-2667(19)30135-5 Thornton, C. M., Cain, K. L., Conway, T. L., Kerr, J., Saelens, B. E., Frank, L. D., Glanz, K., & Sallis, J. F. (2017). Relation of adolescents’ physical activity to after-school recreation environment. Journal of Physical Activity and Health, 14(05), 382–388. https://doi.org/10.1123/jpah.2016-0365 Thornton, C. M., Conway, T. L., Cain, K. L., Gavand, K. A., Saelens, B. E., Frank, L. D., Geremia, C. M., Glanz, K., King, A. C., & Sallis, J. F. (2016). Disparities in pedestrian streetscape environments by income and race/ethnicity. SSM - Population Health, 2, 206–216. https://doi.org/10.1016/j.ssmph.2016.03.004 Thornton, C. M., Kerr, J., Conway, T. L., Saelens, B. E., Sallis, J. F., Ahn, D. K., Frank, L. D., Cain, K. L., & King, A. C. (2017a). Physical Activity in Older Adults: an Ecological Approach. Annals of Behavioral Medicine, 51(2), 159–169. https://doi.org/10.1007/s12160-016-9837-1 Thornton, C. M., Kerr, J., Conway, T. L., Saelens, B. E., Sallis, J. F., Ahn, D. K., Frank, L. D., Cain,   221 K. L., & King, A. C. (2017b). Physical Activity in Older Adults: an Ecological Approach. Annals of Behavioral Medicine, 51(2), 159–169. https://doi.org/10.1007/s12160-016-9837-1 Timperio, A., Crawford, D., Ball, K., & Salmon, J. (2017). Typologies of neighbourhood environments and children’s physical activity, sedentary time and television viewing. Health & Place, 43, 121–127. https://doi.org/https://doi.org/10.1016/j.healthplace.2016.10.004 Todd, M., Adams, M. A., Kurka, J., Conway, T. L., Cain, K. L., Buman, M. P., Frank, L. D., Sallis, J. F., & King, A. C. (2016). GIS-measured walkability, transit, and recreation environments in relation to older Adults’ physical activity: A latent profile analysis. Preventive Medicine, 93, 57–63. https://doi.org/10.1016/j.ypmed.2016.09.019 Tranter, P. (2015). Children’s Play in their Local Neighborhoods: Rediscovering the Value of Residential Streets. In S. T. Evans B., Horton J. (Ed.), Play, Recreation, Health and Well Being (pp. 211–236). Springer. https://doi.org/10.1007/978-981-4585-96-5_37-1 Tremblay, M. S., Gray, C. E., Akinroye, K., Harrington, D. M., Katzmarzyk, P. T., Lambert, E. V, Liukkonen, J., Maddison, R., Ocansey, R. T., Onywera, V. O., Prista, A., Reilly, J. J., Rodríguez Martínez, M. P., Sarmiento Duenas, O. L., Standage, M., & Tomkinson, G. (2014). Physical activity of children: a global matrix of grades comparing 15 countries. Journal of Physical Activity & Health, 11 Suppl 1(Supp 1), S113-25. https://doi.org/10.1123/jpah.2014-0177 Trost, S. G., & Loprinzi, P. D. (2011). Parental Influences on Physical Activity Behavior in Children and Adolescents: A Brief Review. American Journal of Lifestyle Medicine, 5(2), 171–181. https://doi.org/10.1177/1559827610387236 TROST, S. G., PATE, R. R., SALLIS, J. F., FREEDSON, P. S., TAYLOR, W. C., DOWDA, M., & SIRARD, J. (2002). Age and gender differences in objectively measured physical activity in youth. Medicine and Science in Sports and Exercise, 34(2), 350–355. https://doi.org/10.1097/00005768-200202000-00025 Tucker-Seeley, R. D., Subramanian, S. V., Li, Y., & Sorensen, G. (2009). Neighborhood Safety, Socioeconomic Status, and Physical Activity in Older Adults. American Journal of Preventive Medicine, 37(3), 207–213. https://doi.org/10.1016/j.amepre.2009.06.005 Tversky, A., & Kahneman, D. (1974). Judgment under Uncertainty: Heuristics and Biases. Science, 185(4157), 1124–1131. https://doi.org/10.1126/science.185.4157.1124 Ulfarsson, G. F., & Shankar, V. N. (2008). Children’s travel to school: Discrete choice modeling of correlated motorized and nonmotorized transportation modes using covariance heterogeneity. Environment and Planning B: Planning and Design, 35(2), 195–206. https://doi.org/10.1068/b3360 Van Cauwenberg, J., De Donder, L., Clarys, P., De Bourdeaudhuij, I., Owen, N., Dury, S., De Witte, N., Buffel, T., Verté, D., & Deforche, B. (2014). Relationships of individual, social, and physical environmental factors with older adults’ television viewing time. Journal of Aging and Physical Activity, 22(4), 508–517. https://doi.org/10.1123/JAPA.2013-0015   222 Van Cauwenberg, J., Van Holle, V., Simons, D., Deridder, R., Clarys, P., Goubert, L., Nasar, J., Salmon, J., De Bourdeaudhuij, I., & Deforche, B. (2012). Environmental factors influencing older adults’ walking for transportation: a study using walk-along interviews. International Journal of Behavioral Nutrition and Physical Activity. https://doi.org/10.1186/1479-5868-9-85 Van Dyck, D., Deforche, B., Cardon, G., & De Bourdeaudhuij, I. (2009). Neighbourhood walkability and its particular importance for adults with a preference for passive transport. Health & Place, 15(2), 496–504. https://doi.org/https://doi.org/10.1016/j.healthplace.2008.08.010 Van Dyck, D., Veitch, J., De Bourdeaudhuij, I., Thornton, L., & Ball, K. (2013). Environmental perceptions as mediators of the relationship between the objective built environment and walking among socio-economically disadvantaged women. International Journal of Behavioral Nutrition and Physical Activity, 10, 1–9. https://doi.org/10.1186/1479-5868-10-108 Van Holle, V., Van Cauwenberg, J., Gheysen, F., Van Dyck, D., Deforche, B., Van de Weghe, N., & De Bourdeaudhuij, I. (2016). The Association between Belgian Older Adults’ Physical Functioning and Physical Activity: What Is the Moderating Role of the Physical Environment? PLOS ONE, 11(2), e0148398. https://doi.org/10.1371/journal.pone.0148398 Van Holle, V., Van Cauwenberg, J., Van Dyck, D., Deforche, B., Van de Weghe, N., & De Bourdeaudhuij, I. (2014). Relationship between neighborhood walkability and older adults’ physical activity: Results from the Belgian Environmental Physical Activity Study in Seniors (BEPAS Seniors). International Journal of Behavioral Nutrition and Physical Activity, 11(1), 1–9. https://doi.org/10.1186/s12966-014-0110-3 Van Loon, J., Frank, L. D., Nettlefold, L., & Naylor, P. J. (2014). Youth physical activity and the neighbourhood environment: Examining correlates and the role of neighbourhood definition. Social Science and Medicine, 104, 107–115. https://doi.org/10.1016/j.socscimed.2013.12.013 van Zwieten, A., Saglimbene, V., Teixeira-Pinto, A., Howell, M., Howard, K., Craig, J. C., & Wong, G. (2018). The Impact of Age on Income-Related Health Status Inequalities from Birth to Adolescence: A Systematic Review with Cross-Country Comparisons. Journal of Pediatrics, 203, 380-390.e14. https://doi.org/10.1016/j.jpeds.2018.07.030 VanderWeele, T. J., & Knol, M. J. (2014). A Tutorial on Interaction. Epidemiologic Methods, 3(1). https://doi.org/10.1515/em-2013-0005 Vatcheva, K. P., Lee, M., McCormick, J. B., & Rahbar, M. H. (2016). Multicollinearity in Regression Analyses Conducted in Epidemiologic Studies. Epidemiology (Sunnyvale, Calif.), 6(2), 227. https://doi.org/10.4172/2161-1165.1000227 Veerman, J. L., Zapata-Diomedi, B., Gunn, L., McCormack, G. R., Cobiac, L. J., Mantilla Herrera, A. M., Giles-Corti, B., & Shiell, A. (2016). Cost-effectiveness of investing in sidewalks as a means of increasing physical activity: a RESIDE modelling study. BMJ Open. https://doi.org/10.1136/bmjopen-2016-011617 Veitch, J., Salmon, J., & Ball, K. (2008). Children’s active free play in local neighborhoods: a behavioral mapping study. Health Education Research, 23(5), 870–879.   223 https://doi.org/10.1093/her/cym074 Verloigne, M., Van Lippevelde, W., Maes, L., Yildirim, M., Chinapaw, M., Manios, Y., Androutsos, O., Kovács, É., Bringolf-Isler, B., Brug, J., & De Bourdeaudhuij, I. (2012). Levels of physical activity and sedentary time among 10- to 12-year-old boys and girls across 5 European countries using accelerometers: An observational study within the ENERGY-project. International Journal of Behavioral Nutrition and Physical Activity, 9, 1–8. https://doi.org/10.1186/1479-5868-9-34 Villanueva, K., Knuiman, M., Nathan, A., Giles-Corti, B., Christian, H., Foster, S., & Bull, F. (2014). The impact of neighborhood walkability on walking: Does it differ across adult life stage and does neighborhood buffer size matter? Health & Place, 25, 43–46. https://doi.org/https://doi.org/10.1016/j.healthplace.2013.10.005 Walk Score. (2021). Find Apartments for Rent and Rentals - Get Your Walk Score. https://www.walkscore.com/ Wang, H., & Yang, Y. (2019). Neighbourhood walkability: A review and bibliometric analysis. Cities, 93(May), 43–61. https://doi.org/10.1016/j.cities.2019.04.015 Wang, R., & Ware, J. H. (2013). Detecting Moderator Effects Using Subgroup Analyses. Prevention Science, 14(02), 111–120. https://doi.org/10.1007/s11121-011-0221-x Wang, X., Conway, T. L., Cain, K. L., Frank, L. D., Saelens, B. E., Geremia, C., Kerr, J., Glanz, K., Carlson, J. A., & Sallis, J. F. (2017). Interactions of psychosocial factors with built environments in explaining adolescents’ active transportation. Preventive Medicine, 100, 76–83. https://doi.org/10.1016/j.ypmed.2017.04.008 Watson, K. B., Carlson, S. A., Gunn, J. P., Galuska, D. A., O’Connor, A., Greenlund, K. J., & Fulton, J. E. (2016). Physical Inactivity Among Adults Aged 50 Years and Older — United States, 2014. MMWR. Morbidity and Mortality Weekly Report, 65(36), 954–958. https://doi.org/10.15585/mmwr.mm6536a3 Webber, S. C., Porter, M. M., & Menec, V. H. (2010). Mobility in older adults: A comprehensive framework. Gerontologist, 50(4), 443–450. https://doi.org/10.1093/geront/gnq013 Weber Corseuil, M., Hallal, P. C., Xavier Corseuil, H., Jayce Ceola Schneider, I., & D’Orsi, E. (2012). Safety from Crime and Physical Activity among Older Adults: A Population-Based Study in Brazil. Journal of Environmental and Public Health, 2012, 1–7. https://doi.org/10.1155/2012/641010 Whyte, W. H. (1980). The Social Life of Small Urban Spaces (Vol. 10, Issue 4). Project for Public Spaces. Wilcox, S., Bopp, M., Oberrecht, L., Kammermann, S. K., & McElmurray, C. T. (2003). Psychosocial and Perceived Environmental Correlates of Physical Activity in Rural and Older African American and White Women. Journals of Gerontology - Series B Psychological Sciences and Social Sciences, 58(6), 329–337. https://doi.org/10.1093/geronb/58.6.P329   224 Won, J., Lee, C., Forjuoh, S. N., & Ory, M. G. (2016). Neighborhood safety factors associated with older adults’ health-related outcomes: A systematic literature review. Social Science and Medicine, 165, 177–186. https://doi.org/10.1016/j.socscimed.2016.07.024 World Health Organization. (2002). ACTIVE AGEING: A POLICY FRAMEWORK Active Ageing. http://www.who.int/hpr/ World Health Organization. (2014). Global status report on noncommunicable diseases 2014 (Issue WHO/NMH/NVI/15.1). World Health Organization. https://apps.who.int/iris/bitstream/handle/10665/148114/9789241564854_eng.pdf Yin, L. (2017). Computers , Environment and Urban Systems Street level urban design qualities for walkability : Combining 2D and 3D GIS measures. Computers, Environment and Urban Systems, 64, 288–296. https://doi.org/10.1016/j.compenvurbsys.2017.04.001 Young, D. R., Cradock, A. L., Eyler, A. A., Fenton, M., Pedroso, M., Sallis, J. F., & Whitsel, L. P. (2020). Creating built environments that expand active transportation and active living across the United States: A policy statement from the American Heart Association. In Circulation (Vol. 142, Issue 11, pp. 167–183). Lippincott Williams and Wilkins. https://doi.org/10.1161/CIR.0000000000000878 Young, M. D., Plotnikoff, R. C., Collins, C. E., Callister, R., & Morgan, P. J. (2014). Social cognitive theory and physical activity: A systematic review and meta-analysis. Obesity Reviews, 15(12), 983–995. https://doi.org/10.1111/obr.12225 Zandieh, R., Martinez, J., Flacke, J., Jones, P., & Van Maarseveen, M. (2016). Older adults’ outdoor walking: Inequalities in neighbourhood safety, pedestrian infrastructure and aesthetics. International Journal of Environmental Research and Public Health, 13(12), 1–24. https://doi.org/10.3390/ijerph13121179 Zecevic, C. A., Tremblay, L., Lovsin, T., & Michel, L. (2010). Parental Influence on Young Children’s Physical Activity. International Journal of Pediatrics, 2010, 1–9. https://doi.org/10.1155/2010/468526 Zhao, G., Ford, E. S., Li, C., & Balluz, L. S. (2011). Physical activity in U.S. older adults with diabetes mellitus: Prevalence and correlates of meeting physical activity recommendations. Journal of the American Geriatrics Society, 59(1), 132–137. https://doi.org/10.1111/j.1532-5415.2010.03236.x Zhu, X., & Lee, C. (2008). Walkability and Safety Around Elementary Schools. American Journal of Preventive Medicine, 34(4), 282–290. https://doi.org/10.1016/j.amepre.2008.01.024       225            226 APPENDIX Appendix A   Survey Questionnaires A.1 Neighborhood Impact on Kids (NIK) Survey Transport Related Physical Activity V. Walk/Bike  How many days per week does your child usually walk or bike to/from the following (on their own or with an adult)?  Never Once a month or less Once every other week Once a week 2 or 3 times/week 4 times/ week or more 1. Indoor recreation or exercise facility (public or private) e.g. YMCA / boys and girls club 1 2 3 4 5 6 2. Friend or relative’s house 1 2 3 4 5 6 3. Outdoor recreation place (park, playground) 1 2 3 4 5 6 4. Food store or restaurant 1 2 3 4 5 6 5. Other: (please specify) ____________________________ 1 2 3 4 5 6  Leisure and Neighborhood Physical activity T. Physical Activity at Home & In the Neighborhood How often is your child PHYSICALLY ACTIVE (including active play) in the following places?   Never Once a month or less Once every other week Once a week 2 or 3 times/week 4 times/ week or more 1. Inside your home 1 2 3 4 5 6 2. In your yard or common area 1 2 3 4 5 6 3. In your driveway or alley 1 2 3 4 5 6 4. At a neighbor’s house, yard or driveway 1 2 3 4 5 6 5. In a local street, sidewalk, or vacant lot 1 2 3 4 5 6 6. In a nearby cul-de-sac or dead end street 1 2 3 4 5 6    227 Parent support for physical activity Y. Friends and Family  During a typical week, how often does your child sit and watch TV, play videogames, on the computer, or with other electronic devices with…  Never 1-2 days 3-4 days 5-6 days Everyday 1. Siblings (if no siblings, circle ‘never’) 1 2 3 4 5 2. A parent/guardian/caregiver 1 2 3 4 5 3. Friends 1 2 3 4 5 During a typical week, how often have you or another adult in the household: 4. Watched your child participate in physical activity or sports 1 2 3 4 5 5. Encouraged your child to do sports or physical activity 1 2 3 4 5 6. Provided transport to a place where your child can do physical activity or play sports 1 2 3 4 5 7. Done a physical activity or played sports with your child 1 2 3 4 5 During a typical week how often do your child’s siblings or friends: 8. Do physical activity or play sports with your child 1 2 3 4 5   228 9. Ask your child to walk or bike to school or to a friend’s house 1 2 3 4 5  Parents perception of the neighborhood safety FF. Getting Around In Your Neighborhood Please circle the answer that best applies to you and your neighborhood.  Both local and within walking distance means within a 10-15 minute walk from your home.  Strongly disagree Somewhat disagree Somewhat agree Strongly agree  1. Parking is difficult in shopping areas. 1 2 3 4 2. Streets in my neighborhood are hilly, making it difficult to walk. 1 2 3 4 3. There are not many dead end streets.  1 2 3 4 4. There are many different routes for getting from place to place.   1 2 3 4 5. There are sidewalks on most streets.  1 2 3 4 6. Sidewalks are separated from the road/traffic by parked cars. 1 2 3 4 7. There is grass/dirt between the streets and the sidewalks. 1 2 3 4 8. There are trees along the streets. 1 2 3 4 9. There are many interesting things for my child to look at while walking. 1 2 3 4 10. There are many beautiful natural things for my child to look at. 1 2 3 4 11. There are many buildings/homes that are nice for my child to look at. 1 2 3 4 12. The traffic makes it difficult or unpleasant for my child to walk.  1 2 3 4 13. The speed of traffic on most streets is usually slow (30 mph or less).  1 2 3 4 14. Most drivers go faster than the posted speed limits. 1 2 3 4 15. I’m afraid of my child being taken or hurt by a stranger in a local park. 1 2 3 4 16. Streets have good lighting at night. 1 2 3 4 17. Walkers and bikers can be easily seen by people in their homes.  1 2 3 4   229 18. There are crosswalks and signals on busy streets.  1 2 3 4 19. There is a high crime rate. 1 2 3 4 20. I’m afraid of my child being taken or hurt by a stranger on local streets. 1 2 3 4 21. I’m afraid of my child being taken or hurt by a stranger in my yard, driveway, or common area. 1 2 3 4 22. I’m afraid of my child being taken or hurt by a known “bad” person (adult or child) in my neighborhood 1 2 3 4  Socio-demography LL. General information  Parent’s Demographics 7. Your Age: _________   8. Gender:   Male       Female   9. Are you of Hispanic, Mexican, or Latino ethnicity?  Yes         No   10. Your Race (you can check one or more):   Caucasian   African-American or Black  Asian-American   Pacific Islander   American Indian or Alaskan Native   Other _____________   11. What was your highest education level you completed? (please check one).   Less than 7th grade   Junior high/middle school   Some high school   Completed high school   Some college or vocational training   Completed college or university   Completed graduate or professional degree  12. How many hours per week do you (or your child’s primary caregiver) work outside of the home?  None or less then part time (0-15 hours)  Part time (15-35 hours)  Full time (35+ hours)    230 NIK Child’s Demographics  13. Child’s Birth Date:    _______ _______ _______     Month  Day  Year  14. Child’s birth weight: _______ lbs. _______ oz   15. Gender:   Male         Female   16. Is your child of Hispanic, Mexican, or Latino ethnicity?  Yes           No   17. Child’s Race (you can check one or more):   Caucasian   African-American or Black  Asian-American   Pacific Islander   American Indian or Alaskan Native   Other _____________   18a. Was this child breastfed (either directly or given breastmilk in a bottle) during infancy/young childhood?   Yes           No   18b. If ‘yes’, how long was this child given mostly breastmilk (not formula or other milk products) when consuming milk?  <1 week  1-3 weeks  1-3 months  4-6 months  6 months – 1 year  >1-2 years  >2 years  Household Information   19. How many people (including yourself) live in your household? _______ people  20. How many children under 18 live in your household? ________   21. What are the ages and gender (select one) of all children living in your household?  a) _________   male / female    b) _________   male / female  c) _________   male / female     d) _________   male / female    e) _________   male / female   f) _________   male / female      22a. Are you currently pregnant (or if you are the father/paternal caregiver, is this child’s mother/maternal caregiver pregnant)?   Yes           No 22b. If ‘yes’, when is the baby’s expected due date? _______  _______  _______   231                  Month        Day         Year   23. What is the highest level of education among the adults in your household?  Less than 7th grade   Junior high/middle school   Some high school   Completed high school   Some college or vocational training   Completed college or university   Completed graduate or professional degree   24. What type of residence do you live in? (please check one).   Single family house   Multi-family house  Apartment  Condominium/townhouse  Other _______________   25. Do you rent or own your home?  Own/buying  Rent   26. Do you have a valid driver’s license?    Yes    No   27. How many driveable motor vehicles (cars, trucks, motorcycles) are there at your household? _______   28. How many licensed drivers are in your household (including yourself)? _______     29. What is your marital status? (please check one).    Married    Widowed/divorced/separated    Single and never married    Living with partner   30. Approximate annual household income (please check one)   <$10,000     $60,000-$69,000  $10,000-$19,000   $70,000-$79,000  $20,000-$29,000   $80,000-$89,000  $30,000-$39,000   $90,000-$99,000  $40,000-$49,000    > $100,000  $50,000-$59,000     232   A.2 Teen Environment and Neighborhood Study (TEAN) Survey  Transport Related Physical Activity  Leisure and neighborhood physical activity   8. Watch out for cars Yes No 9. Check in frequently Yes No 10. Stay on paths, trails or sidewalk Yes No 11. Do not cross busy streets Yes No 12. Wear hat and/or sunscreen in summer Yes No 13. Do n t fight with o her kids Yes No 14. Do not disrespect others (particularly adults) Yes No   AB.  Walking and Biking:  Remember, think about the PAST YEAR. How often do you usually walk or bike to/from the following?  Never Once a month or less Once every other week Once a week 2 or 3 times per week 4 or moretimes per week  1. Indoor recreation or exercise facility (public or private; YMCA, Boys & Girls Club, dance, martial arts) 0 1 2 3 4 5 2. Friend’s or relative’s house 0 1 2 3 4 5 3. Outdoor recreation place (park, sports field, open space, creek) 0 1 2 3 4 5 4. Food store or restaurant/cafe 0 1 2 3 4 5 5. Other retail stores (e.g., music, clothes) 0 1 2 3 4 5 6. Non-school social or educational activities (e.g., church group, band) 0 1 2 3 4 5 7. Public transportation stop (bus, train, light rail) 0 1 2 3 4 5 8. Work  (check if not applicable □  ) 0 1 2 3 4 5 9. Other: (please specify) ____________________________ 0 1 2 3 4 5 10. How often do you skateboard to go places?           0   Never        1   Once  a month   or less            2      Once  every other       week         3  Once a     week             4      2 or 3        times    per week 5 4 or more  times per  week   17S.   Physical Activity Outside of School   1. Over the past seven days, on how many days were you physically active for a total of at least 60 minutes per day (do not include school PE or gym class)? 0 days 1 day 2 days 3 days 4 days 5 days 6 days 7 days  2. Over a typical or usual week, on how many days are you physically active for a total of at least 60 minutes per day (do not include school PE or gym class)? 0 days 1 day 2 days 3 days 4 days 5 days 6 days 7 days 3. In the past year, how many sports teams or physical activity classes have you participated in outside of school? If you play for more than 1 team of the same sport or across 2 seasons (e.g.,  two softball leagues), count this as 2.               0                    1                          2                      3                      4 or more   T.  Places for Physical Activity Near Your Home How often are you PHYSICALLY ACTIVE in/at the following places?   Never Once a month o