UBC Theses and Dissertations

UBC Theses Logo

UBC Theses and Dissertations

Environmental attitudes and travel behavior of youth Tse, Yat Ching Cindy 2012

You don't seem to have a PDF reader installed, try download the pdf

Item Metadata

Download

Media
ubc_2012_fall_tse_cindy.pdf [ 4.3MB ]
[if-you-see-this-DO-NOT-CLICK]
Metadata
JSON: 1.0073322.json
JSON-LD: 1.0073322+ld.json
RDF/XML (Pretty): 1.0073322.xml
RDF/JSON: 1.0073322+rdf.json
Turtle: 1.0073322+rdf-turtle.txt
N-Triples: 1.0073322+rdf-ntriples.txt
Original Record: 1.0073322 +original-record.json
Full Text
1.0073322.txt
Citation
1.0073322.ris

Full Text

  Environmental Attitudes and Travel Behavior of Youth  by  Yat Ching Cindy Tse  BASc., The University of British Columbia, 2007   A THESIS SUBMITTED IN PARTIAL FULFILLMENT OF THE REQUIREMENTS FOR THE DEGREE OF  MASTER OF APPLIED SCIENCE in The Faculty of Graduate Studies  (Civil Engineering)  THE UNIVERSITY OF BRITISH COLUMBIA (Vancouver)  October 2012  ©  Yat Ching Cindy Tse, 2012    i  Abstract Environmental problems relating to transportation have increased in severity in many cities within Canada. Reduction in emissions generated from the transport sector will require far reaching technological advances as well as behavioral shifts by the population.  This study is carried out with a focus on influencing youths to increase the use of sustainable transportation. Studying the attitudes of young people towards the environment is important given they will be the ones affected by environmental issues arising from our current actions and consequently will be the individuals who will ultimately have to come up with resolutions. In preparation to address future transportation issues, it is critical to conduct researches with a focus on children and young people as attitudes toward the environment begin to develop in childhood.  This study surveyed over 1000 students from 11 secondary schools in Richmond, Canada and Vancouver, Canada.  The purpose of the survey is to examine the relationship between environmental knowledge, environmental attitudes and travel behavior.  Structural equation models (SEMs) were used to quantify environmental attitudes as latent variables, as well as to examine the relationship between environmental knowledge, attitudes and travel behavior while controlling demographic and socioeconomic variables.  The results indicate that 1) Students who have better knowledge on environmental issues hold higher levels of pro- environmental attitudes; 2) Students who use active transportation, public transit and school buses to/from school have stronger pro-environmental attitudes than students who travel to/from school by car; 3) Environmental attitude is found to be a mediating variable between environmental knowledge and travel behavior; and 4) Environmental knowledge has a significant impact on the relationship between environmental attitudes and travel behavior.    Based on these findings, it is evident that continual improvement of the environmental education curriculum is needed. It is recommended that the curriculum should not be limited to knowledge-based education on transportation.  Rather, it should include activities that will help induce a positive environmental attitude.  It was also found that social media and school ii  courses are major sources of environmental information for the students.  Therefore, educational institutes, government, environmental NGOs should continue to leverage these channels to effectively disburse information.            iii  Preface The thesis research received an approval of the University of British Columbia Behavioural Research Ethics Board (UBC BREB NUMBER H11-00467) on January 6, 2012.   iv  Table of Contents Abstract ............................................................................................................................................ i Preface ............................................................................................................................................ iii Table of Contents ............................................................................................................................ iv List of Tables ................................................................................................................................. viii List of Figures ................................................................................................................................... x Acknowledgements ........................................................................................................................ xii Dedication ..................................................................................................................................... xiii 1 Introduction ............................................................................................................................. 1 1.1 Motivation ........................................................................................................................ 1 1.2 Significance ....................................................................................................................... 1 1.3 Research Objectives ......................................................................................................... 3 1.4 Research Approach .......................................................................................................... 4 1.5 Thesis Organization .......................................................................................................... 4 2 Literature Review .................................................................................................................... 7 2.1 The Concept of Environmental Knowledge, Attitudes and Travel Behavior ................... 7 2.1.1 Environmental Knowledge ......................................................................................... 7 2.1.2 Environmental Attitudes ............................................................................................ 9 2.1.3 Travel Behavior ........................................................................................................ 11 2.2 Psychological Models Linking Knowledge, Attitudes and Behavior ............................... 12 2.2.1 The Linear Model ..................................................................................................... 12 2.2.2 Theory of Planned Behavior .................................................................................... 13 2.3 The Casual Link between Attitudes and Behavior ......................................................... 15 2.3.1 Experimental Designs ............................................................................................... 15 2.3.2 Instrumental Variable .............................................................................................. 16 2.3.3 Granger Causality Test ............................................................................................. 17 2.4 Research on Environmental Knowledge, Attitudes and Behavior of Youth .................. 17 2.5 Previous Research on Environmental Knowledge, Attitudes and Travel Behavior ....... 19 2.6 Mediating, Moderating and Confounding Effect ........................................................... 20 v  2.7 Analytical Framework ..................................................................................................... 21 3 Data and Method ................................................................................................................... 22 3.1 Survey Design and Implementation ............................................................................... 22 3.1.1 Study Area ................................................................................................................ 22 3.1.2 Survey Implementation ........................................................................................... 25 3.1.3 Questionnaire Design ............................................................................................... 29 3.2 Data Entry and Data Screening ...................................................................................... 34 3.3 Data Validity ................................................................................................................... 35 3.4 Analytical Methods ........................................................................................................ 38 3.4.1 Exploratory and Confirmatory Factor Analysis ........................................................ 39 3.4.2 Structural Equation Modeling .................................................................................. 40 3.4.3 Multiple-Indicator Multiple-Cause (MIMIC) Analysis .............................................. 43 3.4.4 Multiple Group Analysis (MGA) ............................................................................... 43 3.4.5 SEM and its Application in Transportation .............................................................. 43 4 Environmental Knowledge, Attitudes and Travel Behavior .................................................. 45 4.1 Demographic and Socioeconomics Information ............................................................ 45 4.2 Environmental Knowledge ............................................................................................. 47 4.3 Environmental Attitudes ................................................................................................ 52 4.3.1 Factor Analysis ......................................................................................................... 57 4.3.2 Determinants of Environmental Attitudes .............................................................. 61 4.4 Travel Behavior ............................................................................................................... 64 4.5 Comparison between In-class and Lunch Time survey .................................................. 71 5 The Impact of Environmental Knowledge on Environmental Attitudes ............................... 76 5.1 Introduction.................................................................................................................... 76 5.2 Model Structures and Estimations ................................................................................. 76 5.3 Model Results ................................................................................................................. 77 5.4 Summary ........................................................................................................................ 84 6 Disentangle Casual Direction between Environmental Attitudes and Travel Behavior ........ 85 6.1 Introduction.................................................................................................................... 85 vi  6.2 Comparison of Travel Behavior between PARENT-DECISION-MAKER and STUDENT- DECISION-MAKER ...................................................................................................................... 86 6.3 The Impact of Travel Behavior on Environmental Attitudes ......................................... 89 6.4 Summary ...................................................................................................................... 101 7 The Relationship between Environmental Knowledge, Attitudes and Travel Behavior ..... 103 7.1 Introduction.................................................................................................................. 103 7.2 The Mediating Effect of Environmental Attitudes on the Relationship between Environmental Knowledge and Travel Behavior ..................................................................... 105 7.3 The Confounding Effect of Environmental Knowledge on the Relationship between Environmental Attitudes and Travel Behavior ........................................................................ 107 7.4 The Moderating Effect of Environmental Knowledge on the Relationship between Environmental Attitudes and Travel Behavior ........................................................................ 109 7.5 Summary ...................................................................................................................... 110 8 Environmental Education Program and Policies ................................................................. 111 8.1 Environmental Education Curriculum in British Columbia .......................................... 111 8.2 Interviews ..................................................................................................................... 114 8.3 Environmental Information Acquisition ....................................................................... 116 8.4 Summary ...................................................................................................................... 117 9 Discussion and Conclusions ................................................................................................. 119 9.1 Major Findings .............................................................................................................. 119 9.1.1 Environmental Knowledge Affect Environmental Attitudes ................................. 120 9.1.2 Travel Behavior Affect Environmental Attitudes ................................................... 121 9.1.3 The Relationship between Environmental Knowledge, Attitudes and Travel Behavior………….………….……………..……………………………………………………………..……………………122 9.2 Research and Policy Implications ................................................................................. 122 9.3 Research Limitations and Future Research .................................................................. 124 References .................................................................................................................................. 126 Appendices ................................................................................................................................ ..136 Appendix A: Application Package to School Board………………………………………………………………137  vii  Appendix B: One way ANOVA of Environmental Knowledge and Attitudes by socioeconomic        status …………………………………………………………………………………………………………………………………167 Appendix C: Measurement Equations for the Latent Variables ……………………….……..………….172 Appendix D: Interview Questions……………………………………………………………………………………….175 Appendix E: Sample MPlus Code………………………………………………………………………………………..178      viii  List of Tables Table 3.1 Demographic and Transportation Statistics for the Greater Vancouver, City of Vancouver and City of Richmond ................................................................................................. 24 Table 3.2 Implementation Method and Number of Responses by Schools ................................. 29 Table 3.3 Survey Comparison ....................................................................................................... 30 Table 3.4 Survey Component ........................................................................................................ 31 Table 4.1 Respondent and Household Characteristics ................................................................. 46 Table 4.2 Comparison between Samples and Average Household .............................................. 47 Table 4.3 Environmental Knowledge Questions and Responses .................................................. 51 Table 4.4 Responses to Environmental Attitudes Statement ....................................................... 56 Table 4.5 Goodness of Fit Results for Confirmatory Factor Analysis (1st model) ........................ 58 Table 4.6 Goodness of Fit Results for Confirmatory Factor Analysis (2nd model) ....................... 59 Table 4.7 Correlation Matrix ......................................................................................................... 60 Table 4.8 Goodness of Fit Results for Confirmatory Factor Analysis (3rd model) ....................... 60 Table 4.9 Confirmatory Factor Analysis Results for Environmental Attitudes ............................. 61 Table 4.10 Travel Alone to School by Age and Sex ....................................................................... 67 Table 4.11 Percentages of Frequent Use of Modes of Transport ................................................ 70 Table 4.12 Travel Mode to Work vs Travel Mode to School ........................................................ 70 Table 5.1 Goodness of Fit of Statistics for SEM ............................................................................ 78 Table 5.2 Measurement Equations for the Latent Variables ........................................................ 78 Table 5.3 Structural Equations between Environmental Knowledge and Attitudes .................... 79 Table 5.4 Structural Equations for the Determinants of Environmental Knowledge and Attitudes ....................................................................................................................................................... 81 Table 5.5 Direct and Indirect Effects on General Environmental Attitudes and Transportation- Specific Environmental Attitudes ................................................................................................. 82 Table 5.6 Comparison of Explanatory Power between Models ................................................... 83 Table 5.7 Structural Equations for the Auxiliary Model ............................................................... 83 Table 6.1 Frequent Use of Modes of Transport by STUDENT-DECISION-MAKER......................... 87 Table 6.2 Percentages of Frequent Use of Modes of Transport by PARENT-DECISION-MAKER .. 88 ix  Table 6.3 Comparison of Travel Mode Share ............................................................................... 88 Table 6.4 Model Structure ............................................................................................................ 89 Table 6.5 Goodness of Fit of Statistics – The Impact of Travel Behavior on Environmental Attitudes........................................................................................................................................ 92 Table 6.6 Structural Equations between Environmental Attitudes and Travel Mode to School . 93 Table 6.7 Direct and Indirect Effects on Environmental Attitudes via Travel Mode to School .... 94 Table 6.8 Structural Equations between Environmental Attitudes and Travel Mode from School ....................................................................................................................................................... 95 Table 6.9 Direct and Indirect Effects on Environmental Attitudes via Travel Mode after School 97 Table 6.10 Structural Equations between Environmental Attitudes and Travel Mode Frequency ....................................................................................................................................................... 98 Table 6.11 Direct and Indirect Effects on ENVATT via Travel Mode Frequency ........................... 99 Table 6.12 Structural Equations between Environmental Attitudes and Mode Share .............. 100 Table 6.13 Direct and Indirect Effects on Environmental Attitudes via Travel Mode Frequency ..................................................................................................................................................... 101 Table 7.1 Goodness of Fit of Statistics –Environmental Attitudes as Mediating Variable ......... 105 Table 7.2 Structural Equations – Environmental Attitudes as Mediating Variable .................... 106 Table 7.3 Goodness of Fit of statistics –Environmental Knowledge as Confounding Variable .. 107 Table 7.4 Structural Equations – Environmental Knowledge as Confounding Variable ............. 108 Table 7.5 Environmental Knowledge Levels ............................................................................... 109 Table 7.6 Structural Equations – Environmental Knowledge as Moderating Variable .............. 110    x  List of Figures Figure 1.1 2008 Canadian Transportation Greenhouse Gas Emissions by Mode .......................... 2 Figure 1.2 Thesis Organization ........................................................................................................ 6 Figure 2.1 Linear ‘deficit’ model of pro-environmental behavior ................................................ 13 Figure 2.2 Icek Ajzen’s Theory of Planned Behavior ..................................................................... 14 Figure 3.1 Secondary School Map of the Vancouver School District ........................................... 26 Figure 3.2 Secondary School Map of Richmond School District ................................................... 27 Figure 4.1 Number of Environmental Knowledge Questions Answered Correctly ...................... 48 Figure 4.2 Average levels of GENKNOW and TRANKNOW ........................................................... 49 Figure 4.3 Average levels of GENATT and TRANATT..................................................................... 62 Figure 4.4 Travel Mode to/from School........................................................................................ 65 Figure 4.5 Activities after School .................................................................................................. 66 Figure 4.6 Travel Time to/from School ......................................................................................... 66 Figure 4.7 Factors that Influence the Travel Mode Choice to School .......................................... 68 Figure 4.8 Gender by survey type ................................................................................................. 71 Figure 4.9 Grade by survey type ................................................................................................... 72 Figure 4.10 Environmental Knowledge by survey type ................................................................ 73 Figure 4.11 Environmental Attitudes by survey type ................................................................... 73 Figure 4.12 Travel Mode to/from School by survey type ............................................................. 74 Figure 4.13 Decision Maker by survey type .................................................................................. 75 Figure 5.1 Specification of the Structural Equation Model .......................................................... 77 Figure 6.1 Comparison of Travel Mode to School ........................................................................ 86 Figure 6.2 Comparison of Travel Mode from School .................................................................... 87 Figure 7.1 The Mediating Effect of Environmental Attitudes on the Relationship between Environmental Knowledge and Travel Behavior......................................................................... 103 Figure 7.2 The Confounding Effect of Environmental Knowledge on the Relationship between Environmental Attitudes and Travel Behavior............................................................................ 104 Figure 7.3 The Moderating Effect of Environmental Knowledge on the Relationship between Environmental Attitudes and Travel Behavior............................................................................ 104 xi  Figure 8.1 Source of Environmental Information ....................................................................... 117    xii  Acknowledgements The completion of this thesis marks the end of an exciting and enlightening academic journey.  I would like to acknowledge the many people who made contributions along the way. My thanks go first to my supervisor, Dr. Jinhua Zhao, who has consistently supported, guided, challenged, and helped me through all the research projects that I worked on.  Dr. Zhao’s high expectations and attention to detail encouraged me to give my best throughout this process.  I would also like to thank Dr. Brad Flamm for a very thoughtful review of the thesis.    This research would not be possible without the principals and teachers support, as well as the students who spent time in filing the surveys.  Special thanks to the research team for survey implementation and data input.  I would also like to express my appreciation to the University of British Columbia for giving me the opportunity to complete this Master's program which has been unbelievably effective to me. My very sincere thanks go to my former colleagues.  They have evoked my interests in transportation planning, guided me intellectually, socially and personally during and after my work with them. Many people supported me personally throughout this journey including my family and friends who showed interest and encouraged me in my academic pursuits.  Countless thanks go to my grandmother who always cooks me soup to keep me healthy during my study.  Finally, I want to thank my parents who first introduced me to the academic life and encouraged me to always put forward my best.      xiii  Dedication                   To my family    1  1 Introduction 1.1 Motivation In early 2009, Mayor Gregor Robertson formed the Greenest City Action Team with the goal to make Vancouver the world’s Greenest City by 2020.  The group of community leaders who hold knowledge and experience across a range of the most pressing environmental issues Vancouver faces researched best practices from leading green cities around the world.  Goals and targets have been established that will help Vancouver to accomplish its goal to be the Greenest City.  Amongst the goals and best practices, green transportation is an area where the most people can take part in, which includes transit, as well as active transportation like cycling and walking.  This is the initiative that motivated this research on environmental attitude and travel behavior. How we move around a city makes a big difference to our quality of life. The air we breathe, the amount of land we need, our physical health and well-being, and the cost of travel are all impacted by our transportation choices.  Transportation is also about the places we see and experience on the way to our destinations.  To achieve the Green Transportation goal, we will need to transform Vancouver into a city where moving on foot or by bike is safe, convenient, and enjoyable.  “Transit should be fast, frequent, reliable, accessible, and comfortable, getting you where you need to go when you need to get there.  Streets, public spaces, and neighborhoods should be vibrant places that are alive with people, plants, and activities” (City of Vancouver 2012).  Aside from transportation infrastructure improvements, intervention opportunities identification is also important.  Understanding where youth go, how they get there, how much time they spend, what they do there and what influences their allocation of time and activities are areas of interest that influence travel behavior. 1.2 Significance  Personal transportation decisions made by Canadians impact the environment through emissions of pollutants, greenhouse gases as well as use of land for roads and streets.  According to Environment Canada (2010) , the transportation sector is one of the biggest sources of emissions in Canada, which produced 22 percent (162Mt) of total greenhouse gas 2  (GHG) emissions (734Mt) in 2008.  Of the 162Mt of emissions resulting from transportation sources, 54.6 percent of that amount was emitted by passenger cars and light trucks (Figure 1.1).  GHG emissions from private motor vehicles have gone up 35% from 1990 to 2007, almost twice the growth rate of Canada’s population (19%) during the same period (Statistics Canada 2010). Figure 1.1 2008 Canadian Transportation Greenhouse Gas Emissions by Mode  The Government of Canada is making progress towards the target of reducing Canada’s total GHG emissions by 17 per cent from 2005 levels by 2020.    The Government of Canada recognizes that reducing emissions from the transportation sector can achieve its climate change goals.  As a result, tougher regulations for cars and light trucks were introduced to reduce emissions from these vehicles by 25 percent.   Environmental professionals argue that while technical improvements can decrease pollution per vehicle, this change alone cannot resolve the problem caused by continuously increasing car production and use (Greening, Greene et al. 2000; Isenhour 2010).  Reducing the emissions from the transport sector will require far reaching technological as well as behavioral shifts (Bristow, Pridmore et al. 2004; Anable, Lane et al. 2006; Banister and Hickman 2006).     3  This study is carried out in light of the idea of influencing youth to increase the use of sustainable transportation at the present and in the future.  Sustainable transport here refers to any means of transport with low impact on the environment including but not limited to walking, cycling, transit and carpooling.   Young people are recognized as an important group for the development of a sustainable environment (The United Nations Conference on Environment and Development 1993).  Studying young people’s attitudes toward the environment is particularly important as they are the ones who will be affected and will have to provide resolutions to the environmental problems arising from our current actions. As attitudes towards the environment begin to develop in childhood, it may be critical to focus on children and young people in order to address transportation issues in the future (Kopnina 2011).  Young people develop their attitudes based on what they have learned and experienced through their behavior.  These attitudes may sustain themselves into the future and exert impact on behavior as children grow up and become decision makers of their action.   Having a better understanding of the relationship between environmental knowledge, environmental attitudes and travel behavior provide important insights into how educational institutes, the government and environmental NGOs can develop and improve environmental education programs in connection to travel behavior.  The ultimate goal of the study of environmental attitudes and young people is to influence students toward sustainable behavior in the future. 1.3 Research Objectives This research aims to  1) Benchmark environmental knowledge and environmental attitudes among youth 2) Explore the differences between general and transportation-specific environmental knowledge and environmental attitudes 3) Examine the impact of environmental knowledge on environmental attitudes 4) Disentangle the causal direction between environmental attitudes and travel behavior 4  5) Examine the relationship between environmental knowledge, environmental attitudes and travel behavior   6) Explore the environment educational program and policies in British Columbia 1.4 Research Approach In an attempt to achieve the stated objectives, the first step is to complete a literature review of models and theories that can be used to examine the linkage between knowledge, attitudes and behavior.  After a comprehensive literature review, the analytical framework for this study is developed to test the complex interrelationship among the variables. The survey is then designed and conducted in 11 high schools in the City of Richmond and the City of Vancouver.  To conduct analysis of survey results, methods used in behavioral science research have been reviewed.  Based on the sample size, structure, and detail of the available data, structural equation modeling is chosen as the appropriate analytical method for this study.  The model results are summarized so that recommendations can be developed to improve the environmental education programs in connection to travel behavior.   1.5 Thesis Organization The thesis is organized as follows (see Figure 1.2): Chapter 1 provides an introduction to this thesis.  Chapter 2 explores previous research on knowledge, attitudes and behavior linkages.  It also describes the methods used to test the relationships between these factors.  Finally, it introduces the analytical framework, followed by research questions and study hypotheses.  Chapter 3 explains the study methodology, sampling method, survey design and distribution method used to collect data and models used to conduct the analysis.  It also describes the procedures used to screen the data set and remove inconsistent records.  After, Chapter 4 provides the descriptive statistics of the sample population, followed by the analysis on three key aspects of this research: environmental knowledge, environmental attitudes and travel behavior.  Chapter 5 presents a model examining the impact of environmental knowledge and environmental attitudes.  Chapter 6 is devoted to the findings of the structural equation model developed specifically to explore the 5  directions of causality between attitudes and behavior.  Chapter 7 presents a series of structural equation models to examine the relationship between environmental knowledge, attitudes and travel behavior.  Chapter 8 presents a discussion on environmental educational programs and policies in British Columbia.  Chapter 9 summarizes the findings, discusses policy implications and suggests directions for future research.   6  Figure 1.2 Thesis Organization  Ch 1: Introduction: motivation, objective and approach    Ch 2: Literature Review   Ch 3: Data and Method   Ch 4: Survey Results                Ch 5: Environmental Knowledge and Attitudes                   Ch 6: Environmental Attitudes and Travel Behavior   Ch 7: Integrated Model of Environmental Knowledge, Attitudes and Travel Behavior   Ch 8: Environmental Educational Program and Policies   Ch 9: Discussion and Conclusions   7  2  Literature Review This chapter first introduces the three key concept of this research: environmental knowledge, environmental attitudes and travel behavior.  Then it provides an overview of the frameworks and theories used to examine the link between knowledge, attitudes and behavior.  In particular, it explores the different approaches to analyze the causality between attitudes and behavior.  Since no previous studies directly addressed the relationships between environmental knowledge, environmental attitudes and travel behavior on youth, this chapter reviews studies examining the relationships between environmental knowledge, attitudes and behavior of youth, followed by studies on environmental knowledge, attitudes and travel behavior of adults.  The analytical framework for this study is revealed at the end of this section.  2.1 The Concept of Environmental Knowledge, Attitudes and Travel Behavior 2.1.1 Environmental Knowledge Environmental knowledge reflects people’s understanding and ability to identify environmental issues.  It has been an interdisciplinary subject drawing relevant knowledge from natural and social sciences as well as humanities.  At times, it is also used to deal with moral decisions about values and powers (Hausbeck, Milbrath et al. 1992) Maloney and Ward (1973) compared members of a conservation group, a group of college students, and a non-randomly selected group of adults with no college experience from California.  They found that environmental knowledge was associated with conservation group membership and education.  Arcury et al. (1986) examined the relationship between environmental worldview, measured with the New Environmental Paradigm (Liere and Dunlap 1980), and general environmental knowledge. Worldview, income, education and gender were found to have positive influence on knowledge.  The NEP scale is designed by Dunlap and Van Liere (1978) to measure the extent to which people would accept the ideas of the New Environmental Paradigm which will be explained  in detail in the next section.   8  Blum (1987) found that high school students in four countries possessed low levels of environmental knowledge.  He compared five surveys conducted in the United States, Australia, England, and Israel that assessed environmental knowledge and beliefs of 9th grade and 10th grade students.  Results indicated that the students’ beliefs in environmental causes were generally stronger than their factual and conceptual knowledge.  He concluded that students’ level of factual and conceptual knowledge was rather low and that schools have much to do to improve the knowledge base of all students.  Hausbeck et al. (1992) surveyed  3,200 students from a sample of 30 secondary schools in the State of New York  to assess levels of environmental knowledge, environmental awareness, and environmental concern.  They found that environmental knowledge was positively correlated with age, family, income, gender, and exposure to mass media and personal sources outside of school.  They also found that students scored rather low on knowledge questions; however, the students reported that they would like additional environmental education to be offered in school.  The authors concluded with policy suggestions for improved environmental education at the secondary school level in New York State and the United States at large.    Kuhlemeier (1999) examined the environmental knowledge of more than 9000 students from 206 secondary schools in the Netherlands.  Results show large groups of students lacked knowledge on environmental topics such as energy usage; soil, air, and water pollution; recycling; agricultural activities; tourism; transportation; and recreation.  It was found that teachers seldom discussed with students the causes of environmental problems or potential solutions due to a lack of suitable teaching materials.  As a result, the environmental knowledge that students already possessed was likely to determine by topicality and to be fragmentary, incorrect, or both (Hausbeck, Milbrath et al. 1992; Munson 1994)   Despite the importance of environmental knowledge in environmental studies, there is a lack of consistency in measurement.  Some measured the environmental knowledge on global issues including greenhouse effect and acid deposition while others measured environmental 9  knowledge on local issues, such as energy and water supply.  There is a great variation in method of inquiry, types of questions, and scope of environmental issues.   In most of these studies, environmental knowledge was measured by asking participants to respond to multiple-choice or true-false questions (Chan 1996; Gambro and Switzky 1996; Bradley, Waliczek et al. 1999; Kuhlemeier, Bergh et al. 1999; He, Huo et al. 2004; He, Hong et al. 2011).  Other types of measurement of environmental knowledge employed include the use of a self-reported assessment by asking respondents to estimate their level of knowledge on specific topics (Arcury, Johnson et al. 1986).  In some other studies, respondents’ environmental knowledge was assessed based on responses to open-ended questions on selected global or local environmental issues.   2.1.2 Environmental Attitudes Environmental attitudes reflect people’s mental-psychological projection on environmental issues.  It could be considered as an enduring positive or negative feeling towards a particular aspect of an environment object or issue.  Environmental attitude studies have served an important psychological function in that they have helped people make decisions involving the use and care of the physical environment, such as a decision to put litter into waste recycling receptacles, to join environmental organizations and to use public transportation to school or work (1996).     Environmental attitudes are often discussed in the form of environmental concerns.  Grganon- Thompson and Barton (1994) described environmental concern as motivated either by a true care for the nature, or by a care for nature as a human resource.  Early research showed that environmental concern would determine actions promoting a sustainable environment (Arbuthnot 1977; Kallgren and Wood 1986; Stern and Oskamp 1987).  Another factor of environmental attitude is the perceived threat of environmental degradation.  However, previous studies suggest that environmental attitudes may also be jeopardized by perceptions of the quality of the local environment (Liere and Dunlap 1980; Blake, Guppy et al. 1997). 10  The most common approach to environmental attitude measurement has been the use of self- report scales on a number of local and global ecological issues and the scales have been concerned with interaction between human activities and natural resources.  Maloney and Ward (1973) developed a scale for the measurement of ecological attitudes that comprise of three subscales: verbal commitment, which measures what a person states he is willing to do in reference to pollution-environment issues; actual commitment, which measures what a personal actually does in reference to pollution environmental issues; and affection, which measures the degree of emotionality related to such issues.   Weigel and Newman (1976) developed a general attitude measure to assess people’s attitudes towards the environment, namely the environmental concern scale.  It considered 16 items that measured respondents’ attitudes towards a variety of ecological issues, such as pollution, conservation of natural resources and wildlife preservation.  The environmental concern scale was first tested with a random sample of 141 respondents in 1970 and achieved adequate internal consistency, with a Cronbach alpha coefficient of 0.88.  Shanahan  (1993) updated the environmental concern scale to measure more contemporary issues.  The updated version consisted of 17 items under four subscales including environmental optimism, the relative importance of environmental issues compared to economic and technological progress, attitudes toward specific environmental issues and personal impacts.   Dunlap and Van Liere (1978) designed the New Environmental Paradigm (NEP) scale.  It consisted of 12 items that measured respondents’ attitudes towards ecological issues on man’s influence on the balance of nature, limits to growth on the human population size and whether humans should have rightful dominion over nature.  The NEP scale was tested on a sample of 806 respondents and another sample of 407 members of an environmental organization.  The alpha coefficients for the general public and environmental group samples were 0.81 and 0.76, respectively, suggesting that the NEP scale was internally consistent.   Dunlap et al. (2000) revised the NEP scale to improve upon the original one as it provides more comprehensive coverage of key facets of an ecological worldview, avoids the unfortunate lack of balance in item direction of the original scale, and removes the outmoded, sexist terminology in some of 11  the original scale’s items.  The revised NEP scale has slightly more internal consistency than did the original version (Cronbach’s alpha of 0.83 vs 0.81).  Many studies directly used the developed scales, while others adapted the scales to better suit their local area condition, sample group and survey implementation (Chan 1996; Ivy, Road et al. 1998; Worsley and Skrzypiec 1998; Kuhlemeier, Bergh et al. 1999; Kilbourne, Beckmann et al. 2002).  2.1.3 Travel Behavior Travel behavior is the study of how people travel across distances and the mode at which they travel by.  The questions studied can be of a broad range.  Information on how many trips are made, origin and destination, mode choice, frequency and pattern, and intentions are usually captured via travel survey (Line, Chatterjee et al. 2010). In particular, behavior intention is an area of interest as it relates to environmental knowledge and environmental attitudes.   Intention is an important influence on actual behavior. If an individual does not intend to behave in a given way, it is unlikely they will carry out the behavior, even if they have the opportunity to do so (Line, Chatterjee et al. 2010).  People are more likely to express support for environmental behavior than they are to fulfill their promises. It has been suggested that verbal commitment to act more likely reflects the probable future than it does current behavior (He, Hong et al. 2011). In addition, responses are subject to the researcher’s choice of words when defining the questions.  It is important to recognize the need to examine the subjective, individualistic nature of such influences.    Previous research has shown that the travel behavior intentions of young people between the ages of 11 and 18 are dominated by the desire to drive and/or own a car in the future (Turner and Pilling 1999; Storey and Brannen 2000; Derek Halden Consultancy 2003).  At the age of 17, young people move from being predominantly car passengers to being car drivers.  While little research has been carried out to explore what determines young people’s intentions in this context, it can be assumed that not all will have the intention to drive when given the opportunity to do so.  At the same time, other young people may have the intention to drive 12  but no opportunity to do so, while others may have to create the opportunity to drive (by obtaining a job to finance driving lessons for example) in order to fulfill their intention.    The study by Line et al (2010) found that young people have a strong desire to continue driving or to learn to drive in the future and expressed an unwillingness to change their travel behavior intentions in light of climate change.  It was found that values act as the underlying influence on their travel behavior intentions.  Young people value identity, image and social recognition in particular.  It is also important to note that, although young people recognized that using alternative fuel vehicles or modes other than the car may be more ‘environmentally friendly’, they appear to place little value on ‘the environment’.  This suggests while pro-environmental values may exist, other values of higher perceived importance may take precedence and influence the behavior intention in the opposite direction.  A Scottish research study by Derek Halden Consultancy (2003) found that peer pressure was one of the most significant factors affecting young people’s travel mode choices.  For example, if cycling is not regarded as a socially-acceptable form of transport amongst a certain age group then it will be very hard to encourage more people to cycle.   2.2 Psychological Models Linking Knowledge, Attitudes and Behavior There has been considerable work done in trying to understand behavior by looking at the influences and processes involved in individuals’ decision-making.  They differ according to the degree to which they incorporate internal and external influences including wider social processes and opportunity to act (Anable, Lane et al. 2006). 2.2.1 The Linear Model The early models examining the connection between knowledge and action focused on humans as being rational.  These frameworks utilized a linear progression, in which environmental knowledge led to environmental attitudes, which in turn were thought to lead to pro- environmental behavior (Kollmuss and Agyeman 2002).   The early linear frameworks have been termed ‘deficit’ models of public understanding and action (Figure 2.1).  The assumption is that if people knew and understood more about connections between their own behavior and 13  environmental threats, they would act in a more sustainable way, thus responding more rationally to environmental risks.   Figure 2.1 Linear ‘deficit’ Model of Pro-Environmental Behavior Environmental Knowledge Environmental Attitude Pro- Environmental Behavior  (Source: Kollmuss and Agyeman 2002) This approach has been used in many behavior change programs, including large scale awareness-raising campaigns such as “Are You Doing Your Bit?” launched in 1998 in England and Wales (Department of the Environment Transport and the Regions 1999).  However, this approach is not without shortcomings.  Hounsham (2006), after reviewing the literature and consultation with experts on how to motivate green behavior, concluded that change from the provision of information alone is very minor.  He stated that “most of the lifestyle decisions we seek to influence are not determined mainly by rational consideration of the facts, but by emotions, habits, personal preferences, fashions, social norms, personal morals and values, peer pressure and other intangibles”.  Anable et al. (2006)  asserted that a large proportion of studies assume a linear link between attitudes and behavior without adequate consideration and measurement of other psychological and contextual factors that may shape behavior and the dynamic interrelationships between them.   2.2.2 Theory of Planned Behavior  Ajzen’s (1991) Theory of Planned Behavior is the most common and influential theory used to understand behavior by examining beliefs, values, concerns, and attitudes in the social, environmental and health psychology fields.  The theory predicts that attitudes do not determine behavior directly; rather, Ajzen suggested that attitudes toward a given behavior, subjective norms about that behavior, and perceived level of control over the behavior all contribute to behavioral intentions.  Behavioral intentions combined with an individual’s perception of control produce the actual behavior (Figure 2.2). 14  Figure 2.2 Icek Ajzen’s Theory of Planned Behavior  Many studies have used the theory of planned behavior to predict mode choice.  These studies have generally concluded that the choice of travel mode is largely a reasoned decision related particularly to attitudes and perceived barriers to behavior (Bamberg and Schmidt 1998; Forward 1998).  However, other studies suggest that much of people’s daily travel mode choices are habitual and not always preceded by the deliberation of alternatives.  The addition of habit measure will improve the predictive capability of attitude-behavior studies (Lanken, Aarts et al. 1994; Gärling 1998; Bamberg, Ajzen et al. 2003).  Knowledge is another important variable to the modification of the theory of planned behavior.  Beliefs are strongly influenced by a person’s values and are dependent to some degree on knowledge.  This knowledge may also determine which beliefs are salient and establish the value of the beliefs.   While psychologists often consider knowledge a necessary precursor to attitudes, it is not included in Ajzen’s original model.    Kaiser et al. (1999) have made it an integral part of their adapted theory of planned behavior, and stated that attitudes include not just the evaluation of a certain outcome but also the estimation of the likelihood of the outcome, with factual knowledge being a necessary pre-condition for attitude.   This provides the theoretical grounding for the use of a conceptual model that combines knowledge, attitudes and behavior in a single framework.   15  2.3 The Casual Link between Attitudes and Behavior Both the linear model and the theory of planned behavior assume that attitudes influence behavior directly or through behavioral intentions.  However, there are competing hypotheses regarding the relationship between attitudes and behavior (Tardiff 1977).  Attitudes are formed through experience as a result of behavior, while attitudes prompt certain types of behavior (Kitamura, Mokhtarian et al. 1997) .  Many studies have found mutual causality between attitudes and behavior (Tardiff 1977; Dobson, Dunbar et al. 1978; Golob, Horowitz et al. 1979; Tischer and Phillips 1979; Reibstein, Lovelock et al. 1980; Lyon 1984; Pendyala 1998; Handy, Cao et al. 2005). Golob (2001) tested a series of joint models of attitude and behavior to explain how both mode choice and attitudes regarding tolled high-occupancy vehicle lanes in San Diego differ across the population.  The sample of approximately 1,500 individuals was divided into three groups:  one-third ExpressPass subscribers, former subscribers, and persons on the waiting list; one- third other I-15 commuters, and one-third commuters in another freeway corridor in the San Diego Area.  None of the models tested found any significant effects of attitude on choice; all casual links were from behavior to attitudes.  Tardiff (1977) came to similar findings that the effect of attitudes on behavior to be weaker than that of behavior on attitudes.  He further hypothesized that if attitudes really do cause behavior, the policy emphasis might be on a marketing campaign to change attitudes. However, if the attitudes were merely a response to the behavior, a marketing campaign would have little lasting effect if the conditions of the transportation mode in question were not conducive to a favorable attitude. In this case, an emphasis on physical modal quality would appear to be a more fruitful policy.   There are several ways of assessing if attitudes cause behavior or vice versa: experimental designs, instrumental variable method and granger causality test. 2.3.1 Experimental Designs  Causal relationships are most validly established through experimental designs, in which individuals are randomized to treatment and control groups.  This addresses non-spuriousness 16  and behavior is measured for both groups before and after the treatment of interest (Singleton and Straits 1999).  It is assumed that the experimental and control group are equivalent on all important dimensions and that there are no systematic differences between the two groups.  The experimental group receives the influence of the independent variable (treatment) in addition to the shared conditions of the two groups.  All other conditions are under control to assure changes (if any) are observed in the values of the dependent variable.  By holding all other variables constant, a controlled experimental design can help determine if there is a difference in attitude from the introduction of a built environment versus no change in a built environment.   2.3.2 Instrumental Variable When controlled experiments are not available, instrumental variable can be used to estimate causal relationships.  The method of instrumental variables can be used when standard regression estimates of the relation of interest are biased due to reverse causality, selection bias, measurement error, or the presence of unmeasured confounding effects.  A third variable, also known as an instrumental variable can be introduced to estimate the causal relationship by extracting variation in the variable of interest.  The instrument correlates with the endogenous explanatory variable where it can be used to validate causality exists in the expected direction.  In the example of causality between attitude and behavior related to transportation, if a variable can be identified to only affect behavior, then the correlation between the introduced variable and attitude is evidence that behavior is causal to attitude, assuming there is no reverse causality.  It should also be noted that in most transportation situations, the structure of relationships among variables is not as simple as having a three variable model, namely attitude, behavior and one antecedent variable.  A more realistic approach might be to include a set of antecedent variables as opposed to only one antecedent variable.  Tradiff (1977)  estimated various models in which attitudes and behavior are jointly dependent on a third set of antecedent variables, e.g. personal and situational descriptors.  He suggested that if attitudes and behavior are jointly determined by antecedent variables and only correlated spuriously, a proper policy to change behavior would emphasize changes in the antecedent variables.  17  2.3.3 Granger Causality Test Another method is the Granger causality test, which is a statistical hypothesis test for determining whether one time series is useful in forecasting another.  A time series X is said to Granger-cause Y if it can be shown, usually through a series of t-tests and F-tests on lagged values of X, that those X values provide statistically significant information about future values of Y.  Vector autoregression (VAR) in particular is a statistical model used to capture the linear interdependencies amount multiple time series.  All the variables in a VAR are treated symmetrically; each variable has an equation explaining its evolution based on its own lags and the lags of all the other variables in the model.  “One approach to the causality question is to examine changes in attitudes and behavior and to inspect the time sequence. A study that monitors attitudes at one time and behavior at a second can be used to assess the impact of antecedent attitudes on subsequent behavior” (Reibstein, Lovelock et al. 1980).  For example, Assael and Day (1968) discovered that aggregate brand attitude change tended to lead to changes in brand sales over time.  In application related to transportation, the hypothesis testing can be run against attitude and behavior individually in attempts to find out the causality relationship. 2.4 Research on Environmental Knowledge, Attitudes and Behavior of Youth A key debate in the environmental education research literature revolves around the relationship between environmental knowledge, attitudes and behavior.  In order to have positive environmental attitudes, one would expect that an individual must first have the relevant knowledge to hold particular attitudes (Arcury 1990).   Ramsey and Rickson (1976) tested the relationship between environmental knowledge and attitudes among students at three high schools in Wisconsin and found increased knowledge leads to increased awareness of environmental issues. Bradley et al. (1999) came to a similar finding after assessing 475 Texas high school students’ environmental knowledge and attitudes before and after exposure to a 10-day environmental science course. The results indicated a significant increase in both knowledge and attitudes of students who took the course. A 18  significant relationship was found between pretest knowledge scores and pretest attitudes scores and between posttest knowledge scores and posttest attitudes scores.  Similarly, Mangas and Martinez (1997) conducted an analysis on environmental concepts and attitudes among 56 biology degree students enrolled in an elective environmental education course at the University of Alicante in Spain.   The students’ environmental knowledge increased at the end of the course and was accompanied by an increase in environmental attitudes.   Tikka et al. (2000) investigated the effect of completing various science courses on the environmental attitudes of 464 students in Central Finland. The results suggested that students who completed biology courses exhibited more positive attitudes towards environmental issues than did students who enrolled in courses related to technology and economics. Makki et al. (2003) assessed 660 secondary school students in the Greater Beirut area also found that environmental knowledge was significantly related to environmental attitudes.   While most studies reported that youth who have more knowledge about the environment hold pro-environmental attitudes (Ramsey and Rickson 1976; Jaus 1982; Mangas, Martinez et al. 1997; Bradley, Waliczek et al. 1999; Makki, Abd-El-Khalick et al. 2003), environmental knowledge does not seem to promote environmental behavior (Kuhlemeier, Bergh et al. 1999; Negev, Sagy et al. 2008; He, Hong et al. 2011)  Negev et al. (2008) studied 1591 6th grade students and 1530 12th grade students in Israel and found significant correlation between environmental knowledge and attitudes, and between environmental attitudes and behavior, but insignificant correlation between environmental knowledge and behavior. Kuhemeier et al. (1999), on the other hand, after studying a nationwide sample of more than 9,000 students from 206 secondary schools in Netherlands, found the relation between environmental knowledge and attitudes and between environmental knowledge and behavior to be weak, but substantial relation between environmental attitudes and behavior.  He et al. (2011) found insignificant relationships between environmental knowledge, attitudes and behavior after studying college students 16 to 20 years old in the City of Shanghai and Gansu Province, China.   Meinhold and Malkus (2005) studied 848 students between 14 and 18 years old on the West coast of the United States 19  (Seattle, Portland, Los Angeles) and found that environmental knowledge was a significant moderator for the relationship between environmental attitudes and environmental behaviors.  2.5 Previous Research on Environmental Knowledge, Attitudes and Travel Behavior Few studies have examined the relationships between environmental attitudes and travel behavior (Kitamura, Mokhtarian et al. 1997; Golob and Hensher 1998; Kaiser, Wölfing et al. 1999; Choo and Mokhtarian 2002; Handy, Cao et al. 2005), fewer have included measures of both environmental attitudes and environmental knowledge (Nilsson and Küller 2000; Walton, Thomas et al. 2004; Flamm 2009).  Walton et al. (2004) addressed the relationships between environmental concern, environmental knowledge, and travel mode choice by distributing a survey to 566 commuters in New Zealand.  Environmental knowledge and attitudes were found to have insignificant impact on travel mode choice.  Nilsson and Küller (2000) studied the environmental knowledge, environmental attitudes, and travel behavior of samples of citizens and public officials in the town of Lund. Using information from 422 respondents, they measured travel behavior by annual driving distance, frequency of trips, choice of travel mode, and level of acceptance of potential traffic restrictions.  The study found that environmental attitudes were more potent than factual knowledge in promoting pro-environmental travel behavior.  Nevertheless, Nilsson and Küller (2000) argue that “despite the weak link between factual knowledge and pro-environmental behavior, knowledge must be an operand in establishing environmental concern and should not be neglected”.   Flamm (2009) assessed the effects of environmental knowledge and environmental attitudes on the number and types of vehicles owned per household, annual vehicle miles traveled and fuel consumption in the Sacramento, California metropolitan region.  The results found that environmental knowledge and attitudes are strongly related: respondents who indicate that protecting the natural environmental is important to them know more about the environmental impacts of vehicle ownership and use.  The households of respondents who own more fuel-efficient vehicles know more about the environmental impacts of vehicle ownership 20  and use.  However, environmental knowledge is not statistically significant in relation to the number of vehicles owned, miles driven, and fuel consumption.  Overall, pro-environmental attitudes and environmental knowledge increase, number of vehicles owned, miles driven, and fuel consumption decreases.  Other than the multiple regression analysis, Flamm (2009) also developed structural equation model to test the bi-directional relationships among environmental knowledge, environmental attitudes and vehicle ownership and use.  The result suggests that knowledge and attitudes have a far stronger impact on behavior than behavior does on knowledge and attitudes.  2.6  Mediating, Moderating and Confounding Effect A common approach of identifying the observed relationship between an independent variable and dependent variable is to hypothesize a direct relationship between the two; an alternative method is via a mediational model.  The mediational model hypothesizes that the independent variable causes the mediator variable, which in turn causes the dependent variable (Baron and Kenny 1986).  Full mediation occurs when the inclusion of the mediation variable fully accounts for the observed relationship between the independent variable and dependent variable.  Partial mediation implies that aside from a significant relationship between the mediator and the dependent variable, there are also some direct relationship between the independent and dependent variable. Moderators in statistics and regression analysis are variables that can make the relationship between two variables either stronger or weaker; it further characterizes interactions in regression by affecting the direction and/or strength of the relationship between two variables (Cohen, Cohen et al. 2003).  Mediation and moderation can co-exist in the form of moderated mediation or mediated moderation.  In moderated mediation, mediation is first established, and then factors that affect the magnitude of the treatment effect are investigated.   (Muller, Judd et al. 2005) Confounders are variables that may have a causal impact on both the independent variable and dependent variable.  They include common sources of measurement error (as discussed above) 21  as well as other influences shared by both the independent and dependent variables.  Ignoring a confounding variable may bias empirical estimates of the causal effect of the independent variable (Pearl 1998). 2.7 Analytical Framework This study investigates the relationship between environmental knowledge, attitudes and travel behavior of youth.  Since previous studies suggested that the relationship between attitudes and behavior could be bi-directional, this study proposes three model structures to test the complex interrelationship among the variables.  In particular, this study distinguishes transportation-specific environmental knowledge and attitudes from general environmental knowledge and attitudes. Model 1) Examine the impact of environmental knowledge on environmental attitudes. Model 2) Disentangle the causal direction between environmental attitudes and travel behavior.   Model 3) Examine the relationship between environmental knowledge, environmental   attitudes and travel behavior.     22  3 Data and Method This chapter first describes the study area, survey design and survey distribution used to collect data.  Then it explains how the analysis was conducted, the steps used to screen the data set and remove inconsistent records.  Finally, it discusses the methods used to test the validity and reliability of the data, the modeling methodologies used to analyze the relationship between environmental knowledge, environmental attitudes and travel behavior. 3.1 Survey Design and Implementation 3.1.1 Study Area Two municipalities of Greater Vancouver, the City of Vancouver and the City of Richmond, are chosen as the focus of this research study.  In 2006, Greater Vancouver had a population of 2,116,581 living in 870,992 dwellings.  It has a land area of 2,877.36 km2 and a population density of 735.6 per km2.    The demographic and transportation statistics for Greater Vancouver, City of Vancouver and City of Richmond are presented in Table 3.1.  The combination of the two cities represents the demographic and transportation statistics for the Greater Vancouver region well, suggesting the survey results collected from the two cities can be generalized for the overall region. With only four percent of Greater Vancouver’s land, the City of Vancouver has over a quarter of the population and over a third of the jobs in the region.  The City of Vancouver is renowned for its innovative programs in the areas of sustainability, accessibility and inclusivity.  While Vancouver is recognized as one of the world’s most livable cities with one of the smallest carbon footprints of major cities in North America, the City of Vancouver had a vision to become a leader in initiating a green movement worldwide.  In February 2009, the city launched the Greenest City initiative with a goal to become the greenest city in the world by 2020.   The Greenest City 2020 Action Plan was developed and announced in July 2011, consisting of a list of goals, strategies and action items to fulfill the green initiative.  The area of interest relating to this research paper in particular is Green Transportation.  The strategies in this area include land use, walking and cycling, transit, demand management, low carbon 23  vehicles, and goods movement.  All of which aligns with the topic of environmental attitudes and travel behavior.  This well-rounded approach has attracted the participation of group of independent experts representing knowledge and interest across a range of the most pressing environmental issues the City of Vancouver faces. The City of Richmond, located to the south of the City of Vancouver, represents approximately 10% of the population and jobs in the region.  It is a unique island city, located in close proximity to downtown Vancouver and the U.S. border.  It is connected to neighboring lower mainland cities by a series of bridges and a tunnel.   The City of Richmond offers a number of comprehensive services designed to ensure a clean, livable and sustainable community for current and future generations. One project is the City’s Energy Management Program.  The City of Richmond has been improving its corporate energy use for over 15 years, resulting in reduced greenhouse gas emissions and contribution to climate change (over 1,800,000 kg of CO2/year in reduced emissions). Richmond has gone through significant improvements in its transportation system.  Prior to the 2010 Winter Olympics hosted in Vancouver and construction of the Canada Line, which connects the YVR airport and downtown Vancouver, transit ridership in Richmond was merely 12% in 2006.  Transit ridership of the Canada Line and was forecasted to see 100,000 boardings per day by 2013.  On Aug. 6, 2010, TransLink released figures that show average daily ridership almost reached 100,000 in July.  In February 2011, ridership reached 110,000 per day, exceeding the original ridership forecast well in advance.  Carrying this many passengers eliminates approximately 14,000 tonnes of greenhouse gas emissions annually.  The ridership increases were fueled by buses feeding the Canada Line and transit demand created by City of Richmond’s policy of increasing population density in its city centre.        24  Table 3.1 Demographic and Transportation Statistics for the Greater Vancouver, City of Vancouver and City of Richmond    2006 Greater Vancouver City of Vancouver City of Richmond Population and Households Population 23133281 6035021 1904731 Average Household Size 2.6 2.2 2.8 Average number of cars per household 1.28 1.04 1.74 Age Groups       19 and under 22.8% 17.9% 22.4% 20-39 28.5% 34.5% 26.5% 40-64 35.8% 34.5% 38.4% 65 and over 12.8% 13.1% 12.7% Labour force       Employed labour force 1104760 310640 87180 Unemployment rate 5.6% 6.0% 4% Household Income       Median Income $55,231  $47,299  $53,489  Average Income $73,258  $68,271  $67,440  Dwelling       Owned 40.4% 48.1% 76.3% Rented 59.6% 51.9% 23.7% Single-detached house 10.4% 19.1% 41.1% Semi-detached house 1.2% 1.5% 3.0% Row house 14.7% 3.2% 19.1% Apartment 73.4% 76.1% 36.3% Mode of travel to work Car, truck, van as driver 67.3% 51.5% 73.0% Car, truck, van as passenger 7.1% 6.1% 9.0% Public transit 16.5% 25.1% 12.0% Walk 6.3% 12.2% 4.0% Bicycle  1.7% 3.7% 1.0% Other method 1.1% 1.3% 1.0% 1 Data in 2011   25  3.1.2 Survey Implementation Research applications were submitted to the Richmond School Board (RSB) and Vancouver School Board (VSB) in May 2011 and August 2011, respectively.  The application included the study description, survey questionnaire, conditional approval from the University of British Columbia Behavioral Research Ethics Board and a consent form for parents and assent form for students. (Appendix A).  The RSB Research Committee approved the application in April, 2011, while The VSB Research Committee approved the application in January, 2012.    It took a much longer time for the VSB Research Committee to approve this research due to teacher’s job action.  Starting in September, 2011, teachers were in a legal strike which restricts their ability to participate in administrative duties and other non-instructional activities.   As a result, the survey implementation methods had to be revised to provide an alternative so that surveys can be conducted with or without teachers’ assistance.  Below are the implementation methods that were available and were customized according to individual school’s need.  A complete and a short version of the questionnaire were developed depending on the implementation methods. The short version of the questionnaire was developed so that surveys can be conducted without teachers’ assistance.   1) In-Class survey Our research team prepared and delivered the questionnaires to the schools. The team and/or the partner teachers gave a brief introduction of the project to the students, and students had dedicated 30 minutes during class time to complete the long version of the questionnaire.  This enabled the students to give their full attention to the questionnaire.  The team and/or the partner teachers were available to facilitate the implementation of the survey and to provide clarification if students had questions.   2) Lunch Time Survey  The research team prepared and delivered the questionnaires to the schools. Booths were set up in the cafeteria or anywhere in the school that could attract the students’ attention during lunch time.  The research was explained to the students and they were 26  invited to participate.  This short version of the questionnaire took the students 10 minutes to complete.  As soon as approval was received from the school board to carry out this research, all 18 secondary schools in the City of Vancouver and 10 secondary schools in the City of Richmond were visited.  Figure 3.1 and Figure 3.2 shows the location of the secondary schools in the Vancouver and Richmond School District.   Figure 3.1 Secondary School Map of the Vancouver School District  (Source: https://www.vsb.bc.ca/schools)         27  Figure 3.2 Secondary School Map of Richmond School District  (Source: http://www.sd38.bc.ca/schools/Secondary_Boundaries)  During school visits, the team was introduced and then the research project and the survey implementation options were described to the principals or vice-principals.  If the principals or vice-principals were not available at the time, the team’s contact information, description of the study, survey questionnaires (long and short), and the approval letter from the school board were left to the administrator to pass along to the principals or vice-principals.   For schools who agreed to have their students participate in the in-class surveys, the principals or vice-principals sent out invitation to the teachers.  Teachers who were interested in this study would then contact our research team to schedule a time for survey implementation.  For 28  schools who were interested in conducting lunch time surveys, we sent emails to contact the principals and vice-principals to arrange the survey day.  For schools which the team was not able to meet with the principals or vice-principals, follow-up emails were sent to describe our study and mention the team’s earlier visit to the school and relevant documents were left to the administrator.   Surveys were conducted between October, 2011 and February, 2012.   A total of eleven schools, six schools in the City of Vancouver and five schools in the City of Richmond participated in this study.  One school participated in in-class survey, eight schools participated in lunch time survey, and two schools participated in both in-class and lunch time survey.  Table 3.2 provides a summary of the implementation method, responses and the size of each school.  A total of 1032 responses were collected, of which 749 and 220 were from the lunch time and in-class survey, respectively.  Three grade 8 Science classes, one grade 9 Science class, one grade 10 Leadership class, one grade 11 Science class, one grade 11 English class, one grade 11 Social Studies class, one grade 11/12 Marketing class and one grade 12 Accounting class participated in the in-class survey. For this thesis, the data used for analysis was mainly from four common sections of the in-class and lunch time survey.  Since the implementation method is different, the survey results are compared in section 4.5 to provide justification for combing the data collected.  The comparison found no significant differences between the samples collected from the in-class and lunch time survey. The team tried on a best effort basis to try and obtain as many comprehensive samples as possible so that the data collected would be a better representation of the students in the regions.  Only the teachers that were interested would delegate their class time for our research team to conduct the survey. Due to this, we acknowledge the results may be biased as the principals and teachers who were interested in this study might have higher levels of pro- environmental attitudes which could have also influenced their students’ attitudes.  29  Table 3.2 Implementation Method and Number of Responses by Schools    In-Class Lunch Time Number responses Enrollment1 City of Vancouver Windermere     68 1332 Vancouver Technical    105 1665 Kitsilano     101 1471 Prince of Wales    65 1294 Lord Byng    91 1341 Sir Winston Churchill     108 2100 City of Richmond H.J. Cambie    93 853 R.C. Palmer    53 736 Matthew McNair   116 930 Robert Alexander McMath   165 1249 Hugh Boyd    63 826 1 2011/12 Source: http://www.bced.gov.bc.ca/reporting/school_data_summary.php 3.1.3 Questionnaire Design  The complete questionnaire contains nine sections: environmental attitudes and behavior, environmental assessment and knowledge, travel to school, attitudes toward travel modes, perception of cycling, environmental information acquisition, personal and household information and environmental belief and values.  The short questionnaire contains four sections: environmental attitudes, environmental knowledge, travel to school, and personal and household information.   Table 3.3 shows the comparison between the complete and short questionnaires.  Table 3.4 provides a summary of the survey components.  Both the complete and short questionnaires can be found in Appendix A.   With the exception of the environmental belief and values section which used the New Ecological Paradigm Scale (NEPs), all other questions were developed originally by the author with inputs from the teachers.  The teachers drew from their experiences in the classroom, observation of the students and review of the curriculum.  In particular, the science teachers provided quizzes and exams which were utilized as reference in the development of the 30  environmental knowledge questions.  Select students volunteered to review the questionnaire and recommend revisions to make the language clearer to them.       Table 3.3 Survey Comparison   Complete Survey Short Survey Environmental Attitudes    Environmental Behavior    Environmental Assessment and Knowledge   Travel to School   Attitudes toward Travel Modes    Perception of Cycling    Environmental Information Acquisition    Personal and Household Information   Environmental Belief and Values     31  Table 3.4 Survey Component Section Total Detailed Components Environmental Attitudes    General Environmental Government Intervention on environmental issues General environmental issues related to transportation Government intervention on environmental issues related to transportation Change of Travel Behavior # of Questions 17 5 3 3 4 2 Environmental Behavior   Perform in Private Space Perform in the Public Space Influence Other People to Perform Environmental Behavior     # of Questions 13 4 5 4     Environmental Assessment and Knowledge   Assessment on the City's Environmental Problem General Knowledge Transportation related knowledge     # of Questions 16 2 8 6     Travel to School   Commuting to School Utilization of Travel Modes       # of Questions 13 12 1       Attitudes toward Travel Modes   Anticipated Travel Mode Driving - Adult Identity Transit and Social Image Car Dependence Car Pride # of Questions 20 2 3 3 6 6 Perception of Cycling   Bike Ownership Trip Purpose Terms associated with Cycling  Statements on different aspect of Cycling   # of Questions 5 2 1 1 1   Environmental Information    General Transportation       # of Questions 6 3 3       Personal and Household Information   Personal Household       # of Questions 15 7 8       Environmental Belief and Values   The reality of limits to growth Antianthropocentrism The fragility of nature's balance Rejection of exemptionalism The possibility of an ecocrisis # of Questions 15 3 3 3 3 3 32  Environmental Attitudes   Environmental attitudes were measured using fifteen statements to describe four types of environmental issues: general environmental problems, general environmental problems with government intervention, environmental issues related to transportation, and environmental issues related to transportation with government intervention. The five point Likert scale (strongly disagree, partially disagree, neutral, partially agree, or strongly agree) was used to establish a rating for each statement. Statements could be phrased positively or negatively, but they were coded so that higher scores indicated stronger pro-environmental attitudes.  In addition, two questions were asked to understand if the respondent has changed and has encouraged other people to change their travel behavior because of environmental concern. Environmental Behavior Environmental behaviors were assessed using statements which describe 13 practices. The 13 practices were grouped into 3 categories: environmental behavior performed in private space, environmental behavior performed in public space and actions influencing others on environmental behavior.  Respondents were asked to report the frequency of their environmental friendly behavior; their choices ranged from always, sometimes, rarely and never. For each statement, a respondent’s environmental behaviors were scored as either “very friendly” (4), “friendly” (3), “neutral” (2), or “unfriendly” (1).  Higher scores indicated greater pro-environmental behavior.   Environmental Assessment and Knowledge At the beginning of the section, respondents were asked to evaluate the degree of seriousness of the environmental problems in Vancouver through two questions.  Following that, respondents’ knowledge about the environment was assessed through fourteen questions.  The questions were designed in various difficulties in order to distinguish students with different knowledge levels: eight on general environmental knowledge and six on environmental knowledge specifically related to transportation. Each of the questions had four choices, one of which was the correct answer. To discourage participants from skipping over the questions, a 33  “do not know” option was not included. Each respondent’s percentage of correct answers was used to assess his/her level of environmental knowledge. General knowledge and transportation-specific knowledge were measured separately.   Travel to School Respondents were asked to report their to-school and after school trips of the previous school day. The details of the respondent’s school trip were explored through a series of questions concerning the decision maker of the trip, factors that influence the current mode choice, who they travelled with, distance between their home and school, perceived safety of the trip, ideal travel time and ideal travel mode.  Respondents were also asked to indicate the frequency of use for each travel mode.    Attitudes towards Transportation Respondents’ perception of different travel modes were explored through eighteen statements and two questions.  The eighteen statements were used to quantify five subjective measures:  car pride, car dependence, driving and adult identity, and transit and social image.  Respondents were asked whether they strongly agree, agree, neutral, disagree, or strongly disagree with each statement.  A Likert-scale was used to induce a rating (1 to 5) for each item.  Potential scores for each measure are different as the number of statements for each subjective measure varies.  Respondents were also asked to indicate their ideal mode of travel and whether or not they will buy a car in the future.   Perceptions of Cycling Respondents’ were asked to indicate their bike ownership, purpose of bike trips, and terms associated with cycling.  Respondents’ perception toward cycling was then explored through thirteen statements.   Environmental Information The details of the environmental information that the respondent received were obtained through six questions concerning the source and potential impact of the information. 34  Personal and Household Information Demographic information (gender, grade, car ownership, household income, number of household members, number of children in the household, driver licenses, and residency) concerning the respondent and the respondent’s household was acquired through twelve questions.    Environmental Value The New Ecological Paradigm Scale (NEPS), a set of 15 questions that elicits opinions on limits to growth, anthropocentrism, the balance of nature, human exceptionalism and ecological crisis is used to measure the environment value of the respondents.  The respondents were asked whether they strongly agree, agree, neutral, disagree, or strongly disagree with each statement.  A Likert-scale was used to induce a rating (1 to 5) for each item.  Statements could be phrased positive or negatively, but they were coded so that higher scores indicated a more proecological view. 3.2 Data Entry and Data Screening Data entry was accomplished with the assistance of two graduate students and two undergraduate students at the University of British Columbia.  SurveyMonkey, web-based survey solution, was used for questionnaire development and data input.  The data was then exported to Excel for data analysis.  The data entry work commenced concurrently with the survey implementation.  Each questionnaire is coded with a unique survey ID, which allowed for auditing to ensure that the data was inputted correctly.   Data screening was performed to remove records that have repeated patterns, illogical responses, and non-random missing values.   Below are the criteria for removing the records. 1) The statements on environmental attitudes, travel modes, cycling and environmental values are phrased positively or negatively.  Therefore, the records are removed if the respondent selected the same answer for all the statements.  The only exception is that the respondent selected “Neutral”.   35  2) The records are removed if the respondent answered their household consists of more than ten children, more than ten members with driver licenses and have zero or more than ten household members.   3) The records are removed if the respondent skipped an entire section 3.3 Data Validity Confirmation on the validity and reliability of instruments are highly important to assure the integrity of study findings.  Traditionally, validity of instruments has been determined by examining construct, content, and criterion-related concepts.  In the paper “A Psychometric Toolbox for Testing Validity and Reliability”, validity is defined as the ability of an instrument to measure the attributes of the construct under study; where construct is an explanatory variable which is not directly observable (DeVon, Block et al. 2007).  Reliability on the other hand refers to the ability of an instrument to measure an attribute consistently. As described in the previous section, environmental attitudes were measured using fifteen indicators.  Testing the validity and reliability of the factor-indicators groupings increases the strength of the subsequent structural equation models by introducing less error in the latent variables.  This section first reviews the concepts of reliability and validity, and then describes the methods that are used in this study. Cronbach & Meehl (1955) defined construct validity as the degree to which an instrument measures the construct it is intended to measure.  An instrument is constructually valid if it has the capability to exclusively measure concepts that are theoretically and structurally related.  However, if the instrument is also capable of measuring closely related concepts, then it might not have adequate construct validity as a measure of the construct.  All types of validity fall under the broad heading of construct validity; under which are translational validity and criterion validity.  Content and face validity are types of translational validity; whereas concurrent, predictive, convergent and discriminant are types of criterion validity (Trochim 2001).  36  Construct validity of an instrument can be evaluated via contrasted groups, hypothesis testing, and factor analysis.  In the contrasted groups approach, two groups known to be high and low in the construct being measured are sampled.  If the mean scores of the two groups differ significantly, then the instrument is valid.  In other words, an instrument is constructually valid if it is capable of differentiating between two contrast groups.  Another evaluation method is hypothesis testing, which is based on a theoretical approach.  Construct validity is supported if the direction of scores on the measure reflects the framework as hypothesized.  Factor analysis is a statistical method commonly used during instrument development to analyze relationships among large number of variables.  Exploratory factor analysis (EFA) is used to identify the greatest variance in scores with the smallest number of factors.  Confirmatory factor analysis (CFA) generally follows EFA and includes theoretical knowledge to further test the construct validity by validating the extent to which the statistical model fits the actual data (Waltz, Strickland et al. 2005).  Munro (2005) suggests, in general, a minimum of five items per variable is recommended for factor analysis. Related test items define a part of the construct and are grouped together; whereas unrelated items do not define the construct and should be deleted from the tool.   Trochim (2001) defines content and face validity as types of translational validity.  Content validity is indicated if the items in the tool sample the complete range of the attribute under study.  A large pool of potential items is defined, which will then be reduced based on experts’ review.  However, establishing strong support for content validity can be a challenge.  An estimate of validity is often inflated as experts endorse most items.  In contrast, face validity is the easiest way to claim support for construct validity but also the weakest form of validity.  It is a subjective assessment on the face of the construct which provides insight into how potential participants might interpret and respond to the measure.   Trochim (2001) further defines concurrent, predictive, convergent and discriminant as types of criterion validity.  Criterion-related validity is indicated when measures on the predictor and the criterion variables are correlated; and the strength of the correlation substantially supports the extent to which the instrument estimates performance on each criterion (Waltz, Strickland et al. 37  2005).  However, a faulty criterion-related validity estimate could occur due to criterion contamination.  Criterion contamination, error in measurement of the criterion as an example, leads to an exaggerated correlation between instrument and criterion variables. Concurrent criterion-related validity is confirmed when scores on a tool are correlated to a related criterion at the same point in time (Carmines and Zeller 1979).  Predictive validity is referred to the degree to which test scores predict performance on some future criterion.  High correlations between the original measure and criterion variables reinforce the conclusion that the tool is a valid predictor of the specified criteria.  Convergent validity is a correspondence or convergence between constructs that are theoretically similar.  The interitem correlation coefficients would be high in an instrument that has convergent validity.  Conversely, discriminant validity is the instrument’s capability to differentiate or discriminate between constructs that are theoretically different.  Interitem correlation coefficients would be low in an instrument with discriminant validity. Reliability refers to the ability of an instrument to measure an attribute consistently.  It is a necessary but not sufficient component of the validity of an instrument.  While the true score and error score cannot be known, the amount of both random and systematic error can often be controlled for.  Stability reliability is tested when the attributes under study are not expected to change.  On the other hand, equivalence reliability indicates whether all items in the tool reliably measure the attributes and if participants score similarly on like measures. A form of stability reliability is performed through test-retest, which is estimated by administering the same test to the same group of respondents at different times.  The correlation between the two scores, and often between individual questions, indicates the stability of the instrument.  It should be noted that the major weaknesses of test-retest measures of reliability are the memory reactivity effects.  The time interval should be long enough that respondents do not remember their original responses, but not long enough for their knowledge of the material to have changed.  In general, the longer the time, the lower the reliability and the more likely that knowledge or attitudes actually have changed (Trochim 2005).  Test-retest reliability is relevant for cognitive and trait scales that are not expected to 38  change over time.  However, it is not appropriate for states that are expected to change over time, such as attitude, mood, or knowledge following an intervention.   As for equivalence reliability, Cronbach’s alpha coefficient is the most frequently used statistics to show internal consistency reliability.  Internal consistency indicates how well the items on a tool fit together conceptually.  Cronbach’s alpha coefficient is the only reliability index that can be performed with one test administration.  It is a measure of the internal consistency for the test responses from the current participants and should be computed each time the test is administered (Waltz, Strickland et al. 2005).  Higher coefficient alpha values can be achieved by adding correlated items (Nunnally and Bernstein 1994; DeVellis 2003).  When non-correlated items are added, the value of alpha is reduced.  Inflated alpha values are achieved when computed for an entire scale composed of two or more subscales.  Each subscale should be computed individually as opposed to for the entire scale.  In general, reliability coefficients around 0.90 are considered “excellent”, values around 0.80 are “very good” and values around 0.70 are “adequate”.  If the reliability coefficient is below 0.50, most of the observed score variance is due to random error, an unacceptable amount of imprecision in most research (Kline 2010).   Another method for accessing equivalence reliability is through alternative forms, also known as parallel forms.  Alternative forms reliability pertains to different versions of an instrument to determine reliability of scores.  It can prevent participants from using knowledge gained from the first test.  However, it should be noted that it may be difficult or impossible to generate sufficient items for two forms of a test if the entire content of items is small. 3.4 Analytical Methods Factor analysis and structural equation modeling are used to test hypotheses made about the relationship between environmental knowledge, environmental attitudes and travel behavior.  The first step is factor analysis which prepares the data for analysis by grouping a variety of indicators into cohesive factors.  After the strongest factor-indicator relationships have been established, structural equation modeling is used to test the hypothesized relationships between the latent variables (factors) and the observed constructs.  Both factor analysis and 39  structural equation modeling in this study are estimated in Mplus Version 5 (Muthen and Muthen 2007). 3.4.1 Exploratory and Confirmatory Factor Analysis When studying travel behavior, the viewpoint is commonly transportation-agency-centric.  The objective is focused on prediction based on standard economy theory.  It is believed that choice behavior can be approximated via Random Utility Maximization (RUM) models, where perception, construal, and cognitive processes are well approximated by RUM models (Ben- Akiva, McFadden et al. 2002). Meanwhile, behavioral scientists have focused on demonstrations of how the assumptions of standard economics models often fail or why the decision processes are far from those assumed in standard theory (Loewenstein 2001).  On the other hand, psychologists focus on understanding travel behavior via studying preference changes, how to manage and influence them.  This customer-centric approach is defined from the viewpoint of the customer, encouraging planning agencies to recognize the connection between traveler preferences and the transportation systems.  Gopinath (1995) suggests the economic and psychological choice theories are not antithetical but they can be utilized in conjunction with developments in psychometrics and econometrics to advance a richer class of choice models.  McFadden (2000) foresaw the future of behavioral travel demand analysis as “The standard RUM model, based on a mildly altered version of the economists’ standard theory of consumer behavior that allows more sensitivity of perceptions and preferences to experience, augmented with stated preference, perception, and attitude measures that uncover more of the process by which context molds choice, will increasingly become the dominant methodology for behavioral travel demand analysis.”  McFadden also acknowledges that reliable scales for stated preferences, perceptions, attitudes and reliable mappings from experience and information to perceptions and attitudes would be useful and can be obtained via comprehensive research effort.   This study takes a customer-centric approach in understanding travel behavior.  Answers to survey questions regarding attitudes, perceptions, motivations, intentions are investigated via 40  factor analysis, which includes both exploratory factor analysis (EFA) and confirmatory factor analysis (CFA). Exploratory factor analysis (EFA) identifies the underlying structure of a large set of variables.  .  Researchers have no direct influence on the interaction between indicators and factors, hence the term “exploratory”.  The factor structure of the data is discovered via factor loadings and is tested against unrestricted factor models (Kline 2010).  Data reduction can be achieved via selecting those variables that have the highest correlation with each factor.  In terms of EFA application related to transportation, indicators can be grouped into factors to represent the different types of environmental attitudes, attitudes toward transportation, and perception toward cycling.  Data reductions are then used to ensure that only the most appropriate indicators are analyzed.  Confirmatory factor analysis (CFA) is used to determine if the number of factors and the loadings of indicators on them follow the hypothesis formed in pre- established theory or EFA.  Factor loadings are the correlation coefficients between the indicators and the factors.  If a factor loading is extraordinarily low or lower than that of the other indicators for that factor, the corresponding indicator should be considered for elimination from the analysis and should only be kept if the researcher feels it is important to the definition of the factor.   Goodness of fit measures should be examined to ensure that they are at an acceptable level.  This ensures that the most appropriate factor-indicator groupings are being used, which increases the strength of the subsequent structural equation models by introducing less error in the latent variables.   3.4.2 Structural Equation Modeling Structural equation modeling (SEM) tools are increasingly being used in behavioral science research for the causal modeling of complex, multivariate data sets in which the researcher gathers multiple measures of proposed constructs (Hair, Anderson et al. 1998).  In particular, it has been used in modeling travel behavior and values (Golob 2003).  SEM is a statistical technique for testing causal relationships between observed and unobserved variables where each unobserved variable is represented by a collection of observed variables.  For example 41  attitudes could cause behavior, behavior could cause attitude or the causality could be in both directions, and SEM can be used to test such relationships.  The SEM method composes of a measurement model and a structural model.  The measurement model is used to find the relationship between indicators and factors, while the structural model specifies and tests the theory-based hypotheses made about the relationship between endogenous (dependent) and exogenous (independent) factors and observed constructs.  The links between a cause variable and an effect variable is a direct effect that is observable.  Indirect effects between two variables are the effects along the paths between the two variables that involve intervening variables.  The combination of direct and indirect effects creates the total effect and is used to specify latent variables as linear functions of other indicators in the model by deciding which of the parameters defining the factors are restricted to be zero, and which are freely estimated.  Specification of each parameter allows the modeler to conduct a rigorous series of hypothesis tests regarding the factor structure. Similar to other statistical modeling, each hypothesis is accepted or rejected by examining the statistical significance of coefficients and the overall fit of the model is assessed using a variety of goodness of fit measures.  Finally, R2 values are reported for each of the endogenous variables which indicates how much of that construct’s variance is explained by the model structure.   Many criteria have been developed for assessing goodness-of-fit of an SEM.  This ensures that the most appropriate factor-indicator grouping are being used, which increases the strength of the subsequent structural equation models by introducing less error in the latent variables (Bentler 1990; Hu and Bentler 1999).  Most of these criteria are based on the chi-square statistic given by the product of the optimized fitting function and the sample size.  The objective is to attain a non-significant model chi-square since the statistic measures the differences between the observed variance- covariance matrix and the one produced by the model. There are problems associated with the use of the fitting function chi-square mostly due to the influence of sample size.  For large samples, it may be very difficult to find a model that cannot 42  be rejected due to the influence of sample size.  Many of the goodness-of-fit indices use normalization to cancel out the sample size in the chi-square functions, such as root mean square error of approximation (RMSEA) which measures the discrepancy per degree of freedom (Steiger and Lind 1980). RMSEA is one of the favored statistics because it is possible to compute its confidence interval as well as the mean value.   Steiger (1989), Browne and Mels (1990), and Browne and Cudeck (1992) offered guidelines for interpretation of the value of RMSEA.  By analyzing many sets of empirical data and evaluating the behavior of RMSEA in relation to previous conclusions about model fit.  Steiger (1989) and Browne and Mels  (1990) arrived independently at the recommendation that values of RMSEA less than 0.05 be considered as indicative of close fit.  Browne and Cudeck (1992) also suggested that values in the range of 0.05 to 0.08 indicate fair fit and that values above 0.10 indicate poor fit.   There are also goodness-of fit measures based on the direct comparison of the sample and the model-implied variance-covariance matrices including the standardized root mean square residual (SRMR), which ranges from 0-1, where large values mean high residual variance, and that such values reflect a poorly fitting model.  It is generally accepted that the value less than 0.05 being considered a good fit (Hair, Anderson et al. 1998).   Bentler’s comparative fit index (Bentler 1990) is another often used goodness-of-fit measure.  CFI assesses the relative improvement in fit of the researcher’s model compared with a baseline independence model, called the null model, which assumes zero population covariance among observed variables.  A rule of thumb is that values greater than 0.90 may indicate reasonably good fit of the proposed model (Hu and Bentler 1999).   MacCallum et al (1996) suggested these guidelines are intended as aids for interpretation of a value that lies on a continuous scale and not as absolute thresholds.  Kline (2010) recommended not to rely solely on thresholds for approximate fit indexes when deciding whether or not to re-specify a structural equation model because there are limitations to all approximate fit indexes.  A healthy perspective on approximate fit indexes is to view them as providing qualitative or description information about model fit.  The value of this information 43  increases when you report values of indexes that as a set of assess model fit from different perspective.  The drawback is the potential for obfuscation, or the concealment of evidence about poor fit.   This is less likely to happen if following a comprehensive approach to assess model fit that includes taking the model chi-square test seriously and describing patterns of residuals. 3.4.3 Multiple-Indicator Multiple-Cause (MIMIC) Analysis The multiple-indicator multiple-cause (MIMIC) model estimates the measurement equations relating each factor to its indicators.  A MIMIC factor has both cause indicators and effective indicators, and they can be continuous or categorical, where a categorical cause indicator represents group membership.  A MIMIC factor with a single cause indicator and the rest as effect indicators is an equivalent version of a standard one factor CFA model. 3.4.4 Multiple Group Analysis (MGA) Multiple group analysis (MGA), also known as multiple sample analysis or tests of invariance, is used to compare model fit across groups.  The determination of overall model fit and individual parameters are the same as CFA and SEM.  MGA is used to test the invariance or equality across the groups for factor loadings, factor variances/covariance, and structural coefficients. 3.4.5 SEM and its Application in Transportation SEM has been widely applied to travel behavior research starting from 1980.  More than 50 applications are cited by Golob (2003) in his ‘Structural equation modeling for travel behavior research’ review.  These studies range from travel demand modeling using cross-sectional data, dynamic travel demand modeling, activity-based travel demand modeling, applications to capture attitudes, perceptions and hypothetical choices, organizational behavior and values, and driver behavior. In the transportation field, causal relationships can exist in many forms.  Previous research of causality includes: car ownership and distance traveled (Den Boon 1980); car ownership, season ticket ownership and modal usage (Axhausen, Simma et al. 2001); mode choice behavior and 44  attitudes (Tardiff 1977); mode choice and support for policies that benefit the environment (Golob and Hensher 1998); acceptance of road pricing, intention to reduce car use and feelings related to fairness and freedom (Jakobsson, Fujii et al. 2000).    Dobson et al (1978) examined the attitude-behavior relationship and concluded that attitudes and choices form a two-way causality relationship.  Morikawa and Sassaki (1998) conducted a Dutch survey to capture the influence of latent subjective attributes of choice alternatives on choices and came to similar findings that models with causality only from attitudes to behavior perform less well than those with causal links in both directions.  In another study, Golob’s research on mode choice and attitudes related to combined HOV and Toll lanes was inconclusive.  Ory and Mokhtarian (2009)  found travel amounts influence perceptions; and desires are shaped by both perceptions and affections.  Choocharukul, Van and Fujii (2008) in psychological effects of travel behavior on preference of residential location choice found that preference regarding residential location was significantly affected by behavioral intention towards car usage.  Aside from causality, SEM was used to identify a reduced model which consists of factors that have higher correlation.  Outwater et all (2003) identified three attitudinal factors which were used to partition the ferry-riding market into eight segments.  This resulted in recognition that mode choices were different for market segments and these markets were sensitive to travel stress or the desire to help the environment.  Kitamura and Susilo (2005) in a stability of travel patterns over time study found that changes in travel pattern are largely due to the instability of structural relationships while changes in demographic and socio-economic factors play relatively minor roles.    45  4 Environmental Knowledge, Attitudes and Travel Behavior This chapter begins with descriptive statistics of the sample population, followed by a comparison between the sample population and the average household in the study area.   The analysis on three key aspects of this research: environmental knowledge, environmental attitudes and travel behavior is then presented.  Lastly, it compares the survey results between the samples collected from in-class surveys and lunch time surveys. 4.1 Demographic and Socioeconomics Information Basic descriptive statistics and frequencies of responses to the twelve demographic questions are presented in Table 4.1 below.   There are a total of 937 useable responses after data validation. When compared to the average household in metro Vancouver, respondents from the survey have a larger household size, have a higher household income, are less likely to live in an apartment and are more likely to live in single housing and own their own home (Table 4.2).  This is expected as households with children represent larger household sizes and require larger homes.  Also, typically households with children have higher household income and are financially stable to mortgage a home.        46  Table 4.1 Respondent and Household Characteristics   Number responses % of responses           Gender (N = 874)   Number of Household Licensed Drivers (N =830) Male 348 40.3%   0 22 2.7%   Female 515 59.7%   1 131 15.8%   Grade (N = 866)   2 372 44.8%   8 169 19.5%   3 202 24.3%   9 105 12.1%   4+ 103 12.4%   10 186 21.5%   Driver License (N=854  )     11 199 23.0%   Yes 213 24.9%   12 207 23.9%   No 641 75.1%   Number of Household Members (N =824)   Number of Household Vehicles (N = 871) <=2 51 6.2%   0 56 6.4%   3 148 18.0%   1 309 35.5%   4 332 40.3%   2 349 40.1%   5 171 20.8%   3+ 157 18.0%   6+ 122 14.8%   Own Vehicle (N = 866)     Number of Children in household (N = 833)   Yes 85 9.8%   0 13 1.6%  No 781 90.2%   1 276 33.1%   Dwelling Type (N = 814)     2 373 44.8%   House 560 68.8%   3 128 15.4%   Duplex 41 5.0%   4 27 3.2%   Townhouse 132 16.2%   5 8 1.0%   Low Rises 48 5.9%   6+ 8 1.0%  High Rises 33 4.1%   Monthly Household Income (N = 438) Other 0 0.0%   Less than $25,000 94 21.5%   Residency (N = 857)     $25,000 to $49,999 68 15.5%   Own 598 69.8%   $50,000 to $74,999 94 21.5%   Rent 233 27.2%   $75,000 to $99,999 61 13.9%   Other 26 3.0%   $100,000 to $149,999 50 11.4%   Years Living in Current Home     $150,000 and over 71 16.2%   less than 3 months 37 4.3%           3 months to 1 year 77 9.0%           1 year to 3 years 224 26.2%           3 years to 5 years 127 14.8%           more than 5 years 391 45.7%    47  Table 4.2 Comparison between Samples and Average Household   Respondents Greater Vancouver City of Vancouver City of Richmond Population and Households Household Size 4.3 2.6 2.2 2.8 Number of cars per household 1.68 1.28 1.04 1.74 Household Average Income 76700 73258 68271 67440 Dwelling Owned 70.0% 40.4% 48.1% 76.3% Rented 30.0% 59.6% 51.9% 23.7% Single-detached house 69.0% 10.4% 19.1% 41.1% Semi-detached house 5.0% 1.2% 1.5% 3.0% Row house 16.0% 14.7% 3.2% 19.1% Apartment 10.0% 73.4% 76.1% 36.3%  4.2 Environmental Knowledge  Table 4.3 lists the environmental knowledge questions that students responded to, the correct answer for each question and percentage of students who responded correctly.  Students in general have low level of knowledge about the environment.  On average, they answered 6.86 of the fourteen questions correctly (Standard Derivation = 2.45), with only 1% of respondents answering more than 80% of the questions correctly (Figure 4.1).   The average percentages of correct answers for both general and transportation-specific questions were 49%.  Amongst general knowledge (GENKNOW) related questions, question #5 regarding the definition of organic food had the most number of correct responses (67%).  Question #2 regarding ecological footprint had the fewest number of correct responses (15%).  Amongst transportation-specific knowledge (TRANKNOW), question # 9 regarding the transportation method that creates the least pollution had the most number of correct responses (87%).  Question #14 regarding the percentage of the world’s total CO2 emission attributable to the transport sector had the fewest number of correct responses (31%).   Figure 4.2 analyses general and transportation-specific environmental knowledge by demographic and socioeconomic variables.  A series of ANOVA tests, reported in Appendix B, 48  found significant differences in general knowledge (GENKNOW) and transportation-specific knowledge (TRANKNOW) for the whole sample at the 5% level, and significant differences in GENKNOW by driver’s license status and owning personal vehicle, significant differences in TRANKNOW by number of household people, number of children in the household and years living in a home at the 5% level. Overall, there is slight increase along the grades in either general or transportation-specific knowledge but the difference is not statistically significant.    These bivariate analysis results serve as exploratory purpose only.  For example, students with a driver license have better environmental knowledge, possibly because students with a driver license are in higher grades and are more informed, as opposed to a direct correlation with having a driver license.  This becomes clearer in the multivariable analysis in the SEM models: after demographic and socioeconomic variables are controlled, driver’s license status no longer plays a significant role in explaining environmental knowledge.  Figure 4.1 Number of Environmental Knowledge Questions Answered Correctly      0 20 40 60 80 100 120 140 160 0 1 2 3 4 5 6 7 8 9 10 11 12 13 14 Fr e q u e n cy  Number of Environmental Knowledge Questions Answered Correctly 49  Figure 4.2 Average levels of GENKNOW and TRANKNOW     0% 20% 40% 60% 80% 100% All Male Female 8 9 10 11 12 Yes No Yes No Gender Grade Driver License Own Car GenKnow TranKnow 0% 20% 40% 60% 80% 100% Lo w M e d H ig h 0 1 2 3 4 + 0 1 2 3 + 2 3 4 5 6 + 0 1 2 3 4 + Household Income Household License Household Car Household People Household Children GenKnow TranKnow50      0% 20% 40% 60% 80% 100% 0 1 2 3 4+ Own Rent Other less than 3 months 3 months to 1 year 1 year to 3 years 3 years to 5 years more than 5 years Dwelling Residency Years Living in Current Home GenKnow TranKnow51  Table 4.3 Environmental Knowledge Questions and Responses Q  General Correct % 1 Clothes Washing powder causes water pollution mainly because it contains:     Phosphorus 28% 2 Which of the following statement about ecological footprint is not correct?     Using this assessment, it is possible to estimate how much of the Earth (or how many planet Earths) it would take to support humanity if everybody lived a given lifestyle. 15% 3 The carbon of "carbon emissions" refers to:     Carbon dioxide 59% 4 Which of the following waste materials can not be recycled?     Expired drugs 64% 5 Organic food  means:     Food produced without using synthetic pesticides, hormones, food additives and other production 67% 6 Which country is the world's largest oil consumer?     United States 54% 7 Which of the following statements about greenhouse effect is incorrect?     Increased concentration of carbon dioxide in recent decades is a natural occurring process, not as a result of human activity. 46% 8 Biodiversity means:     Diversity of Biological species, ecosystem and gene multiplicity 56%   Transportation   9 Which of the following transportation method create the least pollution?     Cycling 87% 10 Which of the following transport sector has the highest energy consumption in Canada?     Highways 54% 11 Which of the following strategy does not reduce CO2 emission in urban transport?     Reduce public transportation services 32% 12 What percentage of British Columbia's greenhouse gas emissions come from Metro Vancouver cars?     35% 42% 13 Which of the following type of fuel produce the most carbon dioxide emissions?     Petroleum-based fossil-fuels 50% 14 What percentage of the world total CO2 emission does transport sector account for?     25% 31%    52  4.3 Environmental Attitudes  The environmental attitudes statements that students responded to are listed in  53  Table 4.4 along with their corresponding Likert scale results.  As shown in  54  Table 4.4, environmental attitudes can be parsed into four different attributes; including general environmental problems, general environmental problems with government intervention, environmental issues relating to transportation, and environmental issues related to transportation with government intervention.  While all these attributes relate to the larger environmental attitudes factor, they each represent a distinct concept and possibly interact with other factors in varied ways.  In section 4.3.1, factor analysis is used to confirm the appropriate grouping of statements into multiple factors.  In general the students in this study have attitudes that are protective of the environment. The average environmental attitudes score was 3.4 and 75% of students have positive environmental attitudes.  Four indicators on a scale of 1 to 5 (x1, x2, x3 and x4) were used to quantify the students’ general environmental attitudes.  Indicator x1, “being environmentally responsible is important to me”, served as an overall assessment and 76% of the respondents agreed. Indicators x2, x3 and x4 probed students’ environmental attitudes in terms of the compromise between environmental protection and other factors: 23% agreed that we should not have to be inconvenienced in our daily lives in order to protect the environment; 11% thought that he/she did not need to protect the environment if people around him/her did not pay attention to the environment; and 26% said the economic development in Canada is more important than environmental protection.    Three indicators (x5, x6 and x7) were used to quantify the students’ general environmental attitudes toward government intervention.  While 23% stated that “the government shouldn’t force people to change in order to protect the environment”, 64% agreed that “the government should take more lead in protecting the environment even if people don’t like it”, and 67% agreed that “the government should legislate on categorizing wastes into recyclable categories”.        Four indicators (x8, x9, x10 and x11) were used to quantify the students’ environmental attitudes towards transportation issues: 20% thought that “it is important to build more roads to reduce congestion even if it will increase pollution”; 62% believed that “we should reduce car usage to reduce pollution”; 66% said that “in order to protect the environment, I would 55  cycle and walk more because these modes of transportation do not pollute”; and 30% agreed that “the purchase and use of private cars is freedom of choice; I will buy and use them even if it will damage the environment”.  Four indicators (x12, x13, x14 and x15) were used to quantify the students’ environmental attitudes towards government intervention on transportation issues. 55% stated that “the government should provide incentives to people who travel by electric vehicles, public transport, bicycle or on foot.  44% do not think that the government should encourage the use of cars in the city to stimulate economic growth through increase consumption.  However, only 22% agreed that car users should pay higher taxes for the sake of the environment, and only 24% agreed that the government should raise the price of gasoline to reduce congestion and air pollution.      56  Table 4.4 Responses to Environmental Attitudes Statement  Survey Statement Strongly Agree Moderately Agree Neutral Moderately Disagree Strongly Disagree General environmental issues         3.58 x1 Being environmentally responsible is important to me 29% 47% 21% 2% 1% X2 We should not have to bring inconvenience to our daily lives in order to protect the environment 7% 19% 34% 28% 12% X3 I do not need to protect the environment if people around me do not pay attention to the environment 4% 7% 18% 28% 42% X4 In the current state of Canada, economic development is more important than the protection of the environment 8% 18% 37% 24% 13% Government interventions on environmental issues         3.66 X5 The government shouldn’t force people to change in order to protect the environment 7% 16% 36% 29% 11% X6 The government should take more lead in protecting the environment, even if people don't like it 28% 35% 26% 7% 3% X7 The government should legislate on categorizing wastes into recyclable categories 36% 32% 26% 4% 2% Transportation related environmental issues         3.30 X8 It is important to build more roads to reduce congestion even if it will increase pollution 6% 15% 37% 28% 15% X9 We should reduce car usage to reduce pollution 31% 32% 23% 9% 5% x10 In order to protect the environment, I would cycle and walk more, because these mode do not pollute 32% 34% 24% 8% 3% x11 Purchase and use of private cars is freedom of choice, I will buy and use even if it will damage the environment, even if people don't like it 10% 21% 44% 16% 9% Government interventions on environmental issues related to transportation         3.10 X12 For the sake of the environment, car users should pay higher taxes 6% 17% 34% 24% 20% X13 The government should raise the price of gasoline to reduce congestion and air pollution 8% 16% 28% 25% 23% X14 The government should encourage the use of cars in the city to stimulate economic growth through increase in consumption 7% 9% 40% 26% 18% X15 The Government should provide incentives to people  who travel by electric vehicles, public transport, bicycle or on foot 27% 29% 35% 5% 4% 57  4.3.1 Factor Analysis  Confirmatory Factor Analysis (CFA) is used to validate the hypothesized groupings.  Since there were strong beliefs as to which indicators could be mapped into which factors, exploratory factor analysis was not necessary.  In the first iteration of CFA, the goodness of fit measures were not acceptable (see Table 4.5).  Cronbach’s alpha for all four environmental attitudes factors are below 0.60, suggesting low internal consistency among the indicators.  To improve the fit and internal consistency, the fifteen indicators are organized into two main factors. X1 to x7 are grouped into general environmental attitudes (GENATT) and x8 to x 15 are grouped into transportation-specific environmental attitudes (TRANATT).  In the second iteration of CFA, the goodness of fit measures were still not acceptable (see   58  Table 4.6), but the Cronbach’s Alpha was around 0.70 for both GENATT and TRANATT, suggesting a reasonable level of internal consistency among the indicators.   Table 4.5 Goodness of Fit Results for Confirmatory Factor Analysis (1st model) Overall Model Fit Observations 937 Chi-Square 953.5 Degree of Freedom 84 CFI 0.690 TLI 0.613 RMSEA 0.105 90% CI of RMSEA 0.099~0.111 SRMR 0.08    59  Table 4.6 Goodness of Fit Results for Confirmatory Factor Analysis (2nd model)  Overall Model Fit Observations 937 Chi-Square 976.5 Degree of Freedom 89 CFI 0.684 TLI 0.627 RMSEA 0.103 90% CI of RMSEA 0.097~0.109 SRMR 0.081  To improve the fit, the correlation matrix is examined to determine if any indicators could be eliminated (Table 4.7).   x14 is removed from TRANATT since it has low correlation with the other indicators.   The Cronbach’s alpha for TRANATT changed from 0.69 to 0.68 after removing x14.  The next step in improving the goodness of fit measures was to use modification indices to identify indicators whose error terms might be correlated.  There are eleven indicator pairs that were predicted to improve the fit the most by allowing for correlation among error terms.  This correlation suggests that each indicator is being affected by a third unmeasured variable.  For instance, one pair of correlated indicators was “for the sake of the environment, car users should pay higher taxes” and “the government should raise the price of gasoline to reduce congestion and air pollution”.  Both indicators suggest car users should pay higher prices in order to reduce car usage.  Allowing the error terms to be correlated accounts for this similarity and improves the model fit.  Final CFA goodness of fit results are shown in Table 4.8 with the detailed results shown in Table 4.9.    CFI statistics are greater than 0.9, RMSEA statistic less than 0.06 and SRMR statistic less than 0.05 providing a good indication that the model has a reasonable fit.      60  Table 4.7 Correlation Matrix    x1 x2 x3 x4 x5 x6 x7 x8 x9 x10 x11 x12 x13 x14 x15 x1 1.00                             x2 0.13 1.00                           x3 0.29 0.33 1.00                         x4 0.20 0.27 0.34 1.00                       x5 0.10 0.41 0.24 0.27 1.00                     x6 0.31 0.22 0.27 0.23 0.20 1.00                   x7 0.26 0.16 0.22 0.18 0.11 0.39 1.00                 x8 0.09 0.27 0.29 0.29 0.23 0.18 0.09 1.00               x9 0.28 0.16 0.24 0.12 0.14 0.34 0.20 0.21 1.00             x10 0.40 0.14 0.29 0.15 0.12 0.35 0.22 0.12 0.40 1.00           x11 0.20 0.24 0.28 0.32 0.27 0.21 0.06 0.24 0.20 0.22 1.00         x12 0.36 0.14 0.05 0.18 0.17 0.24 0.14 0.12 0.27 0.32 0.24 1.00       x13 0.25 0.05 0.00 0.05 0.06 0.23 0.09 0.07 0.32 0.28 0.16 0.52 1.00     x14 0.12 0.32 0.43 0.35 0.33 0.24 0.20 0.38 0.17 0.17 0.24 0.08 -0.01 1.00   x15 0.19 0.16 0.18 0.11 0.10 0.32 0.26 0.09 0.26 0.29 0.14 0.17 0.11 0.14 1.00  Table 4.8 Goodness of Fit Results for Confirmatory Factor Analysis (3rd model)  Overall Model Fit Observations 937 Chi-Square 270.2 Degree of Freedom 65 CFI 0.917 TLI 0.883 RMSEA 0.058 90% CI of RMSEA 0.051~0.065 SRMR 0.049      61  Table 4.9 Confirmatory Factor Analysis Results for Environmental Attitudes  Factor Indicator Estimate S.E. Est./S.E. P-Value GENATT Being environmentally responsible is important to me 0.491 0.031 16.052 0.000  We should not have to bring inconvenience to our daily lives in order to protect the environment 0.422 0.032 12.989 0.000  I do not need to protect the environment if people around me do not pay attention to the environment 0.633 0.03 20.93 0.000  In the current state of Canada, economic development is more important than the protection of the environment 0.437 0.032 13.684 0.000  The government shouldn’t force people to change in order to protect the environment 0.361 0.034 10.615 0.000  The government should take more lead in protecting the environment, even if people don't like it 0.624 0.031 20.039 0.000   The government should legislate on categorizing wastes into recyclable categories 0.384 0.035 10.975 0.000 TRANATT It is important to build more roads to reduce congestion even if it will increase pollution 0.37 0.034 10.943 0.000  We should reduce car usage to reduce pollution 0.557 0.03 18.804 0.000  In order to protect the environment, I would cycle and walk more, because these mode do not pollute 0.593 0.028 20.868 0.000  Purchase and use of private cars is freedom of choice, I will buy and use even if it will damage the environment, even if people don't like it 0.453 0.032 14.176 0.000  For the sake of the environment, car users should pay higher taxes 0.476 0.033 14.643 0.000  The government should raise the price of gasoline to reduce congestion and air pollution 0.356 0.037 9.661 0.000   The Government should provide incentives to people  who travel by electric vehicles, public transport, bicycle or on foot 0.419 0.033 12.717 0.000   4.3.2 Determinants of Environmental Attitudes Figure 4.3 analyses general and transportation-specific environmental attitudes by demographic and socioeconomic variables. Overall, the students have more positive general attitudes than transportation-specific environmental attitudes.  The average GENATT score was 62  3.6 and the average TRANATT score was 3.2.   A series of ANOVA tests, reported in Appendix B, found significant differences between GENATT and TRANATT for the whole sample at the 5% level, significant differences in GENATT by income and driver’s license status, and significant differences in TRANATT by household driver licenses at the 5% level.  For example, students in mid-income households have more positive general environmental attitudes than students in low and high income households, but the income difference did not influence the students’ transportation-specific environmental attitudes.  For both general and transportation-specific environmental attitudes, there is no significant difference between grade 8 and 12.  It suggests that the students’ environmental attitudes did not change during their education in high school. Figure 4.3 Average levels of GENATT and TRANATT    1.0 2.0 3.0 4.0 5.0 All Male Female 8 9 10 11 12 Yes No Yes No Gender Grade Driver License Own Car GenAtt TranAtt63         1.0 2.0 3.0 4.0 5.0 Lo w M e d H ig h 0 1 2 3 4 + 0 1 2 3 + 2 3 4 5 6 + 0 1 2 3 4 + Household Income Household License Household Car Household People Household Children GenAtt TranAtt 1.0 2.0 3.0 4.0 5.0 0 1 2 3 4+ Own Rent Other less than 3 months 3 months to 1 year 1 year to 3 years 3 years to 5 years more than 5 years Dwelling Residency Years Living in Current Home GenAtt TranAtt64  4.4 Travel Behavior The students were asked to report their school trips for the previous school day.  Of the different modes to get to school, walking is the main utilized mode, which accounts for 36% of all journeys (Figure 4.4).  35% of students travelled to school by car, with 12% as car driver and 23% as car passenger.  23% took public transit or school bus and only 3% of student biked to school.   The travel modes used for after school trips are slightly different. Fewer students are being driven and shifted to increase in walk mode and public transit.  In contrast, students’ ideal travel mode differs from their actual travel mode.  23% of the students prefer to drive and 11% prefer to bike to school; however, only 11% actually drive and 3% actually bike.  On the other hand, fewer students prefer to take public transit or walk to school (7% less for each mode)  Looking into gender comparison, more female students are being driven and fewer bike to school compared to male students.  This is consistent with the results of students’ ideal mode of travel where more female students prefer to be car passengers and more male students prefer to bike to school.  In addition, male students have a stronger desire to drive than female students.    Looking into grade comparison, grade 12 students are more likely to drive, where approximately one-fifth of them drive to/from school.  While this is expected, as citizens in British Columbia are required to be 16 or older to qualify for a learner’s license and need to stay in the learner’s stage for at least 12 months in order to be qualified to take the road test, roughly 20% switched to driving as soon as the option was available.  This suggests the preference to driving might actually be higher as not all grade 12 students have the option to drive, depending on access to vehicle and whether they have successfully obtained the driver license.    65  An interesting observation is the decline in the share of after school trips made by car passenger and increase in the share of after school trips made by transit across the grades.  For example, the share of after school trips for grade 8 students made by car and by public transit is 23% and 7%, respectively.  In contrast, the share of after school trips for grade 12 students made by car and by public transit is 7% and 20%, respectively.    Trip purposes of trips made directly after school are shown in Figure 4.5.   The majority of students (65%) return home after school.  Trips for social and recreational purposes make up the majority (10%) of the after-school trips away from home.  Personal business, dining out, shopping and work purposes together make up 15% of trips  There is a decline in the percentage of trips made directly home for students in higher grades.  This implies that the tendency to engage in out-of-home activities after school increases with age.    The average travel time to school is 16 minutes, with 52% of the students travelled for less than 15 minutes and 84% travelled for less than 30 minutes ( Figure 4.6).  The average travel time to school by walking, biking and car is 14 minutes. The travel time is similar for walking, biking and car suggest that the students who travel to school by car live further away from the school.  Travel to school by public transit takes 28 minutes, which is almost twice as much the time compared to the other modes.    The average travel time after school is 27 minutes, which is much longer when compared to the average travel time to school, with 28% of students travel for more than 30 minutes.  On average, students who do not go directly home take almost twice the time than the students that go directly home after school (41 minutes compared to 21 minutes).   This could be that the distance to the students’ after school activities are further away from their home.  It is also interesting to note that the travel time for the after school trip generally increases with age.  This finding agrees with the earlier finding that older students are more likely to engage in out- of-home activities after school and therefore travel for a longer time.  Figure 4.4 Travel Mode to/from School 66    Figure 4.5 Activities after School       Figure 4.6 Travel Time to/from School 0% 5% 10% 15% 20% 25% 30% 35% 40% 45% Car Driver Car Passenger Public Transit School Bus Walk Bike Other To Sch After Sch Ideal Mode 65% 3% 10% 3% 5% 5% 10% Go Home Personal Business Recreational/Social Dining/Restaurant Shopping Work Other67    The students were asked if they travel to school with a companion.  Table 4.10 shows the percentage of school trips in which students traveled by themselves, by grades and gender.   42% of students travelled to school by themselves, and the others mostly travelled with their classmates and siblings.  Only 3% travelled to school with their parents.  In general, older students (grade 10-12) tend to make a larger share of school trips unaccompanied than younger students (grade 8-9).  As older students are more likely to have after school activities, the share of trips made alone is expected to be greater.  In addition to the differences across grades, there are differences between genders.  On average, male students made more school trips alone than females, where 47% of male students traveling unaccompanied but only 39% of female students traveling unaccompanied.       Table 4.10 Travel Alone to School by Age and Sex 0% 5% 10% 15% 20% 25% <5 5-9 10-14 15-19 20-29 30-39 40-59 60> To Sch After Sch68  Grades Male Female Total 8 31% 32% 32% 9 48% 26% 32% 10 47% 44% 45% 11 54% 43% 47% 12 55% 38% 44% Total 47% 39% 42%  The students were asked about the decision making process of their commuting mode choices to school. 61% of students were the sole decision makers in their commuting mode choice while the rest having their parents and guardians as decision maker.  Male students and higher grade students have a higher tendency to make their own commuting decision.  The students were further asked to identify the factors that influence their current travel mode choice to school.  As presented in Figure 4.7, convenience is the most important factor, followed by environmental and cost.  Personal safety and road safety is not the driver of the mode choice to school because the majority of students rate their daily commute to school to be either very safe or safe.  Figure 4.7 Factors that Influence the Travel Mode Choice to School    0 50 100 150 200 250 300 Cost Reliable Comfortable Convenient Environmental Health Road Safety Personal Safety Social69  Besides their school trips, the students were also asked to report the average use of each travel mode. As shown in Table 4.11, walking is the most commonly used mode of transport, with 86% of students walking at least once a week and 63% at least 5 days a week.  On the other hand, biking is one of the least used modes of transport, with only 7% of students who bikes at least 5 days a week and 20% at least once a week.  41% of students had never biked before.    An interesting observation is the difference between the frequency of bike use and the travel mode share of bike to school.  7% of students bike at least 5 days a week while only 3% of students bike to and from school. An explanation could be that these results are affected by the seasonal variation of cycling which is much more dependent on weather than are other modes.  Students were asked to report their school trip for the previous school day which is during the winter months.  However, cyclists may respond the average frequency over the whole year.  This suggests the mode share to and from school could be different if the survey is conducted during the warmer months.  Another explanation could be that the students bike for recreational purposes after school or during the weekend.    The car is the second most frequent used mode of transport, followed by walking.  45% of students travelled as car passengers at least 5 days a week and 79% at least once a week.  12% of students drive a car at least 5 days a week and 27 % at least once a week.    In terms of public transportation, buses are more commonly used than the SkyTrain and commuter rail.  While 55% of students use buses at least 5 days a week, only 39% use the SkyTrain and commuter rail at least 5 days a week.  On average, 15% of students have never used of public transportation before.       70  Table 4.11 Percentages of Frequent Use of Modes of Transport   6-7 days a week 5 days a week 3-4 days a week 1-2 days a week once a fortnight once a month less often never Walk 43% 20% 12% 11% 3% 3% 4% 4% Bike 3% 4% 6% 7% 5% 6% 28% 41% Public Transit Buses 16% 13% 10% 16% 9% 9% 15% 12% School Buses 2% 4% 1% 2% 1% 1% 6% 82% Sky Train / Commuter Rail 6% 5% 10% 18% 12% 12% 20% 17% Taxi 2% 1% 2% 3% 3% 3% 20% 66% Car Driver 8% 4% 7% 7% 2% 2% 6% 64% Car Passenger 31% 14% 18% 16% 4% 3% 5% 10%  Table 4.12 shows the modes people use to travel to work in the region, and compares with the modes students use to travel to school.  Car driver has a significantly higher mode share for journey to work compared to the journey to school, but car passenger has a smaller share for journey to work compared to the journey to school.  Among the modes of active transportation, walk mode share is substantially higher for journey to school compares to the journey to work.  However, bike mode share is similar for both types of journey.  Table 4.12 Travel Mode to Work vs Travel Mode to School Travel Mode To Work To School Metro Vancouver City of Vancouver City of Richmond Car Driver 67% 52% 73% 11% Car Passenger 7% 6% 9% 23% Public transit 17% 25% 12% 16% Walk 6% 12% 4% 36% Bike 2% 4% 1% 3% Other 1% 1% 1% 10% (Source: Statistics Canada, 2006 Census)   71  4.5 Comparison between In-class and Lunch Time survey  Of the 937 useable responses, 77% were collected from lunch time surveys and the rest were collected from in-class surveys.  The students that participated in the in-class surveys had a dedicated 30 minutes to fill out the complete version of the questionnaire under teachers’ supervision.  On the other hand, the students that participated in the survey during lunch time filled out the short version of the questionnaire at the cafeteria on a voluntary basis without teachers’ supervision.  Since the implementation method is different, this section provides a comparison between the data collected from the lunch time and in-class survey.    The overall distribution of gender was 40% male and 60% female (Figure 4.8).  However, the in- class survey is more evenly distributed, with 48% male and 52% female.  This indicates the lunch time surveys have over sampled female students.  As reported by the surveyors, female students, in general, are more interested in participating in surveys than male students.    Figure 4.8 Gender by survey type  The distribution of grades for the overall, in-class and lunch time samples are presented in Figure 4.9.  For the in-class surveys, a significant higher proportion of responses were collected 0% 10% 20% 30% 40% 50% 60% 70% Male Female In-Class Lunch Time72  from grade 8 students (36%).  Grade 9, 10 and 12 altogether only represented 41% of the samples.   Of the lunch time surveys, grade 10, 11 and 12 students each represented a quarter of the responses.   Grade 8 and 9 altogether only represented 26% of the sample.   Figure 4.10 reports the average percentage of correct answers for general and transportation- specific environmental knowledge questions by survey type and Figure 4.11 reports the average level of general and transportation-specific environmental attitudes by survey type.  No significant differences in environmental knowledge and attitudes were observed between the samples collected from lunch time surveys and in-class surveys.   Figure 4.9 Grade by survey type           0% 5% 10% 15% 20% 25% 30% 35% 40% 8 9 10 11 12 In-Class Lunch Time73  Figure 4.10 Environmental Knowledge by survey type  Figure 4.11 Environmental Attitudes by survey type       0% 20% 40% 60% 80% 100% General Transportation In-Class Lunch Time 1.00 2.00 3.00 4.00 5.00 General Transportation In-Class Lunch Time74  As shown in Figure 4.12, the in-class samples were more likely to be driven to/from school than the lunch time samples (27% compared to 18%).   On the other hand, the lunch time samples were more likely to walk, take public transit and school bus to/from school (2 to 3% more for each travel mode).  More samples from the in-class surveys are lower grades student; therefore more likely to be driven to/from school by their parents or older siblings.  In terms of decision maker of the school trip, the samples collected in-class and lunch time have similar answers, with approximately 60% decided on their own (Figure 4.13).       In conclusion, the samples collected from the in-class survey have more female students and grade 8 students.  There are no significant differences in environmental knowledge and attitudes between the samples collected from in-class time and lunch time.  This is reasonable since there are no observable differences in environmental knowledge and attitudes by gender and grades.  Lastly, the samples collected from the in-class survey were more likely to be driven to/from school.   Figure 4.12 Travel Mode to/from School by survey type       0% 5% 10% 15% 20% 25% 30% 35% 40% 45% In-Class Lunch Time75  Figure 4.13 Decision Maker by survey type     0% 10% 20% 30% 40% 50% 60% 70% 80% Myself Parents In-Class Lunch Time76  5 The Impact of Environmental Knowledge on Environmental Attitudes 5.1 Introduction Childhood environmental education is generally considered the most beneficial; where early years of learning serve as a foundation for the formation of environmental attitudes.  Many scholars have concluded that unless children develop a sense of respect and care for the environment during their early years, they are at risk of never developing such attitudes later in life (Stapp 1978; Tilbury 1994; Wilson 1994). Cohen (1984), for example, documented that if children develop negative attitudes toward the environment during their early years, such attitudes are likely to become deeply entrenched.  While many studies have directly addressed the relationship between environmental knowledge and attitudes of youth (Ramsey and Rickson 1976; Jaus 1982; Mangas, Martinez et al. 1997; Bradley, Waliczek et al. 1999; Kuhlemeier, Bergh et al. 1999; Makki, Abd-El-Khalick et al. 2003; Negev, Sagy et al. 2008; He, Hong et al. 2011), no studies has focused on their environmental knowledge and attitudes specifically related to transportation.   5.2 Model Structures and Estimations This chapter examines the impact of environmental knowledge on environmental attitudes in the context of both general and transportation issues using structural equation model.  Figure 5.1 specifies the hypothesized relationships among the factors.  The two environmental knowledge factors GENKNOW and TRANKNOW are assumed to affect the two attitude factors GENATT and TRANATT.  GENKNOW and TRANKNOW are correlated with each other, as are GENATT and TRANATT.  The model consists of three sets of equations:  measurement equations that connect the latent variables to their corresponding indicators; structural equations that represent the relationship between knowledge factors and attitude factors; and structural equations that quantify the impact of demographic and socioeconomic variables on knowledge and attitudes.   77  Table 5.1 summarizes the overall goodness-of-fit statistics.  Though CFI of 0.884 is slightly below 0.9, RMSEA and SRMR are below 0.05; in particular the full 90% confidence interval 0.039~0.046 falls below 0.05 so the overall data fit is acceptable. Table 5.2 shows the standardized factor loadings and t-statistics for the measurement equation for the two latent variables (GENATT and TRANATT).  All measurement equations are significant at 1% level. Figure 5.1 Specification of the Structural Equation Model  Environmental Attitude on Transportation EE General Environmental Attitude 0.852 0.311 0.174 0.316 0.293 0.165 General Environmental Knowledge Transportation Environmental Knowledge x7x1 x3 x5 x2 x4 x6 x15x8 x10 x12 x9 x11 x13 xx Household Income Household License YearHomeResidencyDwellingChildren Household Car Household People Gender Grade Driver License Own Car    5.3 Model Results Table 5.3 reports the estimated relationships between environmental knowledge and attitudes.  All the relationships are significant and are in the expected direction.  Students with better environmental knowledge hold higher levels of pro-environmental attitudes.  Both general knowledge and transportation-specific knowledge contribute to stronger pro-environmental 78  attitudes.  While general and transportation-specific attitudes are highly correlated (0.852), the correlation between general knowledge and transportation knowledge is significant but much weaker (0.316) in comparison. Table 5.1 Goodness of Fit of Statistics for SEM Overall Mode Fit Observations 937 Chi-Square 650.5 Degree of Freedom 269 CFI 0.884 TLI 0.845 RMSEA 0.037 90% CI of RMSEA 0.033~0.041 SRMR 0.036   Table 5.2 Measurement Equations for the Latent Variables   Latent Factors GENATT TRANATT Measured by Estimate Est./S.E. Estimate Est./S.E. x1 0.484 16.0 n.a. x4 0.441 13.9 n.a. x8 0.634 22.3 n.a. x11 0.444 14.1 n.a. x3 0.379 11.4 n.a. x14 0.591 19.2 n.a. x15 0.401 11.9 n.a. x6 n.a. 0.384 11.5 x7 n.a. 0.555 18.8 x10 n.a. 0.587 20.5 x13 n.a. 0.466 14.8 x2 n.a. 0.474 14.7 x5 n.a. 0.347 9.5 x12 n.a. 0.405 12.0   79  Table 5.3 Structural Equations between Environmental Knowledge and Attitudes Independent Dependent Estimate Est./S.E. GENKNOW GENATT 0.308 8.2 TRANKNOW GENATT 0.165 4.3 GENKNOW TRANATT 0.288 7.2 TRANKNOW TRANATT 0.174 4.3 Correlation between GENATT and TRANATT 0.851 22.8 Correlation between GENKNOW and TRANKNOW 0.317 10.5 Note: Bold numbers are significant at 10% level.  Table 5.4 reports the estimated direct impact of the demographic and socioeconomic variables on environmental knowledge and attitudes.  Six independent variables are significant to the attitudinal factors.  Females have more positive general and transportation-specific environmental attitudes than male.  This is in line with the Finland study reported by Tikka et al. (2000) that females tended to show more responsibility toward the environment than male.  Household car ownership has a negative impact on both environmental attitudes.  This is reasonable since students’ households with more cars are more likely to travel by car and therefore the students’ attitudes become less pro-environmental.  Owning a personal vehicle has a negative impact on general environmental attitudes, but not on transportation-specific environmental attitudes.   This is a bit surprising as owning a personal vehicle should be more likely to have an impact on transportation-specific attitudes than general attitudes.   Dwelling type was treated as a series of dummy variables with house being the reference.  Of the four dwelling types, only duplex turns out to be significant and have a negative impact on general environmental attitudes when compared with house.   Grade level also has a negative impact on general environmental attitudes.  However, as mentioned in Section 4.3, this study did not find the students’ environmental attitudes increasing from grade to grade.  Income was treated as dummy variables with medium income being the reference.  Students in high-income households have more negative transportation-specific attitudes than students in medium income household.  The significant impact of driver license status on GENATT, and number of 80  household driver licenses on TRANATT identified in ANOVA become insignificant after multiple independent variables are included.   Six independent variables are significant to the knowledge factors.  In contrast to environmental attitudes, grade level has a positive impact on both general and transportation- specific environmental knowledge.  This is expected as students in higher grades are expected to have better environmental knowledge.  There is a positive correlation between general environmental knowledge and the number of years a youth is living in his/her current home.  This is consistent with their significance in the bivariate ANOVA analysis.  Household income was treated as dummy variables with medium income being the reference. Students in mid- income households are less knowledgeable on transportation-specific environmental issues than students in low and high income household.  On the other hand, three variables behave differently.  Students owning a personal vehicle have a negative impact on both general and transportation-specific environmental knowledge.  This is reasonable since these students have less environmental knowledge and might not realize the negative impacts in owning a vehicle.  Similarly, students having a driver license have a negative impact on general environmental knowledge.  Students living in high rises also have a negative impact on transportation-specific environmental knowledge compared to living in houses.  This is unexpected as students living in a more compact urban setting are more likely to use sustainable transportation and therefore are expected to have better transportation- specific environmental knowledge.   Table 5.5 reported direct, indirect and total effects of the demographic and socioeconomic variables on GENATT and TRANATT via GENKNOW and TRANKNOW are significant at 10% level.  Owning a personal vehicle has a negative indirect effect on both attitudes via environmental knowledge, while it has no direct effect on transportation-specific attitudes.  Grade level and the number of years living at the current home show indirect effects on both attitudes via knowledge.  Although grade level has a positive significant indirect impact on transportation- specific attitudes, it shows a significant negative direct impact on transportation-specific 81  attitudes which neutralizes the total effect.   Living in high rises does not show significant direct effect, but it shows positive indirect effect on attitudes. Table 5.4 Structural Equations for the Determinants of Environmental Knowledge and Attitudes   Dependent Variables   GENATT TRANATT GENKNOW TRANKNOW Independent Variables Estimate Est./S.E. Estimate Est./S.E. Estimate Est./S.E. Estimate Est./S.E. MALE -0.072 -1.9 -0.091 -2.3 0.031 0.9 -0.028 -0.8 GRADE -0.038 -0.9 -0.131 -2.8 0.146 3.8 0.070 1.8 HOUSECAR -0.159 -3.3 -0.139 -2.8 -0.047 -1.1 -0.012 -0.3 OWNCAR -0.151 -3.8 -0.030 -0.7 -0.102 -2.8 -0.074 -2.0 HOUSEPPL -0.054 -0.9 0.016 0.2 -0.019 -0.3 -0.060 -1.0 CHILDREN 0.001 0.0 0.025 0.4 0.003 0.1 -0.008 -0.2 HOUSELIC 0.076 1.4 -0.009 -0.2 -0.038 -0.8 0.020 0.4 DRILIC -0.028 -0.5 -0.062 -1.2 -0.097 -2.1 0.012 0.3 YEARHOME 0.025 0.6 0.003 0.1 0.138 3.9 0.046 1.3 DUPLEX -0.088 -2.3 -0.057 -1.4 0.053 1.5 -0.023 -0.6 TOWNHOUS -0.052 -1.3 -0.063 -1.5 0.016 0.4 -0.054 -1.4 LOWRISE 0.011 0.3 -0.003 -0.1 0.016 0.4 -0.039 -1.0 HIGHRISE -0.035 -0.9 -0.013 -0.3 -0.044 -1.2 -0.092 -2.6 RENT -0.07 -1.6 -0.038 -0.8 0.018 0.5 0.027 0.7 LOWIN -0.067 -0.9 -0.062 -0.8 -0.06 -0.9 0.147 2.1 HIGHIN -0.079 -1.0 -0.154 -1.9 0.003 0.0 0.166 2.3 Note: Bold numbers are significant at 10% level    82  Table 5.5 Direct and Indirect Effects on General Environmental Attitudes and Transportation- Specific Environmental Attitudes  Explanatory Variables Direct Effect Total Indirect Effect Total Effect Estimate Est./S.E. Estimate Est./S.E. Estimate Est./S.E. GENATT   Male -0.072 -1.9 0.005 0.37 -0.067 -1.72 Grade -0.038 -0.9 0.057 3.46 0.019 0.42 HouseCar -0.159 -3.3 -0.016 -0.95 -0.176 -3.51 OwnCar -0.151 -3.8 -0.044 -2.96 -0.195 -4.69 YearHome 0.025 0.6 0.05 3.35 0.075 1.79 Duplex -0.088 -2.3 0.013 0.89 -0.075 -1.84 HighRise -0.035 -0.9 -0.029 -1.97 -0.064 -1.54   TRANATT   Male -0.091 -2.31 0.004 0.32 -0.087 -2.10 Grade -0.131 -2.83 0.054 3.37 -0.077 -1.60 HouseCar -0.139 -2.75 -0.016 -0.93 -0.154 -2.93 OwnCar -0.03 -0.70 -0.042 -2.92 -0.072 -1.63 YearHome 0.003 0.07 0.048 3.26 0.051 1.15 HighRise -0.013 -0.31 -0.029 -2.02 -0.042 -0.95 High Income -0.154 -1.91 0.030 1.09 -0.124 -1.46 Note: Bold numbers are significant at 10% level.  Despite the significant relationships between environmental knowledge and attitudes, the r- square values of GENATT and TRANATT are only 0.263 and 0.235, respectively, suggesting that environmental knowledge, together with socioeconomic and demographic variables, can explain environmental attitudes only to a limited extent.  To demonstrate environmental knowledge’s capacity to explain environmental attitudes, two auxiliary SEMs (aux1 and aux 2) are estimated and their r-squares are compared to that in the main models in Table 5.6.  The model with only demographic and socioeconomic variables as independent variables can explain a minimal amount of variation in GENATT and TRANATT (11.5% and 9.9%).   Environmental knowledge, on the other hand, can explain 17.8% and 14.7% of the variation in GENATT and TRANATT, respectively.   An auxiliary model is developed to demonstrate the variation in knowledge and attitudes among students from different grades.  This model treats GRADE as a series of dummy variables with Grade 8 being the reference.  As mentioned in section 4.3, students’ environmental 83  knowledge increased slightly and environmental attitudes did not increase from grade to grade.  Table 5.7 confirms the earlier discussion after controlling for demographic and socioeconomic variables.  For general environmental attitudes, there is no significant difference between grade 8 and the other grades.  For transportation-specific attitudes, there is significant difference for grade 9 and 12 compared to grade 8.  Overall, the impact of grade level on transportation- specific attitudes is non-linear.   Conversely, grade has a positive impact on both general and transportation-specific knowledge.  There is significant increase in general knowledge for grade 10 to 12 students compared to grade 8 students, while there is only significant increase in transportation-specific knowledge for grade 12 students compared grade 8 students. This is logical because grade 12 students are of legal age to drive.   Table 5.6 Comparison of Explanatory Power between Models Models R-Square GENATT TRANATT SEM aux 1: model with only demographic and socioeconomic variables 0.115 0.099 SEM aux 2: model with only environmental knowledge 0.178 0.147 SEM: main: model with demographic, socioeconomic variables and environmental knowledge 0.263 0.235  Table 5.7 Structural Equations for the Auxiliary Model  Dependent Variables   GENATT TRANATT GENKNOW TRANKNOW Independent Variables Estimate Est./S.E. Estimate Est./S.E. Estimate Est./S.E. Estimate Est./S.E. GRADE 9 0.066 1.5 0.084 1.8 -0.012 -0.3 0.026 0.6 GRADE 10 0.038 0.8 0.040 0.8 0.093 2.2 0.026 0.6 GRADE 11 0.017 0.3 -0.052 -1.0 0.112 2.5 0.067 1.4 GRADE 12 -0.046 -0.8 -0.153 -2.6 0.162 3.3 0.085 1.7 Note: Bold numbers are significant at 10% level.   84  5.4 Summary In summary, environmental knowledge is positively associated with pro-environmental attitudes.  Students who have better knowledge of environmental issues also hold higher levels of pro-environmental attitudes, but the capacity of environmental knowledge to explain attitudes is limited.   Environmental knowledge is an important determining factor of attitudes, but not a dominant one; many other factors also influence environmental attitudes.  This suggests that knowledge should be emphasized in education in order to improve environmental attitudes, but the impact of environmental knowledge education alone should not be exaggerated.   The results have confirmed that grade level and car ownership status have significant impact on both general and transportation-specific environmental knowledge.  Driver license status and the years living in current home have significant impact on general knowledge.  Dwelling type and income level have significant impact on transportation-specific environmental knowledge.  For example, students in higher grade levels have better environmental knowledge.   The results also found gender, household car ownership, personal car ownership, and dwelling type have significant impacts on general environmental attitudes.  Gender, grade level, household car ownership, income level have significant impacts on transportation-specific environmental attitudes.  For example, females are more environmental than males on both general and transportation-specific issues.   Although grade level has a negative impact on transportation-specific attitudes, the impact is non-linear as demonstrated by the auxiliary model that treated GRADE as a series of dummy variables.  In addition, grade level shows significant positive impact mediating through knowledge which neutralizes the total effect.       85  6 Disentangle Casual Direction between Environmental Attitudes and Travel Behavior 6.1 Introduction The relationship between environmental attitudes and travel behavior has long been recognized in prior studies.  However, there are competing hypotheses regarding the relationship between attitudes and behavior. Attitudes are formed through experiences as a result of behavior; on the contrary, attitudes prompt certain types of behavior. The task to tease out the exact causal direction between attitudes and behavior remains difficult.  This study provides an opportunity to disentangle the direction of the mutual causalities.    In surveys, students were asked about their decision making process of commuting mode choices: whether the students make their own commuting decisions or if their parents make the decisions for them.  For those whose parents make travel decisions, the causality between students’ attitudes and their behavior cannot be realized.  As a result, we can be more certain in making the assumption that it is their behavior that influences attitudes and not the other way around.  Given the defined causal direction, the strength of such influence can be quantified using structural equation models after controlling demographic and socioeconomic variables.  There is indeed a possibility that students’ attitudes influence their parents’ attitudes, which will in turn influence the parents’ decision for the youth’s travel choices; however, the causal path is more indirect and much weaker. Even though the direction from behavior to attitudes is almost certain for the group where parents are decision makers, multiple group analysis is performed to compare the impact of behavior on attitudes between the two groups of students.  Although the impact of behavior on attitudes cannot be determined exactly, the results from the following analysis explain the total combined effect of the mutual causalities.  Parents who make travel decisions for students are denoted as PARENT-DECISION-MAKER, while students who make their own travel decisions are denoted as STUDENT-DECISION-MAKER.  86  This chapter begins with a comparison of the travel behavior between PARENT-DECISION- MAKER and STUDENT-DECISION-MAKER; followed by structural equation models results. 6.2 Comparison of Travel Behavior between PARENT-DECISION-MAKER and STUDENT-DECISION-MAKER Students were asked to report their travel mode to/from school for the previous school day.  As shown in Figure 6.1 and   Figure 6.2, students with parents as decision makers are more likely to be driven to/from school and are less likely to walk to school compared to students who make their own decision.  The substantial differences between the two groups are likely due to parents’ concern about safety and convenience and results in the preference to drive their children to school (DiGuiseppi, Roberts et al. 1998; McMillan 2003; Rhoulac 2005). As mentioned in section 4.4, some students shifted from being car passengers in the morning to walking and taking public transit in the afternoon.  The figure below shows a clearer picture that majority of students who travel as car passengers are those with parents as decision maker.      Figure 6.1 Comparison of Travel Mode to School  0% 10% 20% 30% 40% 50% Car Driver Car Passenger Public Transit School Bus Taxi Walk Bike Other Student-Decision-Maker Parents-Decision-Maker87    Figure 6.2 Comparison of Travel Mode from School  The frequency of transport modes utilized by parents and students as decision makers are reported in Table 6.1 and Table 6.2.  Similar to travel mode to/from school, students with parents as decision makers are driven by car more frequently and walk less frequently compared to students who are decision makers.   Table 6.1 Frequent Use of Modes of Transport by STUDENT-DECISION-MAKER  6-7 days a week 5 days a week 3-4 days a week 1-2 days a week once a fortnight once a month less often never Walk 51% 21% 10% 9% 3% 2% 2% 3% Bike 4% 4% 6% 8% 4% 6% 28% 38% Public Transit Buses 17% 13% 12% 19% 8% 8% 14% 10% School Buses 1% 3% 1% 2% 1% 1% 6% 84% Sky Train / Commuter Rail 6% 5% 12% 19% 10% 12% 19% 16% Taxi 1% 1% 2% 4% 4% 4% 21% 64% Car Driver 8% 4% 7% 9% 2% 2% 8% 60% Car Passenger 21% 12% 22% 19% 5% 3% 7% 10%  0% 10% 20% 30% 40% 50% Car Driver Car Passenger Public Transit School Bus Taxi Walk Bike Other Student-Decision-Maker Parents-Decision-Maker88    Table 6.2 Percentages of Frequent Use of Modes of Transport by PARENT-DECISION-MAKER  6-7 days a week 5 days a week 3-4 days a week 1-2 days a week once a fortnight once a month less often never Walk 31% 19% 15% 14% 4% 4% 7% 6% Bike 2% 3% 4% 6% 6% 7% 28% 44% Public Transit Buses 12% 13% 8% 13% 11% 11% 17% 16% School Buses 2% 6% 2% 1% 1% 2% 7% 80% Sky Train / Commuter Rail 5% 6% 5% 16% 15% 12% 21% 20% Taxi 2% 1% 2% 2% 2% 3% 19% 69% Car Driver 8% 5% 8% 5% 1% 1% 3% 69% Car Passenger 47% 16% 11% 12% 2% 2% 1% 9%  The travel mode share by parents and students as decision makers are reported in Table 6.3.  Mode share is calculated based on the frequent use of transport modes.  The frequent use of transport modes is converted to the number of trips made for each mode.  For example, walking 5 days a week equates to 10 trips.  This assumes most people will likely utilize the same mode to and from their destination.  Mode share is calculated by the number of trips for a particular mode divided by the total number of trips for all modes.  Walk mode share is the highest for STUDENT-DECISION-MAKER and second highest for PARENT-DECISION-MAKER.  On the other hand, car passenger mode share is the highest for PARENT-DECISION-MAKER and second highest for STUDENT-DECISION-MAKER.   Table 6.3 Comparison of Travel Mode Share  Student-Decision-Maker Parents-Decision-Maker Walk 39% 28% Bike 6% 3% Public Transit Buses 16% 12% School Buses 1% 3% Sky Train / Commuter Rail 8% 6% Taxi 1% 1% 89  Car Driver 7% 6% Car Passenger 22% 41%  6.3 The Impact of Travel Behavior on Environmental Attitudes Table 6.4 specifies the overall model structure for this chapter.  Model a and b are based on multiple group analysis that compare the relationship between the two groups of students; whereas model c and d examine the students as a whole.  In model a and c, environmental attitudes are treated as a single factor (ENVATT).  In model b and d, environmental attitudes are separated into general attitudes (GENATT) and transportation-specific attitudes (TRANATT).   Table 6.4 Model Structure  Travel Behavior Environmental Attitudes STUDENT DECISION MAKER PARENT DECISION MAKER ALL Travel Mode to School ENVATT 1a 1c GENATT, TRANATT 1b 1d Travel Mode from School ENVATT 2a 2c GENATT, TRANATT 2b 2d Travel Mode Frequency ENVATT 3a 3c GENATT, TRANATT 3b 3d Mode Share ENVATT 4a 4c GENATT, TRANATT 4b 4d  In model a and c, travel behavior is assumed to affect ENVATT.  In model b and d, travel behavior is assumed to affect the two attitude factors GENATT and TRANATT, and the correlations amongst GENATT and TRANATT.  Three aspects of travel behavior are included in the analysis: travel mode to school, travel mode frequency and travel mode share.   The models consist of three sets of equations: measurement equations that connect latent variables to the indicators; structural equations that represent the relationships between attitude factor and travel behavior factors; and structural equations that quantify the impact of demographic and socioeconomic variables on attitude and travel behavior factors.   90  The definitions of the travel behavior factors are as follows: SCHWB: Walk or bike to school SCHCAR: Drive or being driven to school SCHTB: Taking public transit or school bus to school AFTWB: Walk or bike after school AFTCAR: Drive or being driven after school AFTTB: Taking public transit or school bus after school WBFREQ: Frequency of walking and biking TPFREQ: Frequency of using public transit buses and school buses CARFREQ: Frequency of driving and travelling as car passenger WBSHARE: Walking and biking mode share TBSHARE: School bus and public transit buses mode share CARSHARE: Car driver and car passenger mode share  The first series of models examine the impact of travel mode to school on environmental attitudes.  Travel mode is treated as a series of dummy variables with SCHCAR being the reference.  The two travel mode factors SCHWB and SCHTB were assumed to affect the attitude factors.   The second series of models examine the impact of travel mode from school on environmental attitudes.  Travel mode is treated as a series of dummy variables with AFTCAR being the reference.  The two travel mode factors AFTWB and AFTTB were assumed to affect the attitude factors.     The third series of models examine the impact of travel mode frequency on environmental attitudes.  Students are asked to report the average use of eight different types of travel modes: walk, bike, public transit buses, school buses, SkyTrain/commuter rail, taxi, car driver and car passengers.   In order to be consistent with the travel mode to/from school, SkyTrain/commuter rail and taxi are not included in this analysis.  Three travel modes frequency factors (WBFREQ, TBFREQ, and CARFREQ) were assumed to affect the attitude factors.   91  The fourth series of models examines the impact of mode share on environmental attitudes.  Mode share is calculated based on the frequency use of transport modes.  The travel modes are grouped into three categories: active transportation, transit, and car.  WBSHARE and TBSHARE were assumed to affect the attitude factors with CARSHARE being the reference.   Table 6.5 summarizes the overall goodness-of-fit statistics for models that examine the impact of travel behavior factors on environmental attitudes.  The standardized factor loadings and the t-statistics for the measurement equation of the latent variables for model 1a and 1b are presented in Appendix C.  All measurement equations are significant at 1% level. Table 6.6 reports the estimated relationships between environmental attitudes and travel mode to school.  In model 1a, where students’ parents are decision makers, SCHWB and SCHTB have significant and positive impact compared to SCHCAR.  In contrast, where students make their own decision, SCHWB and SCHTB have an insignificant impact on ENVATT.  In model 1c, all the relationships are significant and are in the expected direction. In model 1b, where students’ parents are decision makers, all the relationships are significant except for SCHWB on GENATT. Students who take public transit or school bus have more positive general environmental than students who travel to school by car.  However, there are no differences in general attitudes between the students who walk or bike to school and students who travel to school by car.   On the other hand, students who walk or bike to school, and students who take public transit or school bus have more positive transportation-specific attitudes than students who travel to school by car.  Again, for those students that are their own decision maker, SCHWB and SCHTB have an insignificant impact on GENATT and TRANATT.  For both groups of students, GENATT and TRANATT are highly correlated.  In model 1d, all the relationships are significant and in the expected directions. 92  Table 6.5 Goodness of Fit of Statistics – The Impact of Travel Behavior on Environmental Attitudes  Goodness of Fit   Model 1a Model 1b Model 1c Model 1d Model 2a Model 2b Model 2c Model 2d Observations 812 812 886 886 812 812 886 886 Chi-Square 1214.8 1143 863.3 796.2 1391.9 1310.5 1012.5 935.6 Degree of Freedom 628 588 301 282 628 588 301 282 CFI 0.79 0.802 0.8 0.817 0.745 0.759 0.761 0.781 TLI 0.749 0.746 0.75 0.756 0.695 0.692 0.702 0.707 RMSEA 0.048 0.048 0.046 0.045 0.055 0.055 0.052 0.051 90% CI of RMSEA 0.044~0.052 0.044~0.052 0.042~0.050 0.042~0.049 0.051~0.059 0.051~0.059 0.048~0.055 0.048~0.055 SRMR 0.049 0.047 0.041 0.039 0.051 0.05 0.044 0.042   Model 3a Model 3b Model 3c Model 3d Model 4a Model 4b Model 4c Model 4d Observations 858 858 937 937 854 854 933 933 Chi-Square 1377.8 1294.1 908.3 842.8 1181.2 1103.7 794.5 726.9 Degree of Freedom 760 712 367 344 628 588 301 282 CFI 0.813 0.824 0.832 0.845 0.809 0.822 0.83 0.847 TLI 0.749 0.748 0.766 0.77 0.772 0.773 0.787 0.795 RMSEA 0.044 0.044 0.04 0.039 0.045 0.045 0.042 0.041 90% CI of RMSEA 0.040~0.047 0.040~0.047 0.036~0.043 0.036~0.043 0.041~0.049 0.041~0.049 0.038~0.045 0.037~0.045 SRMR 0.047 0.045 0.038 0.036 0.049 0.047 0.04 0.038 93  Table 6.6 Structural Equations between Environmental Attitudes and Travel Mode to School Model 1a 1c Decision Maker Students Parents All Independent Dependent Estimate Est./S.E. Estimate Est./S.E. Estimate Est./S.E. SCHWB ENVATT 0.066 1.1 0.143 2.2 0.119 2.9 SCHTB ENVATT  0.053 0.9 0.211 3.3 0.131 3.1         Model 1b 1d Decision Maker Students Parents All Independent Dependent Estimate Est./S.E. Estimate Est./S.E. Estimate Est./S.E. SCHWB GENATT 0.046 0.7 0.099 1.4 0.081 1.8 SCHTB GENATT 0.058 0.9 0.187 2.7 0.118 2.6 SCHWB TRANATT 0.087 1.3 0.178 2.6 0.151 3.3 SCHTB TRANATT 0.039 0.6 0.219 3.1 0.131 2.8 Correlation between GENATT and TRANATT 0.910 21.2 0.843 15.0 0.880 26.8  Note: Bold numbers are significant at 10% level.  Table 6.7 reports the direct, indirect and total effects of the demographic and socioeconomic variables on ENVATT via SCHWB and SCHTB that are significant at the 10% level.  For the STUDENT-DECISION-MAKER group, male, household car ownership, ownership personal vehicle, living in duplex and renting the home have negative and direct impacts on ENVATT.  No variables have an indirect effect on attitudes via travel mode to school.   For the PARENT- DECISION-MAKER group, there is a negative and direct impact by males on ENVATT.  On the other hand, living in high rises has positive and indirect impact on attitudes.   94  Table 6.7 Direct and Indirect Effects on Environmental Attitudes via Travel Mode to School Explanatory Variables Direct Effect Total Indirect Effect Total Effect Estimate Est./S.E. Estimate Est./S.E. Estimate Est./S.E. ENVATT Students Male -0.092 -1.9 -0.001 -0.17 -0.093 -1.87 HouseCar -0.191 -2.9 -0.005 -0.44 -0.197 -3.08 OwnCar -0.184 -3.3 -0.014 -1.08 -0.198 -3.65 Driver License -0.117 -1.6 -0.002 -0.25 -0.119 -1.66 Duplex -0.116 -2.3 -0.005 -0.83 -0.122 -2.38 Rent -0.109 -1.9 -0.001 -0.29 -0.11 -1.90   Parents Male -0.106 -1.76 0.011 0.72 -0.095 -1.54 Highrise -0.107 -1.61 0.031 1.73 -0.075 -1.11 Note: Bold numbers are significant at 10% level.  Table 6.8 reports the estimated relationships between environmental attitudes and travel mode from school.  The models’ results are similar to the impact of travel mode to school on environmental attitudes; with the exception that AFTWB and AFTTB have significant and positive impacts on TRANATT compared to AFTCAR for those students that made their own commuting decisions.  This suggests the mode choice for after school trips for those students have a more significant impact on transportation-specific environmental attitudes than morning trips.  Table 6.10 reports the estimated relationships between environmental attitudes and travel mode frequency.  In model 3a, CARFREQ has a significant impact on ENVATT for both PARENT- DECISION-MAKER and STUDENT-DECISION-MAKER.  This suggests that students who travel by car more frequently, their attitudes are likely to become less pro-environmental. In model 3c, WBFREQ and CARFREQ both have significant impacts on ENVATT.  Students who walk more have higher levels of pro-environmental attitudes, and students who travel by car more have lower levels of pro-environmental attitudes.     95  Table 6.9 reports direct, indirect and total effects of the demographic and socioeconomic variables on ENVATT via AFTWB and AFTTB that are significant at the 10% level.  The results are similar with the impact on ENVATT via SCHWB and SCHTB.        Table 6.8 Structural Equations between Environmental Attitudes and Travel Mode from School  Model 2a 2c Decision Maker Students Parents All Independent Dependent Estimate Est./S.E. Estimate Est./S.E. Estimate Est./S.E. AFTWB ENVATT 0.104 1.5 0.208 3.2 0.167 3.7 AFTTB ENVATT 0.075 1.1 0.277 4.3 0.185 4.1         Model 2b 2d Decision Maker Students Parents All Independent Dependent Estimate Est./S.E. Estimate Est./S.E. Estimate Est./S.E. AFTWB GENATT 0.008 0.1 0.144 2.0 0.082 1.6 AFTTB GENATT 0.009 0.1 0.233 3.2 0.136 2.7 AFTWB TRANATT 0.214 2.9 0.255 3.7 0.243 5.2 AFTTB TRANATT 0.147 2.0 0.300 4.4 0.218 4.6 Correlation between GENATT and TRANATT 0.917 21.7 0.843 14.7 0.885 26.9  Note: Bold numbers are significant at 10% level.  Table 6.10 reports the estimated relationships between environmental attitudes and travel mode frequency.  In model 3a, CARFREQ has a significant impact on ENVATT for both PARENT- DECISION-MAKER and STUDENT-DECISION-MAKER.  This suggests that students who travel by car more frequently, their attitudes are likely to become less pro-environmental. In model 3c, WBFREQ and CARFREQ both have significant impacts on ENVATT.  Students who walk more 96  have higher levels of pro-environmental attitudes, and students who travel by car more have lower levels of pro-environmental attitudes.     97  Table 6.9 Direct and Indirect Effects on Environmental Attitudes via Travel Mode after School Explanatory Variables Direct Effect Total Indirect Effect Total Effect Estimate Est./S.E. Estimate Est./S.E. Estimate Est./S.E. ENVATT Students Male -0.090 -1.8 -0.002 -0.29 -0.092 -1.84 HouseCar -0.194 -3.0 -0.004 -0.37 -0.199 -3.10 OwnCar -0.176 -3.1 -0.023 -1.37 -0.199 -3.67 Duplex -0.116 2.3 -0.001 -0.26 -0.117 2.28 Rent -0.105 -1.8 -0.006 -0.80 -0.111 -1.92   Parents Male -0.109 -1.87 0.015 0.76 -0.094 -1.53 Grade -0.111 -1.59 0.05 2.00 -0.061 -0.84 HouseCar -0.026 -0.32 -0.053 -1.84 -0.079 -0.94 HouseLic 0.011 0.11 -0.059 -1.77 -0.048 -0.49 Note: Bold numbers are significant at 10% level.  In model 3b, for students whose parents are decision makers, DRIVEFREQ has a significant and negative impact on GENATT and TRANATT.  However, WALKFREQ has significant and positive impact on GENATT and not on TRANATT.  For those students who make their own decision, BIKEFREQ has significant and positive impact on GENATT and TRANATT, and BUSFREQ and DRIVEFEQ have significant and negative impacts on GENATT and TRANATT.  In model 3d, WALKFREQ, TRANFREQ have significant and positive impact on GENATT and TRANATT, and DRIVEFREQ has a significant and negative impact on GENATT and TRANATT.  BIKEFREQ has significant and positive impact on TRANATT and not on GENATT.  Similarly, PASSFREQ has significant and negative impact on TRANATT and not on GENATT.  For both model 3b and 3d, GENATT and TRANATT are highly correlated.     98  Table 6.10 Structural Equations between Environmental Attitudes and Travel Mode Frequency Model 3a 3c Decision Maker Students Parents All Independent Dependent Estimate Est./S.E. Estimate Est./S.E. Estimate Est./S.E. WBFREQ ENVATT 0.075 1.5 0.044 0.6 0.066 1.7 TBFREQ ENVATT 0.022 0.4 0.063 1.0 0.041 1.0 CARFREQ ENVATT -0.169 -3.2 -0.151 -2.4 -0.165 -4.1         Model 3b 3d Decision Maker Students Parents All Independent Dependent Estimate Est./S.E. Estimate Est./S.E. Estimate Est./S.E. WBFREQ GENATT 0.038 0.7 0.039 0.5 0.041 1.0 TBFREQ GENATT 0.012 0.2 0.082 1.2 0.051 1.2 CARFREQ GENATT -0.140 -2.5 -0.131 -2.0 -0.124 -2.9 WBFREQ TRANATT 0.121 2.2 0.04 0.5 0.091 2.1 TBFREQ TRANATT 0.030 0.5 0.029 0.4 0.02 0.5 CARFREQ TRANATT -0.201 -3.3 -0.16 -2.2 -0.206 -4.5 Correlation between GENATT and TRANATT 0.904 20.0 0.865 16.4 0.877 26.4 Note: Bold numbers are significant at 10% level.  Table 6.11 reports the direct, indirect and total effects of the demographic and socioeconomic variables on ENVATT that are significant at the 10% level.  For STUDENT-DECISION-MAKER group, male, household car ownership, ownership personal vehicle, and living in duplex have negative and direct impacts on ENVATT.  Household car ownership, ownership personal vehicle, living in duplex and renting the home have negative indirect impacts on ENVATT via travel mode frequency factors.  For PARENT-DECISION-MAKER group, male and high rises have negative and direct impacts on ENVATT; household car ownership and driving license status have negative and indirect impacts on ENVATT via travel mode frequency factors.          99  Table 6.11 Direct and Indirect Effects on ENVATT via Travel Mode Frequency  Explanatory Variables Direct Effect Total Indirect Effect Total Effect Estimate Est./S.E. Estimate Est./S.E. Estimate Est./S.E. ENVATT Students Male -0.095 -2.0 0.017 1.2 -0.078 -1.6 HouseCar -0.115 -1.8 -0.060 -2.5 -0.174 -2.9 OwnCar -0.123 -2.1 -0.077 -2.88 -0.200 -3.8 Duplex -0.121 -2.4 -0.024 -1.74 -0.145 -2.9 Rent -0.083 -1.5 -0.025 -1.69 -0.108 -1.9   Parents Male -0.102 -1.8 -0.010 0.6 -0.092 -1.6 HouseCar -0.063 -0.8 -0.068 -2.3 -0.131 -1.7 Drive License 0.014 0.2 -0.053 -2.1 -0.039 -0.5 Highrise -0.177 -2.7 0.013 0.5 -0.164 -2.7 Note: Bold numbers are significant at 10% level.  Table 6.12 reports the estimated relationships between environmental attitudes and mode share.  In model 4a, for both groups of students, WBSHARE and TBSHARE have significant and positive impacts on ENVATT compared to CARSHARE.  The same result is found in model 4c when the students are examined as a whole. In model 4b, for PARENT-DECISION-MAKER, all the relationships are significant except for WBSHARE on GENATT.  Students who use public transportation have more positive general environmental attitudes than students who travel to school by car, but there are no differences in general attitudes between the students who walk or bike to school and students who travel to school by car.   On the other hand, students who walk or bike to school, and students who take public transit or school bus have more positive transportation-specific attitudes than students who travel to school by car.  For STUDENT-DECISION-MAKER, WBSHARE and TBSHARE have significant impacts on both GENATT and TRANATT.  For both groups of students, GENATT and TRANATT are highly correlated.  The same result is found in model 4d when the students are examined as a whole.    100  Table 6.12 Structural Equations between Environmental Attitudes and Mode Share  Model 4a 4c Decision Maker Students Parents All Independent Dependent Estimate Est./S.E. Estimate Est./S.E. Estimate Est./S.E. WBSHARE ENVATT 0.223 3.9 0.119 1.8 0.186 4.5 TBSHARE ENVATT 0.176 3.0 0.165 2.5 0.162 3.8         Model 4b 4d Decision Maker Students Parents All Independent Dependent Estimate Est./S.E. Estimate Est./S.E. Estimate Est./S.E. WBSHARE GENATT 0.147 2.4 0.098 1.4 0.123 2.7 TBSHARE GENATT 0.126 2.0 0.162 2.3 0.131 2.8 WBSHARE TRANATT 0.320 5.3 0.125 1.7 0.249 5.5 TBSHARE TRANATT 0.234 3.7 0.154 2.1 0.184 4.0 Correlation between GENATT and TRANATT 0.912 19.8 0.86 16.2 0.879 26.3 Note: Bold numbers are significant at 10% level.  Table 6.13 reports the direct, indirect and total effects of the demographic and socioeconomic variables on ENVATT that are significant at the 10% level.  For the STUDENT-DECISION-MAKER group, male, household car ownership, ownership personal vehicle, years of living in current home, duplex and rent have negative and direct impacts on ENVATT. Household car ownership and ownership personal vehicle have negative indirect impacts on ENVATT via mode share factors.  For the PARENT-DECISION-MAKER group, only male and high rises have negative and direct impacts on ENVATT.         101  Table 6.13 Direct and Indirect Effects on Environmental Attitudes via Travel Mode Frequency  Explanatory Variables Direct Effect Total Indirect Effect Total Effect Estimate Est./S.E. Estimate Est./S.E. Estimate Est./S.E. ENVATT Students Male -0.089 -1.9 0.015 1.1 -0.074 -1.5 HouseCar -0.119 -2.0 -0.052 -2.5 -0.171 -2.8 OwnCar -0.142 -2.6 -0.053 -2.6 -0.195 -3.7 YearHome 0.084 1.6 -0.027 -1.7 0.057 1.1 Duplex -0.129 -2.6 -0.013 -0.9 -0.142 -2.8 Rent -0.093 -1.7 -0.011 -0.7 -0.104 -1.9   Parents Male -0.104 -1.8 0.015 1.1 -0.089 -1.5 Highrise -0.189 -3.1 0.029 1.8 -0.159 -2.6 Note: Bold numbers are significant at 10% level.  6.4 Summary In summary, travel behavior has a statistically significant impact on environmental attitudes for students whose parents made commuting decisions on their behalf.  Four main findings emerge.  First, students who walk or bike to school, and students who take public transit or school bus have higher levels of pro-environmental attitudes than students who travel to school by car.  Second, students who drive more frequently are more likely to have lower levels of pro- environmental attitudes.  Third, walking and biking mode share and public transit and school bus mode share have significant and positive impacts on environmental attitudes compared to car mode share.  Lastly, general environmental attitudes and transportation-specific environmental attitudes are highly correlated.  Therefore, there are similar impacts of travel behavior on both attitudes. Since there are mutual causalities between behavior and attitudes for those students who make their own commuting decisions instead of one-way impact from behavior to attitudes, the model results can only suggest the combined effect of the mutual causality between the two factors.   102  Despite the significant relationships between travel behaviors and environmental attitudes, the explanatory power for ENVATT are less than 21% for the three main models, suggesting that travel behavior, together with socioeconomic and demographic variables, can explain environmental attitudes only to a limited extent.     103  7 The Relationship between Environmental Knowledge, Attitudes and Travel Behavior 7.1 Introduction As mentioned in the literature review, few studies have explored the relationship between environmental knowledge, attitudes and travel behavior.  In particular, none of the previous studies have focused on youth.  In this chapter, the relationship between environmental knowledge, attitudes and travel behavior of youth is explored.  There are three hypotheses on the relationships among these three factors.   Hypothesis 1: Environmental attitude is a mediating variable or mediator of environmental knowledge’s effect on travel behavior (Figure 7.3).  Environmental knowledge significantly affects environmental attitudes; environmental attitudes have a significant effect on travel behavior; environmental knowledge affects travel behavior in the absence of environmental attitudes; and the effect of environmental knowledge on travel behavior is diminished to be non-significant with environmental attitudes added to the relationship.   Figure 7.1 The Mediating Effect of Environmental Attitudes on the Relationship between Environmental Knowledge and Travel Behavior  Hypothesis 2: Environmental knowledge is a confounding variable or confounder of environmental attitudes’ effect on travel behavior (Figure 7.1).  Environmental knowledge is associated with travel behavior, varies over the level of environmental attitudes (with relationship both ways), and is not a cause of travel behavior.   104    Figure 7.2 The Confounding Effect of Environmental Knowledge on the Relationship between Environmental Attitudes and Travel Behavior  Hypothesis 3: Environmental knowledge is a moderating variable or effect modifier. The magnitude of environmental knowledge affects the magnitude or direction of environmental attitudes’ effect on travel behavior (Figure 7.2).    Figure 7.3 The Moderating Effect of Environmental Knowledge on the Relationship between Environmental Attitudes and Travel Behavior  For each of the proposed relationship, environmental knowledge (ENVKNOW), environmental attitudes (ENVATT) and four aspects of travel behavior (travel mode to school, travel mode 105  from school, travel mode frequency and travel mode share) are included in the analysis.  The definition of travel behavior factors have been previously described in Chapter 6.  7.2 The Mediating Effect of Environmental Attitudes on the Relationship between Environmental Knowledge and Travel Behavior This section examines the mediating effect of environmental attitudes on the relationship between environmental knowledge and travel behavior.  Environmental knowledge is assumed to affect environmental attitudes, which in turn are assumed to affect travel behavior.   Table 7.1 summarizes the overall goodness-of-fit statistics for the models.  All measurement equations are significant at the 1% level. Table 7.1 Goodness of Fit of Statistics –Environmental Attitudes as Mediating Variable Goodness of Fit   Model 1 Model 2 Model 3 Model 4   Travel Mode to School Travel Mode from School Travel Mode Frequency Mode Share Observations 886 886 933 933 Chi-Square 692.84 698.911 749.74 724.946 Degree of Freedom 315 315 329 315 CFI 0.872 0.878 0.863 0.867 TLI 0.835 0.841 0.817 0.828 RMSEA 0.037 0.037 0.037 0.037 90% CI of RMSEA 0.033~0.041 0.033~0.041 0.033~0.041 0.033~0.041 SRMR 0.036 0.036 0.037 0.037  Table 7.2 reports the estimated relationships between environmental knowledge, environmental attitudes and travel behavior. In all the models, environmental knowledge has significant and positive impacts on environmental attitudes.  In model 1 and model 2, environmental attitudes have a more significant and positive impacts on travel to/from school by active transportation and transit in comparison to car.  In model 3, environmental attitudes have significant and a positive impact on travel mode frequency by active transportation, and a negative impact on travel mode frequency by car.  On the other hand, environmental attitudes 106  have an insignificant impact on travel mode frequency by transit.  In model 4, environmental attitudes have a significant and positive impact on active transportation and transit mode share.  This suggests that students who have better knowledge of environmental issues hold higher levels of pro-environmental attitudes.  Students with pro-environmental attitudes are more likely to use active transportation but not transit.     Four auxiliary SEMs are estimated to demonstrate the direct impact of environmental knowledge on travel behavior factors when environmental attitude is no longer a mediating variable between environmental knowledge and travel behavior factors.  ENVKNOW has a significant impact on AFTTB, but not on AFTCAR.  Overall, the model results suggest that environmental knowledge do not have a significant impact on travel behavior factors directly. Table 7.2 Structural Equations – Environmental Attitudes as Mediating Variable  Model Independent Dependent Estimate Est./S.E. 1 ENVKNOW ENVATT 0.388 11.7 ENVATT SCHTB 0.078 2.0 ENVATT SCHWB 0.069 1.7 2 ENVKNOW ENVATT 0.390 11.7 ENVATT AFTTB 0.090 2.2 ENVATT AFTWB 0.074 1.8 3 ENVKNOW ENVATT 0.396 12.2 ENVATT FREQWB 0.082 2.0 ENVATT FREQTB 0.044 1.1 ENVATT FREQCAR -0.16 -4.2 4 ENVKNOW ENVATT 0.396 12.2 ENVATT WBSHAR 0.125 3.0 ENVATT TBSHAR 0.090 2.2 107  7.3 The Confounding Effect of Environmental Knowledge on the Relationship between Environmental Attitudes and Travel Behavior This section examines the confounding effect of environmental knowledge on the relationship between environmental attitudes and travel behavior.  Environmental knowledge and attitudes are assumed to affect travel behavior; environmental knowledge and attitudes are assumed to affect each other. Table 7.3 summarizes the overall goodness-of-fit statistics for models.  All measurement equations are significant at the 1% level. Table 7.3 Goodness of Fit of statistics –Environmental Knowledge as Confounding Variable Goodness of Fit   Model 1 Model 2 Model 3 Model 4   Travel Mode to School Travel Mode from School Travel Mode Frequency Mode Share Observations 886 886 933 933 Chi-Square 692.201 695.924 741.186 721.299 Degree of Freedom 313 313 326 313 CFI 0.872 0.878 0.865 0.867 TLI 0.833 0.841 0.817 0.827 RMSEA 0.037 0.037 0.037 0.037 90% CI of RMSEA 0.033~0.041 0.033~0.041 0.033~0.040 0.034~0.041 SRMR 0.036 0.036 0.036 0.037  Table 7.4 reports the estimated relationships between environmental knowledge, environmental attitudes and travel behavior.  In all models, environmental knowledge and environmental attitudes are strongly related.  However, only two models have found a significant impact of environmental knowledge and attitudes on travel behavior. In model 1 and 2, environmental knowledge and attitudes do not have a significant impact on travel mode to/from school.  In model 3, environmental attitude has a significant impact on travel mode frequency and in the expected direction.  However, environmental knowledge has a negative 108  impact on FREQWB, and an insignificant impact on FREQTB and FREQCAR.  In model 4, environmental attitude has a significant impact on travel mode share.  Environmental knowledge also has a significant negative impact on active transportation mode share and an insignificant impact on transit mode share in comparison to car mode share.  Overall, the model results suggest that environmental knowledge does not have a significant impact on the relationship of environmental attitudes and travel behavior.  Table 7.4 Structural Equations – Environmental Knowledge as Confounding Variable  Model Independent Dependent Estimate Est./S.E. 1 ENVKNOW SCHTB 0.022 0.5 ENVKNOW SCHWB 0.010 0.3 ENVATT SCHTB 0.067 1.5 ENVATT SCHWB 0.064 1.4 Correlation between ENVKNOW and ENVATT 0.399 11.8 2 ENVKNOW AFTTB 0.035 0.9 ENVKNOW AFTWB 0.026 0.7 ENVATT AFTTB 0.071 1.6 ENVATT AFTWB 0.06 1.3 Correlation between ENVKNOW and ENVATT 0.399 11.8 3 ENVKNOW FREQWB -0.112 -2.7 ENVKNOW FREQTB -0.069 -1.6 ENVKNOW FREQCAR 0.017 0.4 ENVATT FREQWB 0.141 3.1 ENVATT FREQTB 0.081 1.7 ENVATT FREQCAR -0.168 -3.9 Correlation between ENVKNOW and ENVATT 0.408 12.5 4 ENVKNOW WBSHAR -0.076 -1.7 ENVKNOW TBSHAR -0.005 -0.1 ENVATT WBSHAR 0.165 3.5 ENVATT TBSHAR 0.093 2.0 Correlation between ENVKNOW and ENVATT 0.408 12.5  109  7.4 The Moderating Effect of Environmental Knowledge on the Relationship between Environmental Attitudes and Travel Behavior Multiple group analyses are performed to examine the relationship between attitudes and behavior for three different levels of knowledge.  As mentioned in chapter 4, the total score for the environmental knowledge questions is fourteen.  The definitions of the environmental knowledge levels are shown in Table 7.5. Table 7.5 Environmental Knowledge Levels Level of Knowledge Scores # of Samples Low  (KNOWLOW)  0-4 261 Medium (KNOWMED)  5-9 249 High (KNOWHIGH)  10-14 407  The moderating effect of environmental knowledge on the relationship between environmental attitudes and the four travel behavior variables (travel mode to school, travel mode from school, travel mode frequency and travel mode share) are tested using four models.  The results are summarized in Table 7.6, which shows the standardized coefficients (t-statistics in parenthesis).   Environmental attitudes have a significant impact on travel mode to school (active transportation in comparison to by car) for students that have low and medium levels of environmental knowledge.  In contrast, an insignificant relationship is found for students that have high level of environmental knowledge.  Similar results are found for travel mode from school, travel mode frequency and travel mode share. The results suggest environmental knowledge moderates the relationship between environmental attitudes and travel by active transportation in comparison to car.  However, environmental knowledge does not moderate the relationship between environmental attitudes and travel by transit in comparison to car.       110  Table 7.6 Structural Equations – Environmental Knowledge as Moderating Variable   Environmental Attitudes and Travel Mode to/from School   SCHWB SCHTB AFTWB AFTTB KNOWLOW 0.216 (2.5) 0.084 (1.0) 0.168 (1.7) 0.124 (1.3) KNOWMED 0.188 (2.2) 0.049 (0.6) 0.168 (1.9) 0.027 (0.3) KNOWHIGH -0.038 (-0.6) 0.079 (1.3) -0.011 (-0.2) 0.063 (1.0)   Environmental Attitudes and Travel Mode Frequency   FREQWB FREQTB FREQCAR   KNOWLOW 0.187 (2.1) 0.001 (0.0) -0.224 (-2.5)  KNOWMED 0.315 (3.7) 0.033 (0.4) -0.155 (-1.9)  KNOWHIGH -0.006 (-0.1) 0.091 (1.4) -0.071 (-1.2)    Environmental Attitudes and Travel Mode Share   WBSHAR TPSHAR     KNOWLOW 0.342 (3.6) 0.031 (0.3)   KNOWMED 0.344 (3.9) -0.052 (-0.6)   KNOWHIGH -0.058 (-0.9)         0.134 (2.1)     *The significant relationships are in bold  7.5 Summary This chapter has examined the three types of relationships among environmental knowledge, attitudes and travel behavior.  There are three major findings that emerge from the analysis.  First, while environmental knowledge does not have a direct impact on travel behavior, environmental attitude is found to be a mediating variable between environmental knowledge and travel behavior.  Second, environmental knowledge is not a confounder of the effect of environmental attitudes on travel behavior.  Lastly, environmental knowledge moderates the relationship between environmental attitudes and travel by active transportation.     111  8 Environmental Education Program and Policies This chapter explores the environmental education program and policies in British Columbia and seeks to answer three principal questions:  1) What is the current status of environmental education in British Columbia and how might it be improved?  2) What is the role of environmental education on transportation? 3) What are the sources of environmental information for students?   These questions are answered through review of the environmental education curriculum of British Columbia, interviews with teachers, representatives from the school board and transportation authority, as well as students’ survey results.      8.1 Environmental Education Curriculum in British Columbia In June of 1992, at the United Nations Conference on Environment and Development held in Rio de Janeiro, Canada signed the Agenda 21 agreement, which included obligations to promote education, public awareness and training on environment and sustainable development.  Several objectives were outlined including the achievement of environmental and development awareness in all sectors of society on a world-wide scale.  This objective is meant to be realized through the accessibility of environmental and development education, linked to social education, from primary school age through adulthood to all groups of people. Advancing environmental education within the mainstream education system and curriculum is seen as a key to long-term success. Under the Constitution, formal education falls within provincial and territorial jurisdiction, including the setting of the educational curriculum. Curriculum policy directs what is taught in schools and affects resource allocation, teacher training and the development of textbooks and other materials.  While direct federal involvement in formal environmental education is limited by its lack of constitutional authority for education, Environment Canada is collaborating actively with other federal departments, such as Natural Resources Canada, Parks Canada, and Fisheries and Oceans Canada, to steward Canada's response to the UN Decade of Education for 112  Sustainable Development (2005 - 2014) and with helping to move forward education for sustainable development in Canada. At the provincial/territorial level, approaches to curriculum policy and environmental education vary greatly. Most, if not all, provincial/territorial curricula include either goals or language relating to environmental education.  However, almost all provinces/territories lack a coordinated approach to the development and advancement of environmental education.  In British Columbia, the Ministry of Education developed the Integrated Resource Packages (IRPs) which are applicable to all public schools in the province. The IRPs are essentially the curriculum that is to be covered in the classroom and are split up into grade levels and subjects. There are four sections within each of the IRPs: Prescribed Learning Outcomes (PLOs), Suggested Instructional Strategies (SIS), Suggested Assessment Strategies, and Recommended Learning Resources.  The Prescribed Learning Outcomes outline the topics to be covered in the classroom.  It is the only portion of the IRPs that have to be strictly followed.  Suggested Instructional Strategies are suggested methods for the teachers to achieve the PLOs.  Assessment Strategies suggest methods to assess the achievement of the PLOs. The Recommended Learning Resources list all resources that are recommended by the Ministry of Education as being linked to the IRP for that grade level and subject.  The British Columbia Ministry of Education created a framework document in 1995, ‘Environmental Concepts in the Classroom: A Guide for Teachers’ as part of IRPs to serve as a guideline to environmental education.  The document introduces six guiding principles for integrating environmental education into the classroom setting; however, it does not discuss the benefits of integrating environment and sustainability themes, nor does it give examples of specific teaching methods or specific resources. Arai and Sprules (2001) conducted a study that assessed the status of environmental education in British Columbia public schools and found that many environmentally minded teachers feel that the framework document is inadequate.  In addition, this document was not commonly used and many were not aware of its existence; 80% of teachers in Vancouver surveyed were unaware of the Ministry's document.  In general, teachers rated the effectiveness of each subject’s IRP differently.  In addition, the number and 113  the depth of environmental topics covered in the curriculum varied between grade levels. The curriculum is set up in a way that splits all the grades and subjects into separate teaching compartments, making it difficult for the curriculum to be regarded as an integrated whole and interdisciplinary.  In the absence of specific curriculum guidelines, most teachers teach environmental education based on values and experiences rather than knowledge.  These values are based on childhood experiences, recreational pursuits, concern for health and lifestyles, and involvement in social and environmental issues.  88% percent of teachers indicate they incorporate environmental education in lessons.  However, a large proportion only incorporates environmental education on a monthly or weekly basis.  In summary, environmental education should be interdisciplinary, relevant and frequent, but evidence does not show this to be the case in British Columbia school.  At the municipal level, individual school boards are responsible for implementing the provincial curriculum, and each board makes budget allocation decisions and chooses and/or develops their own curriculum materials. As such, they are a key player in how and to what degree environmental education is implemented in the schools.  School principals and teachers also play a key role in implementing the curriculum, which implies that the environmental education reflected in the classroom varies by school and even class to class.  From the questionnaire results, it was apparent that some teachers do in fact incorporate more environmental education in their lessons than others. While some teachers reported that they never include environmental education in their lessons, others reported that they include it on a daily basis. This large discrepancy is expected to relate to differences in the grade level that teachers taught, the level of teaching experience that the teachers had, and/or the personal involvement and interest that teachers had in environmental issues and activities.  The decentralized responsibility for setting and implementing curriculum, spread across the provinces/territories, school boards and even individual schools, principals and teachers, makes it more difficult for groups to achieve policy changes.  114  In an interview with a representative of the Vancouver School Board, the representative stated that, in his opinion, environmental education is not being well taught in the classroom, though he thought it is fairly well woven into the IRPs.  Teachers identified a number of obstacles to environmental education in this province, including insufficient teacher training, scarce teaching resources, inadequate funding and lack of time.  First, the amount of training the teachers received varies between their teaching experiences.  From survey results, it was found that teachers with 21 to 30 years of teaching experience teach a considerable amount more environmental education than teachers with more or less experience.  This finding is consistent with historical events where environmental education became a major concern for politicians, parents and educators in the early 1970’s.  As a result, teachers with more or less than 21 to 30 years had less teaching training on environmental education.  The second obstacle is scarce teaching resources.  With a lack of good resources, teachers often fall back onto textbooks that may not cover environmental issues effectively and/or do not relate to the curriculum learning outcomes.   Lastly, many teachers stated that they found the amount of material and learning outcomes covered in the curriculum was too overwhelming, and they were not always able to cover all the topics. Some teachers apparently do not follow the IRP very closely at all, or feel as if they did not have the time to cover anything much in any depth. 8.2 Interviews There were a total of seven participants for the interview, which included one sustainable coordinator each from the Vancouver School Board and the Richmond School Board, two teachers each from the Vancouver School District and Richmond School District, and the representative for the Translink’s TravelSmart program.  Interviews were conducted after the survey implementation was completed, between March 2012 and May 2012.  Participants were asked to discuss their experience with the current environmental education program, and opinions on improving environmental education in British Columbia.  Obvious sample limitations can be noted.  The teachers who participated in this interview might be more concerned about environmental education.  Each interview consisted of an approximately 20 115  minute interview with an open-ended format.  The interview questions can be found in Appendix D. In general, the participants felt environmental education is not integrated well into the curriculum, and that there is lots of room for improvement.  The sustainability coordinator of the VSB stated that the environment is a board topic and that it is definitely challenging to include all the subjects of environment in the curriculum.  Although currently there is no mandatory environmental course, some classes such as the science and social studies do include topics related to sustainability and the environment.  The sustainability coordinator of the RSB believed the amount of environmental education incorporated in the courses certainly depends on the teachers.  As indicated by the teachers, most of the environmental education incorporated into their lessons based mostly on their own initiative rather than the curriculum.   For the two teachers that were aware of the 2007 “Environmental Learning and Experience Guide” distributed by the BC Ministry of Education, they found the document only somewhat useful.  They prefer to use the resources that they found from the internet and media.  One teacher said, “I use my own time searching through my extensive library of environmental education resources, searching relevant news stories and scientific studies, and using social media to find local news”.   According to the sustainability coordinators, they felt the teachers are overwhelmed with the mandatory curriculum and do not have time to cover the materials from the environmental curriculum.  On the other hand, one teacher suggests the Ministry should include more materials and improve the quality of the materials in the curriculum.   During the interview, the teachers were asked what techniques they use to help their students learn about the environment.  The teachers reported lecturing in the classroom, hands-on learning outdoors, having guest speakers, going on field trip and implementing practical conservation.  However, the teachers indicated there is lack of funding to support many of the activities.  One teacher stated “The students learn a lot from field trips, but there is always lack of funding for transportation for field trips”.  Conversely, the sustainability coordinator from RSB stated “Most activities do not require funding.  The teachers can utilize the resources available through the schools themselves.”  116  Furthermore, in discussing the obstacles of incorporating environmental education in the schools, the teachers and the sustainability coordinators also have different opinions.   The teachers said the major obstacles are the quality of curriculum materials, field trips and equipment, and training for teachers. On the other hand, the sustainability coordinators believed the main problem is that the teachers are overwhelmed with what they are teaching. The mandatory curriculum has too much material to be covered.    In relation to environmental education on transportation, the teachers are flexible in terms of what topics they want to cover.  The resources are mainly from the media on current events.  The representative from the Travel Smart program said they are developing curriculum and materials for teachers and schools interested in including TravelSmart in class discussion, projects, events and resources that will be available for download through their website.  They also have Travelsmart for secondary school programs introducing students to the concept of the benefits of choosing sustainable transportation option.  However, only schools and students that are interested are involved in the program.       8.3 Environmental Information Acquisition A total of 220 students completed the “Environmental Information Acquisition” section from the complete survey.  Six questions were asked concerning the source and potential impact of the environmental information the students’ received. The students were asked if they have ever participated in any environment-related programs in school or in their community.  43% reported they had never participated in any environment- related programs in school or in their community.  Of the students who participated in environment related programs, 75% participated in programs in school and 25% participated in programs in their community.    The students were asked to identify their information source on general and transportation related environmental issues.  As shown in Figure 8.1, the students indicated that television/radio is the primary source, followed by school courses and internet/mobile.  117  Surprisingly, while 57% of students reported they have participated in environment related programs, only 15% stated that they receive the information from extracurricular activities.   Figure 8.1 Source of Environmental Information  The students were further asked if they want to learn more about environmental issues, 70% stated that they want to learn more about general environmental issues, and 63% want to learn more about the relations between transportation and environment.  Lastly, the students were asked to what extent the environmental information they have learnt made them travel more sustainably.  They stated that environmental information had a limited effect on how they travel. 8.4 Summary The current status of environmental education in British Columbia remains in a developing phase.  While there are resources available through the IRPs, the underlying problem remains in the overwhelming amount of information and also the lack of knowledge by the teachers to access this information. This said, it is recommended that the school board should look into enhancing structure around the environmental program. At this point, only teachers that are aware of such programs offered by Translink are actively incorporating them into their study plans. This said, communication of these programs should be made mandatory through information sessions. While social media plays an influential role in the student’s development, 0% 20% 40% 60% 80% Never Extracurricular activities Personal experiences Friends and peers Parents and family Newspaper/Magazine Internet/Mobile School courses Television/Radio General Transportation118  the survey indicates that environmental education will rely on the school system.  With students showing a genuine interest in learning more about environmental issues, it is imperative that the school system better equip the teachers in acting as a conduit of this knowledge.    119  9 Discussion and Conclusions Young people are spending more time in cars than ever before.  Between 1986 and 2006, weekday travel by car for 11-15 years olds increased 83%, while the per capita increase for adults was only 11% (University of Toronto 2008).  Young people who do the majority of their traveling by car while growing up may continue that behavior into adulthood and are reluctant to travel by alternative transportation modes.  Reducing car use is a dominant goal of most climate change strategies and will contribute to healthier environments for young people today as well as future generations.   This research explored the relationship between environmental knowledge, attitudes and travel behavior of youth in Vancouver.  It is focused on the youth because 1) they are the generation which will bear the consequences and have to find solutions to the environment problems rising today; 2) the five years at high school are the most important period for students to gain environmental knowledge; and 3) there is a good opportunity to encourage positive attitudes through formal education.  The ultimate goal of this study is to influence youth to increase the use of sustainable transportation. This final chapter is organized as follows.  Section 9.1 describes the major findings of the study.  Section 9.2 discusses the overall policy and research implications that are supported by the findings.  Finally, section 9.3 describes the future research directions. 9.1 Major Findings Based on the 937 student responses collected from the City of Richmond and City of Vancouver, it is found that the students lack in knowledge about the environment.  On average, students answered half of the questions correctly, with only 1% of students answering more than 80% of the questions correctly.  A series of ANOVA found significant differences in general knowledge and transportation-specific knowledge for the whole sample at the 5% level. The students in general have attitudes that are protective of the environment; three-quarters of the students have positive environmental attitudes.  A series of ANOVA tests  found 120  significant differences between general environmental attitudes and transportation-specific environmental attitudes for the whole sample at the 5% level.  Overall, students have more positive general attitudes than transportation-specific environmental attitudes.  There is a positive correlation between general and transportation-specific attitudes, suggesting that students with positive attitudes toward general environmental issues are likely to have positive attitudes toward transportation-specific issues.   Students across five grades are compared in order to assess education’s impact on environmental knowledge and attitudes.  However, there were no observable increases in either general or transportation-specific knowledge or attitudes.  This finding suggests that the schools’ curricula do not contribute to students’ environmental knowledge and attitudes.   The survey results illustrate the diversity in transportation mode used for trips to school.  Students in this study made 36% of their trips to school by walking, 35% by car (12% as car driver and 23% as car passenger), 16% by public transit, 7% by school bus and 3% by bicycle.   In contrast, students’ ideal travel mode to school differs from their actual travel mode.  More students prefer to drive and bike to school; fewer students prefer to take public transit or walk to school.  42% of students travelled to school by themselves, and the other mostly travelled with their classmates and siblings.  Only 3% travelled to school with their parents.  61% of students were the sole decision makers in their commuting mode choice while the rest have their parents and guardians as decision maker.  Overall, female students are more likely to be driven and fewer bike to school in comparison to male students.  Male students and higher grade students have a stronger desire to drive and bike, have a higher tendency to make their own commuting decisions, and make more school trips alone. 9.1.1 Environmental Knowledge Affect Environmental Attitudes Environmental knowledge is positively associated with pro-environmental attitudes.  Students who have better knowledge of environmental issues have higher levels of pro-environmental attitudes.  Both general knowledge and transportation-specific knowledge contribute to positive environmental attitudes.  While general and transportation-specific attitudes are highly 121  correlated, the correlation between general knowledge and transportation knowledge is significant but much weaker in comparison. Even though the structural equation models showed that students’ environmental knowledge is associated positively with their environmental attitudes, only 15% of the variation in attitudes can be explained by knowledge.  Environmental knowledge is an important determining factor of attitudes, but not a dominant one; many other factors also influence environmental attitudes.   9.1.2 Travel Behavior Affect Environmental Attitudes Four structural equation models are used to assess the impact of travel behavior (Travel mode to school, travel mode from school, travel mode frequency and travel mode share) on environmental attitudes.    In summary, travel behavior shows a statistically significant impact on environmental attitudes for the students whose parents made commuting decisions on their behalf.  First, students who use active transportation, public transit and school buses to/from school have stronger pro-environmental attitudes than students who travel to /from school by car.  Second, students who use active transportation more frequent have higher level of pro- environmental attitudes.  In contrast, students who use car more frequent have lower levels of pro-environmental attitudes. Third, active transportation, public transit and school buses mode share have significant and positive impacts on environmental attitudes compared to car mode share.  Lastly, general environmental attitudes and transportation-specific environmental attitudes are highly correlated.  Therefore, there are similar impacts of travel behavior on both attitudes.   Despite the significant relationships between travel behaviors and environmental attitudes, the r-square values of ENVATT are less than 21%.  This suggests that travel behavior, together with socioeconomic and demographic variables, can explain environmental attitudes only to a limited extent.   122  9.1.3 The Relationship between Environmental Knowledge, Attitudes and Travel Behavior Three possible relationships between environmental knowledge, attitudes and travel behavior were explored.  There are three major findings that emerge from the analysis.  First, environmental attitude is found to be a mediating variable between environmental knowledge and travel behavior.  Second, environmental knowledge is not a confounder to the effect of environmental attitudes on travel behavior.  Lastly, environmental knowledge has a significant impact on the relationship between environmental attitudes and travel behavior.   9.2 Research and Policy Implications The survey results indicate that there is a discrepancy between students’ actual travel mode and ideal travel mode to school. Many students who currently walk and use public transit to school prefer to drive to school.  This is probably because the majority of students are not at the age to obtain a driver license.  Previous research has shown that the travel behavior intentions of young people between the ages of 11 and 18 are dominated by the desire to drive and/or own a car in the future (Turner and Pilling 1999; Storey and Brannen 2000; Derek Halden Consultancy 2003).  Travel behavior usually changes at the age of 17, after the age of license acquisition, when young people move from other modes to being car drivers.  This research is developed with the goal to influence young people to be less reliant on the car and more reliant on alternative modes before the age of license acquisition.  Having better understanding on the relationship between environmental knowledge, attitudes and travel behavior of youth can contribute to the development and improvement of the environmental education programs in their connection to travel behavior.  Based on the review of the environmental education curriculum of British Columbia, interviews with teachers and representatives from the School Board, and students’ survey results, demand for environmental education exceeds what the Ministry of Education and School Boards currently offer.  The students in general have poor knowledge about the environment, but they show a genuine interest in learning more about general and transportation related environment issues.  The teachers want to improve the environmental education but there is a 123  lack of resources available to them (ex. quality of curriculum materials, field trips and equipment, and training for teachers).  Sustainability coordinators also felt environmental education is not integrated well into curriculum and the teachers are overwhelmed with what they are teaching.   This suggests the current state of environmental education in British Columbia remains in a developing phase, and there is a lot of room for improvement. There are significant correlations between environmental knowledge and attitudes, and attitudes and travel behavior.  Even though environmental knowledge does not have a direct impact on travel behavior, their significant relationships suggest environmental attitude is a mediating variable between environmental knowledge and travel behavior.  Therefore, continual improvement of the environmental education curriculum is recommended, but the curriculum should not be limited to knowledge-based education on transportation.  Rather, it should include activities that will help induce positive environmental attitudes.   Social media and school courses are currently the major sources for environmental information for the students.  Hence, educational institutes, government, environmental NGOs should continue to leverage these channels to effectively disburse information. Meanwhile, there should be a focus to improve and develop the other conduits to which students receive environmental information. For instance, while limited amount of students stated that they have received environmental information through extracurricular activities, additional resources should be allocated to extracurricular programs so more students could gain environmental knowledge through hands-on experiences. Lastly, parents play an important role in students’ knowledge, attitudes and travel behavior.  Many students stated that they have received general and transportation related information through their parents.  For the students whose parents made commuting decisions for them, they are more likely to be driven to school and less likely to walk to school. Parents’ perceptions about the safety of a mode positively influences the choice of the corresponding mode for school trips (McMillan 2003; Rhoulac 2005).  Parents claimed that the streets closest to the school are some of the most dangerous locations for children who travel to school on foot due to the high period-specific traffic volumes and erratic driving behavior of car-bound parents 124  (McMillan 2005).  The fear of crime and threats to personal safety is yet another hindering factor; two different surveys of parents revealed that parents were more worried that children may be abducted or harmed by a stranger than that the children would sustain a physical injury (Gilbert and O’Brien 2005).  Hence, it is critical to implement safe routes to school programs that enable policymakers, community leaders and schools to improve safety and encourage more students to walk and bicycle to school. In a study of commuter behavior at the University of North Carolina, Chapel Hill, Rodriguez and Joonwon (2004) found that the likelihood to travel by foot was positively associated with an increase in the percentage of sidewalks available on the shortest route to a destination.  This study also found that the appeal of walking and bicycling was affected by sloping terrain.  Some communities are now being built or redeveloped to include extensive trails, sidewalks, narrow streets and traffic calming in order to facilitate walking and cycling.  These changes not only increase the attractiveness of active transportation, they also reduce the risk and severity of injuries by providing a safe, secure and pedestrians- cyclists- friendly environment.  Through enforcement of traffic laws around schools and engineering of the street environment along the routes to schools, parents could change their perception of safety and allow their children to travel to school by active transportation. 9.3 Research Limitations and Future Research  The limitations of this study that should be addressed in future studies.  First, the conclusions are drawn based on 1032 respondents from 11 schools in the City of Richmond and City of Vancouver.  It is recommended to extend this study to other school districts and engage more schools in research for a more representative sample.  Second, only 220 students completed the in-class survey.  The data used for analysis are mainly from the four common sections of the in-class and lunch time survey.  It is suggested to get more students to participate in the in-class surveys as it explore students’ environmental behavior, attitudes toward travel modes, environmental information acquisition and environmental belief and values that were not covered in the lunch time survey.   Third, the students’ home location cannot be obtained because of parents’ concern of confidentiality.  With the students’ home location, the impact of 125  road network and public transit accessibility on travel behavior can be examined.  It is essential to have full disclosure with the parents on the approval issued by the UBC Behavioral Research Ethics Board which address the parents’ concern.  Future research should also be expanded to investigate the influences of parents, schools, peers on youth’s attitudes and behavior.   Since the survey results indicated that parents are an important source of environmental information and play a significant role in students’ travel decisions, it would be valuable to conduct surveys with the parents to better understand the influence that they have on their children.  In addition, social networks affect the perceptions, beliefs and actions of individuals through structural relations such as direct contact, and indirect interaction through intermediaries (Knoke and Yang 2008).  An individual’s network of peers could be a significant motivation for his or her opinions and actions towards the environment.  Therefore, future research should examine the association between the environmental attitudes and travel behavior of an individual and his/her network.  In addition, students’ academic performance should also be included in the analysis.  More resources should be allocated to the subject/course that has significant impact on knowledge, attitudes and travel behavior.  Finally, replicating this study in other areas will hopefully validate this methodology and also bring to light any differences in environmental knowledge, attitudes and travel behavior among youth from different areas.    126  References Ajzen, I. (1991). "The theory of planned behavior." Organizational Behavior and Human Decision Processes 50(2): 179-211.   Anable, J., B. Lane, et al. (2006). An Evidence Base Review of Public Attitudes to Climate Change and Transport Behaviour, Department for Transport.   Arai, E. and S. Sprules (2001). The Status of Environmental Education in British Columbia Public Schools: Grade 6, 7 and 8 in the Vancouver School District as a Case Study. Vancouver, The University of British Columbia. Undergraduate.   Arbuthnot, J. (1977). "The roles of attitudinal and personality variables in the prediction of environmental behavior and knowledge." Environment and Behavior 9(2): 217-232.   Arcury, T. A. (1990). "Environmental Attitudes and Environmental Knowledge." Human Organization 49: 300-304.   Arcury, T. A., T. P. Johnson, et al. (1986). "Ecological Worldview and Environmental Knowledge: The “New Environmental Paradigm”." The Journal of Environmental Education 17(4): 35-40.   Assael, H. and G. S. Day (1968). "Attitudes and Awareness as Predictors of Market Share." Journal of Advertising Research 8(4): 3-10.   Axhausen, K. W., A. Simma, et al. (2001). "Pre-commitment and usage: seasontickets, cars and travel." European Research in Regional Science 11: 101-110.   Bamberg, S., I. Ajzen, et al. (2003). "Choice of Travel Mode in the Theory of Planned Behavior: The Roles of Past Behavior, Habit, and Reasoned Action." Basic and Applied Social Psychology 25(3): 175 - 187.   Bamberg, S. and P. Schmidt (1998). "CHANGING TRAVEL-MODE CHOICE AS RATIONAL CHOICE." Rationality and Society 10(2): 223-252.   Banister, D. and R. Hickman (2006). "How to design a more sustainable and fairer built environment: transport and communications." Intelligent Transport Systems, IEE Proceedings 153(4): 276-291.   Baron, R. M. and D. A. Kenny (1986). "The moderator–mediator variable distinction in social psychological research: Conceptual, strategic, and statistical considerations." Journal of Personality and Social Psychology 51(6): 1173-1182. 127    Ben-Akiva, M., D. McFadden, et al. (2002). "Hybrid Choice Models: Progress and Challenges." Marketing Letters 13(3): 163-175.   Bentler, P. M. (1990). "Comparative fit indexes in structural models." Psychological Bulletin 107(2): 238- 246.   Blake, D. E., N. Guppy, et al. (1997). "Canadian Public Opinion and Environmental Action: Evidence from British Columbia." Canadian Journal of Political Science/Revue canadienne de science politique 30(03): 451-472.   Blum, A. (1987). "Students' Knowledge and Beliefs concerning Environmental Issues in Four Countries." The Journal of Environmental Education 18(3): 7-13.   Bradley, Waliczek, et al. (1999). "Relationship Between Environmental Knowledge and Environmental Attitude of High School Students " Journal of Environmental Education 30(3).   Bristow, A., A. Pridmore, et al. (2004). How can we reduce carbon emissions from transport? Tyndall Centre Technical Report 15. Leeds, Tyndall Centre for Climate Change Research.   Browne, M. W. and R. Cudeck (1992). "Alternative Ways of Assessing Model Fit." Sociological Methods & Research 21(2): 230-258.   Browne, M. W. and M. G. (1990). RAMONA user’s guide, Ohio State University.   Carmines, E. G. and R. A. Zeller (1979). Reliability and Validity Assessment. Newbury Park, CA, Sage Publications Inc.   Chan, K. (1996). "Environmental attitudes and behaviour of secondary school students in Hong Kong." The Environmentalist 16(4): 297-306.   Choo, S. and P. Mokhtarian (2002). "The relationship of vehicle type choice to personality, lifestyle, attitudinal, and demographic variables." Davis, CA: Institute of Transportation Studies, University of California at Davis.   Choocharukul, K., H. T. Van, et al. (2008). "Psychological effects of travel behavior on preference and lifestyle in influencing vehicle type choice." Transportation Research Part A 42(1): 116-124.   City of Vancouver (2012). Greenest City 2020 Action Plan. Vancouver, BC. 128    Cohen, J., P. Cohen, et al. (2003). Applied Multiple Regression/Correlation Analysis for the Behavioral Sciences. Mahwah, NJ, USA, Lawrence Erlbaum Associates.   Cohen, M. J. (1984). Prejudice against nature. Freeport, ME, Cobble-smith.   Cronbach, L. J. and P. E. Meehl (1955). "Construct validity in psychological tests." Psychological Bulletin 52(4): 281-302.   Den Boon, A. K. (1980). Opvattingen over Autogrebruik en Milieuvervuiling, University of Amsterdam.  Department of the Environmental Transport and the Regions (1999).  Every Little Bit Helps: Are You Doing Your Bit? United Kingdom.   Derek Halden Consultancy (2003). Children’s Attitudes to Sustainable Transport. Edinburgh, Scottish Executive Social Research.   DeVellis, R. F. (2003). Scale development: Theory and application. Thousand Oaks , CA, Sage.   DeVon, H. A., M. E. Block, et al. (2007). "A Psychometric Toolbox for Testing Validity and Reliability." Journal of Nursing Scholarship 39(2): 155-164.   DiGuiseppi, C., I. Roberts, et al. (1998). "Determinants of car travel on daily journeys to school: cross sectional survey of primary school children." BMJ 316(7142): 1426-1428.   Dobson, R., F. Dunbar, et al. (1978). "Structural models for the analysis of traveler attitude-behavior relationships." Transportation 7(4): 351-363.   Dunlap, R. E. (1978). "The new environmental paradigm: a proposed instrument and preliminary results." The Journal of Environmental Education 9: 10-19.   Dunlap, R. E., K. D. Van Liere, et al. (2000). "New Trends in Measuring Environmental Attitudes: Measuring Endorsement of the New Ecological Paradigm: A Revised NEP Scale." Journal of Social Issues 56(3): 425-442.   Flamm, B. (2009). "The impacts of environmental knowledge and attitudes on vehicle ownership and use." Transportation Research Part D: Transport and Environment 14(4): 272-279.   Forward, S. E. (1998). Behavioural Factors Affecting Modal Choice: ADONIS. Sweden, Swedish National Road and Transport Research Institute. 129    Gagnon Thompson, S. C. and M. A. Barton (1994). "Ecocentric and anthropocentric attitudes toward the environment." Journal of Environmental Psychology 14(2): 149-157.   Gambro, J. S. and H. N. Switzky (1996). "A national survey of high school students' environmental knowledge." Journal of Environmental Education 27(3): 28.   Gärling, T. (1998). Behavioural Assumptions Overlooked in Travel-Choice Modeling. Oxford, UK, Elsevier.   Gilbert, R. and C. O’Brien (2005). Child- and youth-friendly land-use and transport planning guidelines. Toronto, Ontario, The Centre for Sustainable Transportation    Golob, T. F. (2001). "Joint models of attitudes and behavior in evaluation of the San Diego I-15 congestion pricing project." Transportation Research Part A: Policy and Practice 35(6): 495-514.   Golob, T. F. (2003). "Structural equation modeling for travel behavior research." Transportation Research Part B: Methodological 37(1): 1-25.   Golob, T. F. and D. A. Hensher (1998). "Greenhouse gas emissions and australian commuters' attitudes and behavior concerning abatement policies and personal involvement." Transportation Research Part D: Transport and Environment 3(1): 1-18.   Golob, T. F., A. D. Horowitz, et al. (1979). Attitude-Behavior Relationships in Travel Demand Modeling. London, Croom Helm.   Gopinath, A. D. (1995). Modeling Heterogeneity in Discrete Choice Processes: Application to Travel Demand. Department of Civil and Environmental Engineering, Massachusetts Institute of Technology. Ph.D.   Government of Canada (2010, June 2, 2010). "2008 Canadian Transportation Greenhouse Gas Emissions by Mode." Retrieved August 1, 2012, from http://www.climatechange.gc.ca/default.asp?lang=En&n=36BDD00A-1.   Greening, L. A., D. L. Greene, et al. (2000). "Energy efficiency and consumption — the rebound effect — a survey." Energy Policy 28(6–7): 389-401.   Hair, J. F., R. E. Anderson, et al. (1998). Multivariate Data Analysis with Readings. Englewood Cliffs, NJ, Prentice Hall.   130  Handy, S., X. Cao, et al. (2005). "Correlation or causality between the built environment and travel behavior? Evidence from Northern California." Transportation Research Part D: Transport and Environment 10(6): 427-444.   Hausbeck, K. W., L. W. Milbrath, et al. (1992). "Environmental Knowledge, Awareness and Concern Among 11th-Grade Students: New York State." The Journal of Environmental Education 24(1): 27-34.   He, K., H. Huo, et al. (2004). "Oil consumption and CO2 emissions in China's road transport: current status, future trends, and policy implications." Energy Policy 33(12): 1499-1507.   He, X., T. Hong, et al. (2011). "A comparative study of environmental knowledge, attitudes and behaviors among university students in China." International Research in Geographical and Environmental Education 20(2): 91-104.   Hounsham, S. (2006). Painting the Town Green: how to persuade people to be environmentally friendly – a report for everyone involved in promoting greener lifestyles to the public. London, Green-Engage.   Hu, L. t. and P. M. Bentler (1999). "Cutoff criteria for fit indexes in covariance structure analysis: Conventional criteria versus new alternatives." Structural Equation Modeling: A Multidisciplinary Journal 6(1): 1-55.   Isenhour, C. (2010). "On conflicted Swedish consumers, the effort to stop shopping and neoliberal environmental governance." Journal of Consumer Behaviour 9(6): 454-469.   Ivy, T. G.-C., K. S. Road, et al. (1998). "A Survey of Environmental Knowledge, Attitudes and Behaviour of Students in Singapore." International Research in Geographical and Environmental Education 7(3): 181- 202.   Jakobsson, C., S. Fujii, et al. (2000). "Determinants of private car users' acceptance of road pricing." Transport Policy 7(2): 153-158.   Jaus, H. H. (1982). "The effect of environmental education instruction on children's attitudes toward the environment." Science Education 66(5): 689-692.   Kaiser, F. G., S. Wölfing, et al. (1999). "ENVIRONMENTAL ATTITUDE AND ECOLOGICAL BEHAVIOUR." Journal of Environmental Psychology 19(1): 1-19.   Kallgren, C. A. and W. Wood (1986). "Access to attitude-relevant information in memory as a determinant of attitude-behavior consistency." Journal of Experimental Social Psychology 22(4): 328-338. 131    Kilbourne, W. E., S. C. Beckmann, et al. (2002). "The role of the dominant social paradigm in environmental attitudes: a multinational examination." Journal of Business Research 55(3): 193-204.   Kitamura, R., P. L. Mokhtarian, et al. (1997). "A micro-analysis of land use and travel in five neighborhoods in the San Francisco Bay Area." Transportation 24(2): 125-158.   Kitamura, R. and Y. O. Susilo (2005). "IS TRAVEL DEMAND INSATIABLE? A STUDY OF CHANGES IN STRUCTURAL RELATIONSHIPS UNDERLYING TRAVEL." Transportmetrica 1(1): 23-45.   Kline, R. B. (2010). Principles and Practice of Structural Equation Modeling. New York, Guildford.   Knoke, D. and S. Yang (2008). Social Network Analysis. Los Angeles, Sage.   Kollmuss, A. and J. Agyeman (2002). "Mind the Gap: Why do people act environmentally and what are the barriers to pro-environmental behavior?" Environmental Education Research 8(3): 239-260.   Kopnina, H. (2011). "Kids and cars: Environmental attitudes in children." Transport Policy 18(4): 573-578.   Kuhlemeier, H., H. V. D. Bergh, et al. (1999). "Environmental knowledge, attitudes, and behavior in Dutch secondary education." Journal of Environmental Education 30(2): 4.   Lanken, B., H. Aarts, et al. (1994). "Attitude Versus General Habit: Antecedents of Travel Mode Choice1." Journal of Applied Social Psychology 24(4): 285-300.   Liere, K. D. V. and R. E. Dunlap (1980). "The Social Bases of Environmental Concern: A Review of Hypotheses, Explanations and Empirical Evidence." Public Opinion Quarterly 44(2): 181-197.   Line, T., K. Chatterjee, et al. (2010). "The travel behaviour intentions of young people in the context of climate change." Journal of Transport Geography 18(2): 238-246.   Loewenstein, G. (2001). "The Creative Destruction of Decision Research." Journal of Consumer Research 28(3): 499-505.   Lyon, P. K. (1984). "Time-dependent structural equations modeling: A methodology for analyzing the dynamic attitude – behavior relationship." Transportation Science 18(4): 395-414.   132  MacCallum, R. C., M. W. Browne, et al. (1996). "Power analysis and determination of sample size for covariance structure modeling." Psychological Methods 1(2): 130-149.   Makki, M. H., F. Abd-El-Khalick, et al. (2003). "Lebanese Secondary School Students' Environmental Knowledge and Attitudes." Environmental Education Research 9(1): 21-33.   Maloney, M. P. and M. P. Ward (1973). "Ecology: Let's hear from the people: An objective scale for the measurement of ecological attitudes and knowledge." American Psychologist 28(7): 583-586.   Mangas, V. J., P. Martinez, et al. (1997). "Analysis of Environmental Concepts and Attitudes Among Biology Degree Students." The Journal of Environmental Education 29(1): 28-33.   McFadden, D. (2000). Disaggregate Behavioral Travel Demand's RUM Side: A 30-Year Retrospective. International Association of Travel Behavior Research (IATBR) Conference. Gold Coast, Queensland, Australia.   McMillan, T. E. (2003). Walking and urban form: modeling and testing parental decisions about children’s travel. Irvine, University of California. PhD.   McMillan, T. E. (2005). "Urban Form and a Child’s Trip to School: The Current Literature and a Framework for Future Research." Journal of Planning Literature 19(4): 440-456.   Meinhold, J. L. and A. J. Malkus (2005). "Adolescent Environmental Behaviors." Environment and Behavior 37(4): 511-532.   Morikawa, T. and K. Sasaki (1998). Discrete choice models with latent variables using subjective data. Travel Behaviour Research: Updating the State of Pla. Pergamon, Oxford: 435-455.   Muller, D., C. M. Judd, et al. (2005). "When moderation is mediated and mediation is moderated." Journal of Personality and Social Psychology 89(6): 852-863.   Munro, B. H. (2005). Statistical Methods for Health Care Research. Philadelphia, PA., Lippincott Williams and Wilkins.   Munson, B. H. (1994). "Ecological Misconceptions." The Journal of Environmental Education 25(4): 30-34.   Muthen, L. and B. Muthen (2007). Mplus Statistical Analysis with Latent Variables User's Guide.   133  Negev, M., G. Sagy, et al. (2008). "Evaluating the Environmental Literacy of Israeli Elementary and High School Students." Journal of Environmental Education 39(2): 3-20.   Nilsson, M. and R. Küller (2000). "Travel behaviour and environmental concern." Transportation Research Part D: Transport and Environment 5(3): 211-234.   Nunnally, J. C. and I. H. Bernstein (1994). Psychometric theory. New York, McGraw-Hill.   Ory, D. T. and P. L. Mokhtarian (2009). "Modeling the structural relationships among short-distance travel amounts, perceptions, affections, and desires." Transportation Research Part A: Policy and Practice 43(1): 26-43.   Outwater, M., S. Castleberry, et al. (2003). "Attitudinal Market Segmentation Approach to Mode Choice and Ridership Forecasting: Structural Equation Modeling." Transportation Research Record: Journal of the Transportation Research Board 1854(-1): 32-42.   Pearl, J. (1998). Why there is no statistical test for confounding, why many think there is, and why they are almost right. Los Angeles, Department of Computer Science, University of California.   Pendyala, R. M. (1998). Causal Analysis in Travel Behaviour Research: A Cautionary Note. Pergamon, Oxford, Elsevier.   Ramsey, C. E. and R. E. Rickson (1976). "Environmental Knowledge and Attitudes." The Journal of Environmental Education 8(1): 10-18.   Reibstein, D. J., C. H. Lovelock, et al. (1980). "The Direction of Causality between Perceptions, Affect, and Behavior: An Application to Travel Behavior." Journal of Consumer Research 6(4): 370-376.   Rhoulac, T. (2005). "Bus or Car?: The Classic Choice in School Transportation." Transportation Research Record: Journal of the Transportation Research Board 1922(-1): 98-104.   Rodr  guez, D. A. and  .  oo (2004). "The relationship between non-motorized mode choice and the local physical environment." Transportation Research Part D: Transport and Environment 9(2): 151-173.   Shanahan, J. (1993). Television and the Cultivation of Environmental Concern. Leicester, Leicester University Press.   Singleton, R. A. and B. C. Straits (1999). Approaches to Social Research. New York and Oxford, Oxford University Press. 134    Stapp, W. (1978). "An instructional model for environmental education." Prospects: Quarterly Review of Education 8(4): 495-507.   Statistics Canada (2010, May 13, 2010). "Greenhouse Gas Emissions from Private Vehicles in Canada, 1990 to 2007." Retrieved August 1, 2012, from http://www.statcan.gc.ca/pub/16-001-m/2010012/part- partie1-eng.htm.   Steiger, J. H. (1989). A supplementary module for SYSTAT and SYGRAPH, Evanston IL: SYSTAT.   Steiger, J. H. and J. C. Lind (1980). Statistically based tests for the number of common factors. Annual Meeting of the Psychometric Society. Iowa City, IA.   Stern, P. C. and S. Oskamp (1987). Managing scarce environmental resources. New York, Wiley.   Storey, P. and J. Brannen (2000). Young People and Transport in Rural Areas. Leicester, National Youth Agency.   Tardiff, T. J. (1977). "Causal inferences involving transportation attitudes and behavior." Transportation Research 11(6): 397-404.   The United Nations Conference on Environment and Development (1993). The Earth Summit the United Nations Conference on Environment and Development 1992. Rio de Janeiro, Graham & Trotman, London.   Tikka, P. M., M. T. Kuitunen, et al. (2000). "Effects of Educational Background on Students' Attitudes, Activity Levels, and Knowledge Concerning the Environment." The Journal of Environmental Education 31(3): 12-19.   Tilbury, D. (1994). "The critical learning years for environmental education." Environmental Education at the Early Childhood Level. Washington, DC: North American Association for Environmental Education: 11-13.   Tischer, M. L. and R. V. Phillips (1979). "The relationship between transportation perceptions and behavior over time." Transportation 8(1): 21-36.   Trochim, W. M. K. (2001). The Research Methods Knowledge Base. Cincinnati , OH, Atomic Dog Publishing.   135  Trochim, W. M. K. (2005). Research Methods: The Concise Knowledge Base. Cincinati, Atomic Dog Publishing.   Turner, J. and A. Pilling (1999). Integrating young people into integrated transport: a community-based approach to increase travel awareness. Young People and Transport Conference.   University of Toronto (2008). 2006 Transportation Tomorrow Survey. Toronto, Data Management Group.   Walton, D., J. A. Thomas, et al. (2004). "Commuters' concern for the environment and knowledge of the effects of vehicle emissions." Transportation Research Part D: Transport and Environment 9(4): 335-340.   Waltz, C. F., O. L. Strickland, et al. (2005). Measurement in Nursing and Health Research. New York, Springer Publishing Co.   Weigel, R. H. and L. S. Newman (1976). "Increasing attitude-behavior correspondence by broadening the scope of the behavioral measure." Journal of Personality and Social Psychology 33(6): 793-802.   Wilson, R. (1994). Environmental Education at the Early Childhood Level. Washington, DC, North American Association for Environmental Eduaiton.   Worsley, A. and G. Skrzypiec (1998). "Environmental attitudes of senior secondary school students in South Australia." Global Environmental Change 8(3): 209-225.      Appendices   136Appendix A: Application Package to School Board  137   Assent Form  Environmental Attitude and Travel Behavior Study of Youth  Principal Investigator: Dr. Jinhua Zhao Assistant Professor Department of Civil Engineering, UBC  Co-Investigator: Cindy Tse Master of Applied Science Student Department of Civil Engineering, UBC  Kevin Millsip Coordinator for Sustainability Vancouver School Board  Dear Student, We are conducting a research study entitled “Environmental Attitude and Travel Behavior Study of Youth”. This research aims to examine the students’ environmental attitude and its correlation with travel behavior. This survey is carried out by University of British Columbia (UBC) in collaboration with Vancouver School Board.   I am inviting you to complete a questionnaire that will take about 20 to 30 minutes. Your parents or legal guardians have already given permission for you to participate in this study, but you do not have to participate if you choose. You may quit this study at any time. Your participation in this study will not affect your grades in any way. There are no known risks involved in this study. All information collected in the survey will be strictly confidential. No one at the high school will have access to any of the data collected. The data records will be stored in the principal investigator and co- investigator’s computers with password protection. Hard copies will be locked in the office of the Principal Investigator at the Department of Civil Engineering at UBC.  The data will be used for academic research only. Should you have any questions about the study, please contact the research team or your homeroom teacher. If you like to learn more about your rights as a research participant, please contact UBC Behavioral Research Ethics Board at 604.822-8598, toll free at 1-877-822-8598, or e-mail to RSIL@ors.ubc.ca. 138  Please give your permission by signing the enclosed assent form and returning it to your homeroom teacher.  Please keep this letter for your own records.    Sincerely,    Jinhua Zhao      139    Agree to Participate   I agree to participate in the study entitled “Environmental Attitude and Travel Behavior Study of Youth” and I have received a copy of this form.    Student’s Name (please print)   Student’s Signature      Date      140 1  8/18/ 20 11  Consent Form Environmental Attitude and Travel Behavior Study of Youth Principal Investigator: Dr. Jinhua Zhao Assistant Professor Department of Civil Engineering, UBC  Co-Investigator: Cindy Tse Master of Applied Science Student Department of Civil Engineering, UBC  Kevin Millsip Coordinator for Sustainability Vancouver School Board  Dear Parents or Guardians, We are conducting a research study entitled “Environmental Attitude and Travel ?ehavior Study of Youth”?  This research aims to examine the environmental knowledge, attitudes and travel behavior of youth in Vancouver public schools. The findings should be of particular interest to educational institutes, government, environmental groups and social organizations to improve the environmental education, with the goal to encourage students’ to?ard more sustaina?le travel? This survey is carried out by the University of British Columbia (UBC) in collaboration with Vancouver School Board.   Participants in the study will be asked to complete a survey. The total time to participate in the study will be approximately 20 to 30 minutes. Participants will be asked to provide their household information (such as income, car vehicle ownership, dwelling type, etc.) in the last section of the survey. Participants can skip any questions that they do not want to answer.       All information collected in the survey will be strictly confidential. No one at the high school will have access to any of the data collected. The data records will be stored in the principal investigator and co-investigator’s computers with password protection.  Hard copies will be locked in the office of the Principal Investigator at the Department of Civil Engineering at UBC.  The data will be used for academic research only. 141 2  8/18/ 20 11  Participation in the study is entirely voluntary and there will be no penalty for not participating. All students for whom we have parent consent will be asked if they wish to participate and only those who agree will complete the forms. Moreover, participants will be free to stop their participation in the study at any time.   Should you have any questions about the study, please contact the research team.  If you like to learn more a?out your youngster’s rights as a research participant, please contact UBC Behavioral Research Ethics Board at 604.822-8598, toll free at 1-877-822-8598, or e-mail to RSIL@ors.ubc.ca. Please give your permission by signing the enclosed consent form and having your youngster return it to his or her homeroom teacher.  Please keep this letter for your own records.    Sincerely,  Jinhua Zhao PhD. Assistant Professor Department of Civil Engineering / School of Community and Regional Planning University of British Columbia    142 3  8/18/ 20 11   Consent to Participate   I have read the attached informed consent letter and agree to have my youngster participate in the study entitled “Environmental Attitude and Travel Behavior Study of Youth?”    I consent / I do not consent (please circle one) to my youngster’s participation in this study.    Student’s ?ame   ?arent’s or ?uardian’s ?ame ?please print?   ?arent’s or ?uardian’s Signature      Date   143Page 1 Environmental Attitude and Travel Behavior of Youth    Thank you very much for participating in this survey.   This research aims to examine students’ environmental attitude and its correlation with travel behavior. This survey is  carried out by the University of British Columbia.    This survey includes 8 sections:     1. Environmental Attitudes and Behavior  2. Environmental Assessment and Knowledge  3. Travel to School  4. Attitudes toward Travel Modes  5. Perceptions of Cycling  6. Environmental Information Acquisition  7. Personal and Household Information   8. Environmental Belief and Values    It takes about 30 minutes to complete the questionnaire. All information provided will be kept strictly confidential. The  data will be used for academic research only. If you have any questions about the survey, please feel free to contact the  research team.    Principal Investigator:   Dr. Jinhua Zhao  Assistant Professor  Department of Civil Engineering, UBC    Co­Investigators:   Cindy Tse  Master of Applied Science Student  Department of Civil Engineering, UBC    Kevin Millsip  Coordinator for Sustainability  Vancouver School Board      1. Have you ever changed your travel behavior because of the concern of the environment? 2. Have you ever encouraged other people to change his/her travel behavior because of the concern of the environment?   1. Introduction   2. Environmental Attitudes and Behavior Yes   nmlkj No   nmlkj Yes   nmlkj No   nmlkj 144Page 2 3. Listed below are statements on environmental attitudes. For each one, please indicate whether you STRONGLY AGREE, MILDLY AGREE, NEUTRAL, MILDLY DISAGREE or STRONGLY DISAGREE. 4. Below is a list of daily life activities. Which of the following activities did you perform in the past year? Strongly  Agree Mildly Agree Neutral Mildly  Disagree Strongly  Disagree Being environmentally responsible is important to me nmlkj nmlkj nmlkj nmlkj nmlkj For the sake of the environment, car users should pay higher taxes nmlkj nmlkj nmlkj nmlkj nmlkj The government shouldn’t force people to change in order to protect  the environment nmlkj nmlkj nmlkj nmlkj nmlkj We should not have to bring inconvenience to our daily lives in order to  protect the environment nmlkj nmlkj nmlkj nmlkj nmlkj The government should raise the price of gasoline to reduce congestion  and air pollution nmlkj nmlkj nmlkj nmlkj nmlkj It is important to build more roads to reduce congestion even if it will  increase pollution nmlkj nmlkj nmlkj nmlkj nmlkj We should reduce car usage to reduce pollution nmlkj nmlkj nmlkj nmlkj nmlkj I do not need to protect the environment if people around me do not  pay attention to the environment nmlkj nmlkj nmlkj nmlkj nmlkj The government should encourage the use of cars in the city to  stimulate economic growth through increase in consumption nmlkj nmlkj nmlkj nmlkj nmlkj In order to protect the environment, I would cycle and walk more,  because these mode do not pollute nmlkj nmlkj nmlkj nmlkj nmlkj In the current state of Canada, economic development is more  important than the protection of the environment nmlkj nmlkj nmlkj nmlkj nmlkj The Government should provide incentives to people who travel by  electric vehicles, public transport, bicycle or on foot nmlkj nmlkj nmlkj nmlkj nmlkj Purchase and use of private cars is freedom of choice, I will buy and use  even if it will damage the environment nmlkj nmlkj nmlkj nmlkj nmlkj The government should take more lead in protecting the environment,  even if people don't like it nmlkj nmlkj nmlkj nmlkj nmlkj The government should legislate on categorizing wastes into recyclable  categories nmlkj nmlkj nmlkj nmlkj nmlkj Always Sometimes Rarely Never Sort batteries or electronic products into specified recycling bins nmlkj nmlkj nmlkj nmlkj Purchase and use of energy­saving light bulbs nmlkj nmlkj nmlkj nmlkj Shut down water taps after use nmlkj nmlkj nmlkj nmlkj To protect the environment, refuse to use disposable tableware when  eating out nmlkj nmlkj nmlkj nmlkj Sort household wastes into recyclable and non­recyclable categories nmlkj nmlkj nmlkj nmlkj Take initiative to pick up and deposit waste found in a public area into  recycling bins and garbage cans nmlkj nmlkj nmlkj nmlkj Bring reusable shopping bags when shopping at supermarket or mall nmlkj nmlkj nmlkj nmlkj 145Page 3 5. Did you perform the following in the past year? 1. How serious are the environmental problems in Vancouver? 2. How serious is the pollution caused by traffic in Vancouver? 3. Which of the following transportation method create the least pollution? 4. Clothes Washing powder causes water pollution mainly because it contains: 5. The carbon of "carbon emissions" refers to: 6. What percentage of the world total CO2 emission does transport sector account for? 7. Which of the following statement about ecological footprint is not correct? Always Sometimes Rarely Never Discuss with friends about problems related to the environment nmlkj nmlkj nmlkj nmlkj Dissuade and stop other people from littering in the public nmlkj nmlkj nmlkj nmlkj Make recommendations to school or relevant government departments to  solve environmental problems nmlkj nmlkj nmlkj nmlkj Encourage others to participate in environmental protection activities nmlkj nmlkj nmlkj nmlkj Participate in activities of environmental organizations or give donations nmlkj nmlkj nmlkj nmlkj Participate in community service related to the environment nmlkj nmlkj nmlkj nmlkj   3. Environmental Assessment and Knowledge Very Serious   nmlkj Serious   nmlkj Not Serious   nmlkj No Problem   nmlkj Very Serious   nmlkj Serious   nmlkj Not Serious   nmlkj No Problem   nmlkj Bus   nmlkj Train   nmlkj Cycling   nmlkj Car   nmlkj Chlorine   nmlkj Nitrogen   nmlkj Phosphorus   nmlkj Potassium   nmlkj Carbon monoxide   nmlkj Carbon dioxide   nmlkj Coal   nmlkj Carbohydrate   nmlkj 2%   nmlkj 25%   nmlkj 50%   nmlkj 90%   nmlkj The ecological footprint is a measure of human demand on the Earth's ecosystems.   nmlkj It represents the amount of biologically productive land and sea area needed to regenerate the resources a human population consumes  and to absorb and render harmless the corresponding waste.  nmlkj Although present human life style continues to increase the ecology footprint, the total is still within Earth's bearing capacity   nmlkj Using this assessment, it is possible to estimate how much of the Earth (or how many planet Earths) it would take to support humanity if  everybody lived a given lifestyle.  nmlkj 146Page 4 8. "Biodiversity"means: 9. Which of the following statements about greenhouse effect is incorrect? 10. Which of the following type of fuel produce the most carbon dioxide emissions? 11. Which of the following waste materials can not be recycled? 12. Which of the following strategy does not reduce CO2 emission in urban transport? 13. "Organic food " means: 14. Which country is the world's largest oil consumer? Each living thing has many characteristics   nmlkj Each kind of living thing has many varieties   nmlkj Organic food has many varieties   nmlkj Diversity of Biological species, ecosystem and gene multiplicity   nmlkj The greenhouse effect is caused by greenhouse gases selective absorption of radiation which results in rise of surface temperature   nmlkj The main greenhouse gas is carbon dioxide   nmlkj Increased concentration of carbon dioxide in recent decades is a natural occurring process, not as a result of human activity.   nmlkj Current atmospheric CO2 concentration is increasing rapidly, while the increase of greenhouse gases cause global warming   nmlkj Compressed Natural Gas   nmlkj Petroleum­based fossil­fuels   nmlkj Ethanol   nmlkj Bio­diesel   nmlkj Waste paper, waste plastic   nmlkj Glass, metal   nmlkj Expired drugs   nmlkj Fabric   nmlkj Use alternative fuel vehicle   nmlkj Limit car ownership and usage   nmlkj Improve fuel economy   nmlkj Reduce public transportation services   nmlkj Food with no environmental impact   nmlkj Food produced without using synthetic pesticides, hormones, food additives and other production   nmlkj Uncontaminated food   nmlkj Food produced using genetic modification   nmlkj United States   nmlkj China   nmlkj Japan   nmlkj Canada   nmlkj 147Page 5 15. Which of the following transport sector has the highest energy consumption in Canada? 16. What percentage of British Columbia's greenhouse gas emissions come from Metro Vancouver cars? 1. What is the distance between your home and school? (kilometers) 2. What was your one­way commuting time to school yesterday (in minutes)?    3. What would be your ideal one­way commuting time to school? (minutes) 4. Do you travel to school with a companion? 5. What was the primary travel mode for your morning commute to school yesterday? 6. What did you do after you leave school yesterday? 7. Which was your main mode of transport after you leave school yesterday?   4. Travel to School Railways   nmlkj Highways   nmlkj Waterways   nmlkj Aviation   nmlkj 50%   nmlkj 35%   nmlkj 25%   nmlkj 10%   nmlkj less than  0.25km  nmlkj 0.25­0.5km   nmlkj 0.5­1km   nmlkj 1­2km   nmlkj 2­5km   nmlkj more than 5km   nmlkj 5   nmlkj 10   nmlkj 15   nmlkj 20   nmlkj 30   nmlkj 40   nmlkj 60   nmlkj By myself   nmlkj With classmates   nmlkj With my  parents/guardians  nmlkj With my siblings   nmlkj Other   nmlkj Car Driver   nmlkj Car Passenger   nmlkj Public Transit   nmlkj School Bus   nmlkj Taxi   nmlkj Walk   nmlkj Bike   nmlkj Other   nmlkj Go Home   nmlkj Personal Business (e.g., bank, dentist)   nmlkj Recreational/Social (e.g., movie, sport)   nmlkj Dining/Restaurant   nmlkj Shopping   nmlkj Work   nmlkj Other   nmlkj Car Driver   nmlkj Car Passenger   nmlkj Public Transit   nmlkj School Bus   nmlkj Taxi   nmlkj Walk   nmlkj Bike   nmlkj Other   nmlkj 148Page 6 8. What was the commuting time to your after school activity yesterday (in minutes)?    9. How often on average do you use each of the following travel modes? 10. Who decide how you travel to school (i.e. which travel mode to use)? 11. What are the factors that influence your current travel mode choice to school?  12. How would you rate the safety level of your daily commute to school? 13. If you were free to choose any mode of transport to and from school, what is the ideal mode you would choose?  6­7 days a  week 5 days a week 3­4 days a  week 1­2 days a  week once a  fortnight once a month less often never Walk nmlkj nmlkj nmlkj nmlkj nmlkj nmlkj nmlkj nmlkj Bike nmlkj nmlkj nmlkj nmlkj nmlkj nmlkj nmlkj nmlkj Public Transit Buses nmlkj nmlkj nmlkj nmlkj nmlkj nmlkj nmlkj nmlkj School Buses nmlkj nmlkj nmlkj nmlkj nmlkj nmlkj nmlkj nmlkj Sky Train / Commuter Rail nmlkj nmlkj nmlkj nmlkj nmlkj nmlkj nmlkj nmlkj Taxi nmlkj nmlkj nmlkj nmlkj nmlkj nmlkj nmlkj nmlkj Car Driver nmlkj nmlkj nmlkj nmlkj nmlkj nmlkj nmlkj nmlkj Car Passenger nmlkj nmlkj nmlkj nmlkj nmlkj nmlkj nmlkj nmlkj   5. Attitudes towards Transportation Myself   nmlkj Parents / Guardian   nmlkj Cost   gfedc Reliable   gfedc Comfortable   gfedc Convenient   gfedc Environmental   gfedc Health   gfedc Road Safety   gfedc Personal Safety   gfedc Social   gfedc Very Safe   nmlkj Safe   nmlkj Unsafe   nmlkj Very unsafe   nmlkj Walk   nmlkj Public Transit   nmlkj School Bus   nmlkj Car Driver   nmlkj Car Passenger   nmlkj Taxi   nmlkj Bike   nmlkj Other   nmlkj 149Page 7 1. Listed below are statements on different travel modes. For each one, please indicate whether you STRONGLY AGREE, MILDLY AGREE, NEUTRAL, MILDLY DISAGREE or STRONGLY DISAGREE. 2. When you grow up and can afford a car by yourself, will you buy it? 3. How would you like to travel to work when you grow up? 1. Do you own a bicycle? Strongly  Agree Mildly  Agree Neutral Mildly  Disagree Strongly  Disagree To own a car ownership gives me higher social status. nmlkj nmlkj nmlkj nmlkj nmlkj There is no relationship between owning a car and vanity. nmlkj nmlkj nmlkj nmlkj nmlkj Riding a private car satisfies my self­esteem to a certain extent. nmlkj nmlkj nmlkj nmlkj nmlkj Owning a car gives me a sense of achievement. nmlkj nmlkj nmlkj nmlkj nmlkj Car is simply a transportation tool; to own one or not has nothing to do with one's  pride. nmlkj nmlkj nmlkj nmlkj nmlkj I feel proud to own a car. nmlkj nmlkj nmlkj nmlkj nmlkj My lifestyle is dependent on car. nmlkj nmlkj nmlkj nmlkj nmlkj My life would be very difficult if I cannot travel by car. nmlkj nmlkj nmlkj nmlkj nmlkj Car is only one of the many transportation methods I use. I don’t feel my lifestyle is  dependent on car. nmlkj nmlkj nmlkj nmlkj nmlkj I don’t have time to think about how I travel, I just get in my car and go. nmlkj nmlkj nmlkj nmlkj nmlkj I decide whether to drive, use transit or walking and cycling according to the specific  circumstances when I travel. nmlkj nmlkj nmlkj nmlkj nmlkj Because of my lifestyle and geographic location, I have no choice but to drive. nmlkj nmlkj nmlkj nmlkj nmlkj Learning to drive and having your own car is an identity of an adult status. nmlkj nmlkj nmlkj nmlkj nmlkj To be able to drive says nothing about becoming grown­up. nmlkj nmlkj nmlkj nmlkj nmlkj Driving removes the dependency on others to get around. It indicates my  independence and maturity. nmlkj nmlkj nmlkj nmlkj nmlkj Using public transit is a decent way of traveling. nmlkj nmlkj nmlkj nmlkj nmlkj People use public transit because they cannot afford a car. nmlkj nmlkj nmlkj nmlkj nmlkj Using public transit will negatively affect my personal image. nmlkj nmlkj nmlkj nmlkj nmlkj   6. Perceptions of Cycling For Sure   nmlkj Maybe   nmlkj No   nmlkj Walk   nmlkj Bike   nmlkj Bus   nmlkj SkyTrain   nmlkj Car   nmlkj Other   nmlkj Yes   nmlkj No   nmlkj 150Page 8 2. How many bicycles does your household own? 3. For what purpose(s) do you ride your bike? 4. Please indicate the term(s) associated with cycling. (You can pick multiple answers) 5. Listed below are statements on cycling. For each one, please indicate whether you STRONGLY AGREE, MILDLY AGREE, NEUTRAL, MILDLY DISAGREE or STRONGLY DISAGREE. Strongly  Agree Mildly  Agree Neutral Mildly  Disagree Strongly  Disagree Cycling is a sustainable way of travel. nmlkj nmlkj nmlkj nmlkj nmlkj We should cycle more because it is environment­friendly. nmlkj nmlkj nmlkj nmlkj nmlkj Cycling or not cannot reflect a person’s environmental attitude. nmlkj nmlkj nmlkj nmlkj nmlkj Teenagers should form the habit of cycling to school because it is good for the  environment. nmlkj nmlkj nmlkj nmlkj nmlkj I enjoy cycling because it makes me fitter. nmlkj nmlkj nmlkj nmlkj nmlkj Teenagers who cycle to school will be fitter and healthier than those who travel by car. nmlkj nmlkj nmlkj nmlkj nmlkj It is not safe for teenagers to cycle because of the risk of being involved in a road  accident. nmlkj nmlkj nmlkj nmlkj nmlkj It is not safe for teenagers to cycle because of the risk of "Stranger Danger". nmlkj nmlkj nmlkj nmlkj nmlkj Cycling is an enjoyable family activity. nmlkj nmlkj nmlkj nmlkj nmlkj I love to cycle because I can get to places on my own. nmlkj nmlkj nmlkj nmlkj nmlkj Cycling is an efficient way of getting around. nmlkj nmlkj nmlkj nmlkj nmlkj People should cycle instead of travel by car to help reduce congestion on the roads. nmlkj nmlkj nmlkj nmlkj nmlkj Cycling is a fun activity for kids but not for grown­ups. nmlkj nmlkj nmlkj nmlkj nmlkj   7. Environmental Information 0   nmlkj 1   nmlkj 2   nmlkj 3 or more   nmlkj Utility (i.e. using bike as a  travel means to get to your  destination)  nmlkj Recreational (i.e. cycling  itself as fun or exercise)  nmlkj Both   nmlkj Neither   nmlkj Fun   gfedc Boring   gfedc Clean and Green   gfedc Polluting and Dirty   gfedc Fast   gfedc Slow   gfedc Cool   gfedc Trendy   gfedc Safe   gfedc Dangerous   gfedc Reliable   gfedc Tiring   gfedc Anything else? (please specify)  151Page 9 1. Have you ever participated in any environment­related programs in school or your community? 2. Where do you get information on environmental issues? (You can pick multiple answers) 3. Where do you learn the relations between transportation and environment? (You can pick multiple answers) 4. Do you want to learn more about environmental issues? 5. Do you want to learn more about the relations between transportation and environment? 6. To what extent will the environmental information you've learned make you travel more sustainably? 2. Your grade   8. Personal and Household Information 1. Your gender Yes, in school   nmlkj Yes, in my community   nmlkj No   nmlkj Television/Radio   gfedc Newspaper/Magazine   gfedc Internet/Mobile   gfedc School courses   gfedc Extracurricular activities   gfedc Parents and family   gfedc Friends and peers   gfedc Personal experiences   gfedc Never   gfedc Other (please specify)  Television/Radio   gfedc Newspaper/Magazine   gfedc Internet/Mobile   gfedc School courses   gfedc Extracurricular activities   gfedc Parents and family   gfedc Friends and peers   gfedc Personal experiences   gfedc Never   gfedc Other (please specify)  Yes   nmlkj No   nmlkj Yes   nmlkj No   nmlkj Greatly   nmlkj A little bit   nmlkj Not at all   nmlkj Male   nmlkj Female   nmlkj 8   nmlkj 9   nmlkj 10   nmlkj 11   nmlkj 12   nmlkj 152Page 10 3. What is your name?   4. What is your home postal code?   5. How many cars does your household own? 6. Do you have your own car? 7. What is your household annual income before taxes? 8. Number of household members   9. Number of children in your household (18 years or younger)   10. Number of household members with driver licenses   11. Do you have a driver license? 12. What is your home dwelling type? 13. Does your family own or rent your home? 14. How long have you been living in your current home? 0   nmlkj 1   nmlkj 2   nmlkj 3 or more   nmlkj Yes   nmlkj No   nmlkj Less than $25,000   nmlkj $25,000 to $49,999   nmlkj $50,000 to $74,999   nmlkj $75,000 to $99,999   nmlkj $100,000 to $149,999   nmlkj $150,000 and over   nmlkj Yes   nmlkj No   nmlkj House   nmlkj Duplex   nmlkj Townhouse   nmlkj Low Rises   nmlkj High Rises   nmlkj Other (please specify)  Own   nmlkj Rent   nmlkj Other   nmlkj less than 3 months   nmlkj 3 months to 1 year   nmlkj 1 year to 3 years   nmlkj 3 years to 5 years   nmlkj more than 5 years   nmlkj 153Page 11 15. What is your e­mail address?   1. Listed below are statements about the relationship between humans and the environment. For each one, please indicate whether you STRONGLY AGREE, MILDLY AGREE, UNSURE, MILDLY DISAGREE or STRONGLY DISAGREE.   9. Environmental Belief and Values Strongly  agree Mildly  agree Neutral Mildly  disagree Strongly  disagree Don't  understand  the  question We are approaching the population limit that earth can support nmlkj nmlkj nmlkj nmlkj nmlkj nmlkj Humans have the right to modify the natural environment to suit their  needs nmlkj nmlkj nmlkj nmlkj nmlkj nmlkj When human interfere with nature it often produces disastrous  consequences nmlkj nmlkj nmlkj nmlkj nmlkj nmlkj Human ingenuity will insure that we do NOT make the earth unlivable nmlkj nmlkj nmlkj nmlkj nmlkj nmlkj Humans are severely abusing the environment nmlkj nmlkj nmlkj nmlkj nmlkj nmlkj The earth has plenty of natural resources if we learn how to develop them nmlkj nmlkj nmlkj nmlkj nmlkj nmlkj Plants and animals have as much right as humans to exist nmlkj nmlkj nmlkj nmlkj nmlkj nmlkj The balance of nature is strong enough to cope with the impacts of  modern industrial nations nmlkj nmlkj nmlkj nmlkj nmlkj nmlkj Despite our special abilities, humans are still subject to the laws of nature nmlkj nmlkj nmlkj nmlkj nmlkj nmlkj The so­called "ecological crisis" facing mankind has been greatly  exaggerated nmlkj nmlkj nmlkj nmlkj nmlkj nmlkj The earth is like a spaceship with very limited room and resources nmlkj nmlkj nmlkj nmlkj nmlkj nmlkj Humans were meant to rule over the rest of nature nmlkj nmlkj nmlkj nmlkj nmlkj nmlkj The balance of nature is very delicate and easily upset nmlkj nmlkj nmlkj nmlkj nmlkj nmlkj Humans will eventually learn enough about how nature works to be able  to control it nmlkj nmlkj nmlkj nmlkj nmlkj nmlkj If things continue on their present course, we will soon experience a  major ecological catastrophe nmlkj nmlkj nmlkj nmlkj nmlkj nmlkj 154APPLICATION FOR PERMISSION TO CONDUCT RESEARCH STUDIES IN SCHOOL DISTRICT NO. 38 (RICHMOND) 7811 Granville Avenue, Richmond, B.C.    V6Y 3E3  Applicant: Cindy Tse Present Position: Master of Applied Science student at University of British Columbia Address: 2002-6250 Applied Science Lane, Vancouver, B.C. Canada V6T 1Z4     Title of Study: Environmental Attitude and Travel Behavior of High School Students  Statement of Problem:  The frequent use of private vehicle in urban areas has a significant impact on the environment as well as on human health.  Although technical improvements such as catalytic converters and fuel efficient engines have decrease the pollution per vehicle, environmental activists argue that this alone is not a sufficient solution to the problem of continuously increasing production and overconsumption of cars.  Since environmental attitudes have significant impacts on vehicle ownership and use, therefore understanding environmental attitudes and travel behavior is of increasing interest to transportation planners and policy makers.    This study will focus on the high school students between the ages of 12 to 15 as their travel behavior intentions are dominated by the desire to drive and/or own a car in the future but are not legally able to carry out this behavior yet.  This study explores the links between environmental value, environmental attitude, environmental behavior, environmental assessment, environmental knowledge, travel behavior, attitudes toward different transport modes and environmental information among youths.  The findings should be of particular interest to educational institutes, government, environmental groups and social organization to improve the environmental education, with the goal to influence the students’ toward more sustainable travel.    Significance of Study: (i.e. How might this study contribute to the improvement of education in the Richmond School District?)  Children are recognized as an important group for the development of a sustainable environment.  Studying young people’s attitudes to transport and the environment is particularly important as they are the ones who will be affected by and will have to provide solutions to the environment problems arising from our current actions.  Attitudes towards the environment begin to develop in childhood, it may be imperative to focus on children and young people in order to address transportation issues in the future.  With the increasing trend in transporting children to various activities by car, their attitude towards the environment and car usage may otherwise become detrimental to their future environment.    According to a Scottish research report, “An investigation into how children and young people view sustainability in relation to their personal transportation”, delivering an understanding of sustainable transport through education is one of the most important factors in achieving change in travel behavior.  Therefore, there is a need to understand the relationship between attitude, behavior and knowledge toward transportation and the environment for the development of environmental education curriculums.  In particular, an emphasis should be placed to educate the students before they develop a reliance on the car after the age of license acquisition.     Number of subjects desired who are: Students: entire grade 8 to gr 10     Students to be selected from Grade(s): 8 to 10 How much time is required from each subject?:  Students: 20- 30 mins    Will students be tested?: Will Applicant actually conduct study? Individually  Small Groups Entire Class Yes No  (If no, please give the name and qualifications of person(s) who will conduct the study):  Duration of Study: One day  Commencement Date:  April, 2011 Completion Date:  November, 2011 Date of submission of the Final Report and and one page Summary Report to the Richmond School Board: January 2012  155Checklist for Application Package: o Attach one page “Invitation to Participate in Research Study” (see template attached to this application). o Attach outline of research methodology (i.e. design, selection of subjects, experimental treatments, etc.). o Attach copies of all tests or questionnaires which will be given to the subjects. o Attach the letter or certificate of approval from appropriate committees at your university or institution, including Ethics Review approval.  FOR OFFICE USE ONLY: Application package received: Reviewed by The Research Study Review Committee:  o Granted o Denied Letter to Applicant sent: Invitation to Participate in Research Study sent to School Principal: Revised October 2010 156T H E  U N I V E R S I T Y  O F  B R I T I S H  C O L U M B I A    Dr. Jinhua Zhao   Assistant Professor    Department of Civil Engineering    #2007 - 6250 Applied Science Lane   Vancouver, B.C. Canada V6T 1Z4 1    Research Application to the Vancouver School Board  Environmental Attitudes and Travel Behavior of Youth   Aug 18, 2011  Researcher Names:  Dr. Jinhua Zhao Assistant Professor Department of Civil Engineering / School of Community and Regional Planning University of British Columbia  Cindy Tse Master of Applied Science Student Department of Civil Engineering University of British Columbia  Kevin Millsip Coordinator for Sustainability Vancouver School Board  Address:  #2007 - 6250 Applied Science Lane, Vancouver, B.C. Canada V6T 1Z4  Researcher-mailing address:   #2007 - 6250 Applied Science Lane, Vancouver, B.C. Canada V6T 1Z4  Study Completion Date:  December 31, 2011  Study Title:  Environmental Attitudes and Travel Behavior of Youth  Target Study Group:  High School Students in Vancouver   157T H E  U N I V E R S I T Y  O F  B R I T I S H  C O L U M B I A    Dr. Jinhua Zhao   Assistant Professor    Department of Civil Engineering    #2007 - 6250 Applied Science Lane   Vancouver, B.C. Canada V6T 1Z4 2   Abstract  Children and young people are recognized as an important group for the development o? a sustaina?le environment?  Studying young people’s knowledge and attitudes towards the environment is particularly important as they are the ones who will be affected by and will have to ultimately provide solutions to the environmental problems arising from our current actions.  Attitudes towards the environment begin to develop in childhood, it may be imperative to focus on children and young people in order to address transportation issues in the future.  This study explores the environmental knowledge, attitudes and travel behavior of youth.  The findings should be of particular interest to educational institutes, government, environmental groups and social organization to improve the environmental education, with the goal to in?luence the students’ to?ard more sustaina?le travel modes.    Research Objectives 1)  To benchmark the environmental knowledge and attitude among youth (high school students) 2)  To quantify to what extent these knowledge and attitudes are materialized into students' daily tra vel behavior 3) To identify contributing factors: formal education, public media, parents' influence, peers' interaction, etc 4) To compare youth across countries: Canada, US, China and UK. Ideal Sample Frame and Size 1) One randomly selected class for each of the five grades in each of the 18 Secondary schools in the VSB. Once a class is chosen, all students in the class will be invited to participate in the survey 2) Total sample size target: 25 students/class * 1 class/grade * 5 grades/school *18 schools = 2,700 students Implementation instruments  (all are available and can be chosen according to each school's need) 1) Paper based questionnaire  UBC can prepare the questionnaires and deliver it to the schools. We and/or the partner teachers can give a brief introduction of the project to the students, and students in the chosen class will have a dedicated 30 minutes to complete the questionnaire at school. 2) Online based questionnaire 158T H E  U N I V E R S I T Y  O F  B R I T I S H  C O L U M B I A    Dr. Jinhua Zhao   Assistant Professor    Department of Civil Engineering    #2007 - 6250 Applied Science Lane   Vancouver, B.C. Canada V6T 1Z4 3   UBC has prepared an online questionnaire and we can give the URL to all the schools and classes. Students can go online and complete the questionnaire at any convenient time.  A test link is here: https://www.surveymonkey.com/s/vsbtest  3) Online based with email solicitation If schools have the email list of students, we can send students a cover letter in the email, explaining the project and inviting them to participate in the online survey. Questionnaire design This is a comprehensive 14 pages questionnaire, covering environmental value, knowledge, attitude and behavior. It consists of 10 categories of information: environmental value, environmental attitudes, environmental behavior, environmental assessment, environmental knowledge, travel behavior, travel mode to school, travel mode evaluation, environmental information acquisition and respondent’s personal and household information.  It takes 20-30 minutes to complete.  Tentative Stages and Dates 1. Pilot in 1~2 classes: September, 2011 2. Full implementation in 90 classes: October 2011   UBC Behavioral Research  Ethical Board (BREB) Approval The UBC BREB has conditionally approved our research proposal (the certificate is attached with this application). The condition is that VSB approves the research as well.    159T H E  U N I V E R S I T Y  O F  B R I T I S H  C O L U M B I A    Dr. Jinhua Zhao   Assistant Professor    Department of Civil Engineering    #2007 - 6250 Applied Science Lane   Vancouver, B.C. Canada V6T 1Z4 4   Extensions 1.  Parents' version of the survey The current questionnaire is targeted at students. One natural extension is to test the environmental attitude and knowledge of parents of the youth who complete the questionnaire, who are one of most important source of information for students. The ideal date of implementation will be right after the student survey, say in November. Or we can implement the students and parents’ surveys together in ?cto?er? students complete the survey in school and then bring the adult version of the questionnaire home for their parents to fill in.  2.  Longitudinal survey across years Our long term plan is to monitor student’s environmental knowledge and attitude over time so that we can track the evolution or progress students make during their high school education. Ideally we would like to work with VSB to formalize this environmental program in VSB schools on an annual basis. 3.  Social network analysis of the peer effects The interactions ?ith peers are another important contri?utor to student’s environmental attitudes? This extension needs the in?ormation on students’ peers? We have developed questionnaires that can help students identify their peers or friends, with whom they interact the most.   Attachment 1. UBC BREB Conditional Approval Certificate 2. Full Questionnaire 3. Consent form for Parents 4. Assent form for Students     160Page 1 1. Have you ever changed your travel behavior because of the concern of the environment? 2. Have you ever encouraged other people to change his/her travel behavior because of the concern of the environment? 3. Listed below are statements on environmental attitudes. For each one, please indicate whether you STRONGLY AGREE, MILDLY AGREE, NEUTRAL, MILDLY DISAGREE or STRONGLY DISAGREE.   1. Environmental Attitudes and Behavior Strongly  Agree Mildly Agree Neutral Mildly  Disagree Strongly  Disagree Being environmentally responsible is important to me nmlkj nmlkj nmlkj nmlkj nmlkj For the sake of the environment, car users should pay higher taxes nmlkj nmlkj nmlkj nmlkj nmlkj The government shouldn’t force people to change in order to protect  the environment nmlkj nmlkj nmlkj nmlkj nmlkj We should not have to bring inconvenience to our daily lives in order to  protect the environment nmlkj nmlkj nmlkj nmlkj nmlkj The government should raise the price of gasoline to reduce congestion  and air pollution nmlkj nmlkj nmlkj nmlkj nmlkj It is important to build more roads to reduce congestion even if it will  increase pollution nmlkj nmlkj nmlkj nmlkj nmlkj We should reduce car usage to reduce pollution nmlkj nmlkj nmlkj nmlkj nmlkj I do not need to protect the environment if people around me do not  pay attention to the environment nmlkj nmlkj nmlkj nmlkj nmlkj The government should encourage the use of cars in the city to  stimulate economic growth through increase in consumption nmlkj nmlkj nmlkj nmlkj nmlkj In order to protect the environment, I would cycle and walk more,  because these mode do not pollute nmlkj nmlkj nmlkj nmlkj nmlkj In the current state of Canada, economic development is more  important than the protection of the environment nmlkj nmlkj nmlkj nmlkj nmlkj The Government should provide incentives to people who travel by  electric vehicles, public transport, bicycle or on foot nmlkj nmlkj nmlkj nmlkj nmlkj Purchase and use of private cars is freedom of choice, I will buy and use  even if it will damage the environment nmlkj nmlkj nmlkj nmlkj nmlkj The government should take more lead in protecting the environment,  even if people don't like it nmlkj nmlkj nmlkj nmlkj nmlkj The government should legislate on categorizing wastes into recyclable  categories nmlkj nmlkj nmlkj nmlkj nmlkj   2. Environmental Assessment and Knowledge Yes   nmlkj No   nmlkj Yes   nmlkj No   nmlkj 161Page 2 1. How serious are the environmental problems in Vancouver? 2. How serious is the pollution caused by traffic in Vancouver? 3. Which of the following transportation method create the least pollution? 4. Clothes Washing powder causes water pollution mainly because it contains: 5. The carbon of "carbon emissions" refers to: 6. What percentage of the world total CO2 emission does transport sector account for? 7. Which of the following statement about ecological footprint is not correct? 8. "Biodiversity"means: 9. Which of the following statements about greenhouse effect is incorrect? Very Serious   nmlkj Serious   nmlkj Not Serious   nmlkj No Problem   nmlkj Very Serious   nmlkj Serious   nmlkj Not Serious   nmlkj No Problem   nmlkj Bus   nmlkj Train   nmlkj Cycling   nmlkj Car   nmlkj Chlorine   nmlkj Nitrogen   nmlkj Phosphorus   nmlkj Potassium   nmlkj Carbon monoxide   nmlkj Carbon dioxide   nmlkj Coal   nmlkj Carbohydrate   nmlkj 2%   nmlkj 25%   nmlkj 50%   nmlkj 90%   nmlkj The ecological footprint is a measure of human demand on the Earth's ecosystems.   nmlkj It represents the amount of biologically productive land and sea area needed to regenerate the resources a human population consumes  and to absorb and render harmless the corresponding waste.  nmlkj Although present human life style continues to increase the ecology footprint, the total is still within Earth's bearing capacity   nmlkj Using this assessment, it is possible to estimate how much of the Earth (or how many planet Earths) it would take to support humanity if  everybody lived a given lifestyle.  nmlkj Each living thing has many characteristics   nmlkj Each kind of living thing has many varieties   nmlkj Organic food has many varieties   nmlkj Diversity of Biological species, ecosystem and gene multiplicity   nmlkj The greenhouse effect is caused by greenhouse gases selective absorption of radiation which results in rise of surface temperature   nmlkj The main greenhouse gas is carbon dioxide   nmlkj Increased concentration of carbon dioxide in recent decades is a natural occurring process, not as a result of human activity.   nmlkj Current atmospheric CO2 concentration is increasing rapidly, while the increase of greenhouse gases cause global warming   nmlkj 162Page 3 10. Which of the following type of fuel produce the most carbon dioxide emissions? 11. Which of the following waste materials can not be recycled? 12. Which of the following strategy does not reduce CO2 emission in urban transport? 13. "Organic food " means: 14. Which country is the world's largest oil consumer? 15. Which of the following transport sector has the highest energy consumption in Canada? 16. What percentage of British Columbia's greenhouse gas emissions come from Metro Vancouver cars? 1. What is the distance between your home and school? (kilometers)    3. Travel to School Compressed Natural Gas   nmlkj Petroleum­based fossil­fuels   nmlkj Ethanol   nmlkj Bio­diesel   nmlkj Waste paper, waste plastic   nmlkj Glass, metal   nmlkj Expired drugs   nmlkj Fabric   nmlkj Use alternative fuel vehicle   nmlkj Limit car ownership and usage   nmlkj Improve fuel economy   nmlkj Reduce public transportation services   nmlkj Food with no environmental impact   nmlkj Food produced without using synthetic pesticides, hormones, food additives and other production   nmlkj Uncontaminated food   nmlkj Food produced using genetic modification   nmlkj United States   nmlkj China   nmlkj Japan   nmlkj Canada   nmlkj Railways   nmlkj Highways   nmlkj Waterways   nmlkj Aviation   nmlkj 50%   nmlkj 35%   nmlkj 25%   nmlkj 10%   nmlkj less than  0.25km  nmlkj 0.25­0.5km   nmlkj 0.5­1km   nmlkj 1­2km   nmlkj 2­5km   nmlkj more than 5km   nmlkj 163Page 4 2. What was your one­way commuting time to school yesterday (in minutes)?    3. What would be your ideal one­way commuting time to school? (minutes) 4. Do you travel to school with a companion? 5. What was the primary travel mode for your morning commute to school yesterday? 6. What did you do after you leave school yesterday? 7. Which was your main mode of transport after you leave school yesterday?  8. What was the commuting time to your after school activity yesterday (in minutes)?    9. How often on average do you use each of the following travel modes? 6­7 days a  week 5 days a week 3­4 days a  week 1­2 days a  week once a  fortnight once a month less often never Walk nmlkj nmlkj nmlkj nmlkj nmlkj nmlkj nmlkj nmlkj Bike nmlkj nmlkj nmlkj nmlkj nmlkj nmlkj nmlkj nmlkj Public Transit Buses nmlkj nmlkj nmlkj nmlkj nmlkj nmlkj nmlkj nmlkj School Buses nmlkj nmlkj nmlkj nmlkj nmlkj nmlkj nmlkj nmlkj Sky Train / Commuter Rail nmlkj nmlkj nmlkj nmlkj nmlkj nmlkj nmlkj nmlkj Taxi nmlkj nmlkj nmlkj nmlkj nmlkj nmlkj nmlkj nmlkj Car Driver nmlkj nmlkj nmlkj nmlkj nmlkj nmlkj nmlkj nmlkj Car Passenger nmlkj nmlkj nmlkj nmlkj nmlkj nmlkj nmlkj nmlkj 5   nmlkj 10   nmlkj 15   nmlkj 20   nmlkj 30   nmlkj 40   nmlkj 60   nmlkj By myself   nmlkj With classmates   nmlkj With my  parents/guardians  nmlkj With my siblings   nmlkj Other   nmlkj Car Driver   nmlkj Car Passenger   nmlkj Public Transit   nmlkj School Bus   nmlkj Taxi   nmlkj Walk   nmlkj Bike   nmlkj Other   nmlkj Go Home   nmlkj Personal Business (e.g., bank, dentist)   nmlkj Recreational/Social (e.g., movie, sport)   nmlkj Dining/Restaurant   nmlkj Shopping   nmlkj Work   nmlkj Other   nmlkj Car Driver   nmlkj Car Passenger   nmlkj Public Transit   nmlkj School Bus   nmlkj Taxi   nmlkj Walk   nmlkj Bike   nmlkj Other   nmlkj 164Page 5 10. Who decide how you travel to school (i.e. which travel mode to use)? 11. What are the factors that influence your current travel mode choice to school?  12. How would you rate the safety level of your daily commute to school? 13. If you were free to choose any mode of transport to and from school, what is the ideal mode you would choose?  2. Your grade 3. What is your name?   4. What is your home postal code?   5. How many cars does your household own? 6. Do you have your own car?   4. Personal and Household Information 1. Your gender Myself   nmlkj Parents / Guardian   nmlkj Cost   gfedc Reliable   gfedc Comfortable   gfedc Convenient   gfedc Environmental   gfedc Health   gfedc Road Safety   gfedc Personal Safety   gfedc Social   gfedc Very Safe   nmlkj Safe   nmlkj Unsafe   nmlkj Very unsafe   nmlkj Walk   nmlkj Public Transit   nmlkj School Bus   nmlkj Car Driver   nmlkj Car Passenger   nmlkj Taxi   nmlkj Bike   nmlkj Other   nmlkj Male   nmlkj Female   nmlkj 8   nmlkj 9   nmlkj 10   nmlkj 11   nmlkj 12   nmlkj 0   nmlkj 1   nmlkj 2   nmlkj 3 or more   nmlkj Yes   nmlkj No   nmlkj 165Page 6 7. What is your household annual income before taxes? 8. Number of household members   9. Number of children in your household (18 years or younger)   10. Number of household members with driver licenses   11. Do you have a driver license? 12. What is your home dwelling type? 13. What is the tenure of your home? 14. How long have you been living in your current home? 15. What is your e­mail address?   Less than $25,000   nmlkj $25,000 to $49,999   nmlkj $50,000 to $74,999   nmlkj $75,000 to $99,999   nmlkj $100,000 to $149,999   nmlkj $150,000 and over   nmlkj Yes   nmlkj No   nmlkj House   nmlkj Duplex   nmlkj Townhouse   nmlkj Low Rises   nmlkj High Rises   nmlkj Other (please specify)  Own   nmlkj Rent   nmlkj Other   nmlkj less than 3 months   nmlkj 3 months to 1 year   nmlkj 1 year to 3 years   nmlkj 3 years to 5 years   nmlkj more than 5 years   nmlkj 166Appendix B: One way ANOVA of Environmental Knowledge and Attitudes by socioeconomic status  167APPENDIX B: One way ANOVA of Environmental Knowledge and Attitudes by socioeconomic status ANOVA of GENKN OW                                      Sum of Squares  df  Mean Square  F  P - Value  F crit  Significant  Gender  Betwee n Gro ups  0.7  1  0.66  0.22  0.64  3.85  No  Within Gro ups  259 0 .6  860  3.01        Total  259 1 .2  861          Grade  Betwee n Gro ups  22.4  4  5.59  1.84  0.12  2.38  No  Within Gro ups  261 0 .3  860  3.04        Total  263 2 .7  864          Inco me  Betwee n Gro ups  5.6  2  2.82  0.91  0.40  3.02  No  Within Groups  134 6 .4  434  3.10        Total  135 2 .0  436          HouseCar  Betwee n Gro ups  2.7  3  0.89  2.37  0.07  2.62  No  Within Gro ups  295 .9  791  0.37        Total  298 .5  794          DriveL ic  Betwee n Gro ups  12.3  1  12.25  4.04  0.04  3.85  Yes  Within Gro ups  257 6 .7  849  3.03        Total  258 8 .9  850          Residency  Betwee n Gro ups  11.2  2  5.61  1.86  0.16  3.01  No  Within Gro ups  255 7 .1  850  3.01        Total  256 8 .3  852          HouseP pl  Betwee n Gro ups  17.4  4  4.36  1.48  0.21  2.38  No  Within Gro ups  241 2 .1  817  2.95        Total  242 9 .5  821          HouseL ic  Betwee n Gro ups  17.0  4  4.24  1.44  0.22  2.38  No  Within Gro ups  241 8 .0  823  2.94        Total  243 5 .0  827          Children  Betwee n Gro ups  10.4  4  2.61  0.88  0.48  2.38  No  Within Gro ups  245 4 .1  826  2.97        Total  246 4 .6  830          OwnCar  Betwee n Gro ups  51.4  1  51.44  17.33  0.00  3.85  Yes  Within Gro ups  256 2 .3  863  2.97        Total  261 3 .8  864          Dwelling  Betwee n Gro ups  12.6  4  3.14  1.04  0.39  2.38  No  Within Gro ups  243 4 .1  806  3.02        Total  244 6 .7  810          YearHome  Betwee n Gro ups  44.6  4  11.16  3.73  0.01  2.38  Yes  Within Gro ups  253 4 .3  847  2.99        Total  257 8 .9  851          168ANOVA of TRANKNOW                                      Sum of Squares  df  Mean Square  F  P - Value  F crit  Significant  Gender  Betwee n Gro ups  0.8  1  0.84  0.56  0.46  3.85  No  Within Gro ups  128 9 .0  860  1.50        Total  128 9 .8  861          Grade  Betwee n Gro ups  6.7  4  1.68  1.12  0.35  2.38  No  Within Gro ups  129 1 .0  860  1.50        Total  129 7 .7  864          Inco me  Betwee n Gro ups  3.7  2  1.83  1.22  0.30  3.02  No  Within Gro ups  652 .3  434  1.50        Total  655 .9  436          HouseCar  Betwee n Gro ups  1.4  3  0.46  1.27  0.28  2.62  No  Within Gro ups  286 .3  791  0.36        Total  287 .7  794          DriveL ic  Betwee n Gro ups  1.5  1  1.51  1.00  0.32  3.85  No  Within Gro ups  127 4 .3  849  1.50        Total  127 5 .8  850          Residency  Betwee n Gro ups  8.8  2  4.39  2.94  0.05  3.01  No  Within Gro ups  126 9 .2  850  1.49        Total  127 8 .0  852          HouseP pl  Betwee n Gro ups  24.0  4  5.99  4.12  0.00  2.38  Yes  Within Gro ups  118 7 .9  817  1.45        Total  121 1 .8  821          HouseL ic  Betwee n Gro ups  9.0  4  2.24  1.52  0.19  2.38  No  Within Gro ups  121 0 .3  823  1.47        Total  121 9 .3  827          Children  Betwee n Gro ups  16.5  4  4.13  2.81  0.02  2.38  Yes  Within Gro ups  121 1 .2  826  1.47        Total  122 7 .7  830          OwnCar  Betwee n Gro ups  8.1  1  8.10  5.38  0.02  3.85  Yes  Within Gro ups  129 8 .8  863  1.50        Total  130 6 .9  864          Dwelling  Betwee n Gro ups  12.2  4  3.05  2.01  0.09  2.38  No  Within Gro ups  122 2 .7  806  1.52        Total  123 4 .9  810          YearHome  Betwee n Gro ups  14.7  4  3.67  2.46  0.04  2.38  Yes  Within Gro ups  126 1 .3  847  1.49        Total  127 6 .0  851            169ANOVA of GENATT                                      Sum of Squares  df  Mean Square  F  P - Value  F crit  Significant  Gender  Betwee n Gro ups  0.0  1  0.01  0.03  0.86  3.85  No  Within Gro ups  300 .6  796  0.38        Total  300 .6  797          Grade  Betwee n Gro ups  1.3  4  0.33  0.89  0.47  2.38  No  Within Gro ups  299 .4  798  0.38        Total  300 .7  802          Inco me  Betwee n Gro ups  2.5  2  1.25  3.58  0.03  3.02  Yes  Within Gro ups  141 .8  405  0.35        Total  144 .3  407          HouseCar  Betwee n Gro ups  2.7  3  0.89  2.37  0.07  2.62  No  Within Gro ups  295 .9  791  0.37        Total  298 .5  794          DriveL ic  Betwee n Gro ups  2.6  1  2.64  7.26  0.01  3.85  Yes  Within Gro ups  285 .9  785  0.36        Total  288 .5  786          Residency  Betwee n Gro ups  0.4  2  0.18  0.49  0.61  3.01  No  Within Gro ups  292 .5  788  0.37        Total  292 .9  790          HouseP pl  Betwee n Gro ups  2.1  4  0.52  1.39  0.23  2.38  No  Within Gro ups  280 .5  754  0.37        Total  292 .9  790          HouseL ic  Betwee n Gro ups  1.1  4  0.27  0.73  0.57  2.38  No  Within Gro ups  285 .9  760  0.38        Total  287 .0  764          Children  Betwee n Gro ups  2.1  4  0.52  1.40  0.23  2.38  No  Within Gro ups  284 .2  761  0.37        Total  286 .3  765          OwnCar  Betwee n Gro ups  0.5  1  0.52  1.38  0.24  3.85  No  Within Gro ups  300 .6  797  0.38        Total  301 .1  798          Dwelling  Betwee n Gro ups  1.1  4  0.28  0.74  0.56  2.38  No  Within Gro ups  284 .4  748  0.38        Total  285 .5  752          Duratio n  Betwee n Gro ups  1.0  4  0.25  0.69  0.60  2.38  No  Within Gro ups  289 .9  788  0.37        Total  290 .9  792           170ANOVA of TRANATT                                      Sum of Squares  df  Mean Square  F  P - Value  F crit  Significant  Gender  Betwee n Gro ups  1.1  1  1.06  2.91  0.09  3.85  No  Within Gro ups  289 .1  796  0.36        Total  290 .2  797          Grade  Betwee n Gro ups  2.6  4  0.64  1.78  0.13  2.38  No  Within Gro ups  286 .5  798  0.36        Total  289 .0  802          Inco me  Betwee n Gro ups  1.1  2  0.54  1.48  0.23  3.02  No  Within Gro ups  149 .3  405  0.37        Total  150 .4  407          HouseCar  Betwee n Gro ups  0.9  1  0.93  2.84  0.09  3.85  No  Within Gro ups  248 .2  758  0.33        Total  249 .1  759          DriveL ic  Betwee n Gro ups  0.2  1  0.17  0.48  0.49  3.85  No  Within Gro ups  279 .1  785  0.36        Total  279 .3  786          Residency  Betwee n Gro ups  1.2  2  0.59  1.67  0.19  3.01  No  Within Gro ups  279 .3  788  0.35        Total  280 .5  790          HouseP pl  Betwee n Gro ups  2.1  4  0.52  1.46  0.21  2.38  No  Within Gro ups  269 .1  754  0.36        Total  271 .2  758          HouseL ic  Betwee n Gro ups  3.8  4  0.94  2.62  0.03  2.38  Yes  Within Gro ups  272 .6  760  0.36        Total  276 .4  764          Children  Betwee n Gro ups  0.2  4  0.05  0.15  0.96  2.38  No  Within Gro ups  272 .1  761  0.36        Total  272 .3  765          OwnCar  Betwee n Gro ups  1.1  1  1.09  3.06  0.08  3.85  No  Within Gro ups  283 .9  797  0.36        Total  285 .0  798          Dwelling  Betwee n Gro ups  0.9  4  0.23  0.64  0.63  2.38  No  Within Gro ups  268 .3  748  0.36        Total  269 .3  752          YearHome  Betwee n Gro ups  0.3  4  0.07  0.19  0.94  2.38  No  Within Gro ups  282 .4  788  0.36        Total  282 .7  792           171Appendix C: Measurement Equation for the Latent Variables  172APPENDIX C:  MEASUREMENT EQUATIONS FOR THE LATENT VARIBLES  M o del 1a: ST UD ENT - DECISION - M AKER   Latent Facto rs  ENVATT  Measured by  Estim ate  Est./ S.E.  x1  0.520  15.7  x4  0.413  11.7  x8  0.605  18.8  x1 1  0.445  12.1  x3  0.371  10.0  x1 4  0.597  18.4  x1 5  0.381  10.3  x6  0.373  10.8  x7  0.514  15.3  x1 0  0.561  17.5  x1 3  0.445  12.7  x2  0.481  13.5  x5  0.345  8.8  x1 2  0.376  10.5   Model 1a: PAREN T- D ECISION - M AKER   Latent Facto rs  ENVATT  Measured by  Estim ate  Est./ S.E.  x1  0.551  14.7  x4  0.412  10.4  x8  0.625  16.7  x1 1  0.428  11.7  x3  0.358  9.3  x1 4  0.610  16.6  x1 5  0.376  9.5  x6  0.416  10.1  x7  0.525  14.3  x1 0  0.563  15.6  x1 3  0.481  12.8  x2  0.512  13.0  x5  0.368  8.7  x1 2  0.394  10.0   173M o del 1b: ST UDEN T- D ECISION - M AKER   Latent Facto rs  GENATT  TRANATT  Measured by  Estim ate  Est./ S.E.  Estim ate  Est./ S.E.  x1  0.511  15.1  n.a.  x4  0.428  12.1  n.a.  x8  0.620  18.4  n.a.  x1 1  0.454  12.3  n.a.  x3  0.373  9.9  n.a.  x1 4  0.608  17.1  n.a.  x1 5  0.389  10.3  n.a.  x6  n.a.  0.374  10.4  x7  n.a.  0.550  16.0  x1 0  n.a.  0.600  18.1  x1 3  n.a.  0.465  12.9  x2  n.a.  0.490  13.1  x5  n.a.  0.375  9.4  x1 2  n.a.  0.372  9.7   Model 1b: PA REN T -D ECISION - M AKER   Latent Facto rs  GENATT  TRANATT  Measured by  Estim ate  Est./ S.E.  Estim ate  Est./ S.E.  x1  0.559  14.6  n.a.  x4  0.446  10.7  n.a.  x8  0.670  16.9  n.a.  x1 1  0.454  11.9  n.a.  x3  0.375  9.3  n.a.  x1 4  0.643  16.2  n.a.  x1 5  0.400  9.6  n.a.  x6  n.a.  0.413  9.7  x7  n.a.  0.572  14.2  x1 0  n.a.  0.604  15.6  x1 3  n.a.  0.505  12.8  x2  n.a.  0.519  12.8  x5  n.a.  0.402  9.1  x1 2  n.a.  0.388  9.3   174Appendix D: Interview Questions  175APPENDIX D:  INT ER VI EW QUESTIONS  Questions for Sustainable Coordinators 1 .  Explain yo ur positio n and the role yo u play in env ironm ental educatio n . 2.  Ho w well do yo u think env iro nmental educatio n is included in the curriculum ?  3.  Are there currently any mandato ry co urses in high scho o l that relate to env ironmenta l studies? 4.  What are some of the “outside the classroom” activities? 5.  The Enviro nm ental Learning and Experience Curriculum Map /Guide assist BC teachers to inte grate enviro nmental concepts into teaching and learning.  Is it mandato ry or vo luntary fo r the teachers to fo llo w the guide? 6.  Any schoo ls are offering sustainability co urse?  In the sustainability co urse guide, one of the modules is on sustainable design and transportat io n.  Do you kno w which schoo l is currently teaching that to pic?   7.  Are teachers being intro duced to sustainability educatio n?  Is there training for the teachers? 8.  Ho w the teachers integrate env iro nmental educatio n into their practice?  Are there any too ls, guides and reso urces fo r the teachers to use? 9.  What are so me of the plans to advance sustainability educatio n?   Are there any plans to teach the students mo re on sustainable t ransportat io n? 10.  What are so me of the obstacles to enviro nmental educatio n in Vanco uv er schoo ls? 11.  What organizatio ns are working with the scho o l board? 12.  What is the so urce of funding for the env ironm ental educatio n pro gram? Questions for Teachers 1 .  At what school do yo u teach? 2.  What grade( s) do yo u teach? 3.  What subject( s) do yo u teach? 4.  D o yo u incorpo rate env ironm ental educatio n in yo ur lesso ns? 5.  Ho w often do yo u incorpo rate env iro nmental educatio n into yo ur lesso ns? 6.  D o yo u teach to pics relate d to the enviro nment and transport atio n? 7.  Can yo u explain what you teach / co ver in yo ur co urse(s) on the enviro nment and transport atio n? 8.  Hav e yo u ever enco uraged yo ur students to beco me env ironm entally responsible? 1769 .  Hav e yo u ever encouraged yo ur students to use sustainable t ransport? 10.  Of the enviro nmental educatio n yo u inco rpo rate into yo ur lesso ns, how much of it is based o n the curriculum and how much is yo ur own initiative? 11.  What techniques do yo u use to help yo ur students learn about the env ironm ent? 12.  What reso urces do yo u use and how do yo u use them? 13.  Ho w well do yo u think env iro nmental educatio n is included in the curriculum ? 14.  What do yo u think is the mo st effective way to include env iro nmental educatio n in the classroo m? What too ls wo uld be helpful? 15.  What are the obstacles to inco rpo rate enviro nmental educatio n in BC schoo ls? 16.  What are yo ur opinio ns on improv ing env ironm ental educatio n in BC? 17.  Are yo u aware of the 200 7 "Enviro nmental Learning and Experience Guide" distribute d by the BC Ministry of Educatio n? 18.  Ho w useful do / did yo u find this docum ent? 19.  Sho uld British Co lum bia dev elo p policy on inco rpo rating env ironm ental e ducation in the schoo ls? 20.  Hav e yo u perso nally been inv o lved in any of the fo llowing env iro nmental activ ities?   177Appendix E: Sample MPLUS Code  178APPENDIX E: SAMPLE MPLUS CODE TITL E: Env ironm ental Kno wledge and Attitudes DATA: FIL E IS EnvAtt.tx t; VARIABL E:       NAMES ARE       CollID  StudID   x1 -x61       Male Grade HouseCar OwnCar Income Ho useP pl       Children Ho useL ic DriL ic Dwell ing       Tenure YearHom e       Kno w GenKno w TranKno w       SchDri SchPass SchTran SchBus SchTaxi       SchWalk SchBike SchOth SchCar SchTB SchWB;        Missing are All (-99 99 );       USEVARIABLES are       x3 x4 x5 x6 x7 x8 x9       x10 x12 x13 x14       x15 x16 x17       Male Grade OwnCar HouseP pl       Children Ho useL ic DriL ic GenKno w TranKnow YearHo me       Duplex To wnho us Lo wRise HighRise Rent       LowIn HighIn;    DEFINE:             x5=6 -x5 ;             x6=6 -x6 ;             x8=6 -x8 ;             x10=6 -x10 ;             x13=6 -x13 ;             x15=6 -x15 ;             Duplex = (Dwelling == 2);             Townho us = (Dwelling == 3);             LowRise = (Dwe lling == 4);             HighRise = (Dwe lling == 5);             Rent = (Tenure == 1);             LowIn = (Incom e <=3);             HighIn = (Incom e >=5 AND Incom e <=6 ); 179  ANALYSIS:           PROCESSORS=8;   MOD EL :           GenAtt BY x3 x6 x10 x13 x5 x16 x17;           TranAtt BY x8 x9 x12 x1 5 x4 x7 x14;           GenAtt TranAtt GenKno w TranKnow ON Male Grade OwnCar           HouseP pl Children HouseL ic DriLic YearHom e Duplex To wnho us Lo wRise           HighRise Rent LowIn HighIn;           GenKno w WITH TranKno w;           GenAtt TranAtt ON GenKnow TranKno w;           GenAtt WITH TranAtt;           x7 with x4 x9;           x6 with x5 ;           x13 with x15 x8;           x17 with x16 ;           x10 with x4 x7 x16 ;           x3 with x1 2 x4;    MOD EL IND IRECT:         GenAtt IND Male;         GenAtt IND OwnCar;         GenAtt IND Duplex;         TranAtt IND Male;         TranAtt IND Grade;         TranAtt IND HighIn;    OUTP UT:             TECH4 MOD IND ICES SAM P STAT STAND ARDIZED ; ______ ___ ___ ___ _____ ____ ____ ___ ___ ___ _____ ____ ____ ___ ___ ___ _____ ____ ____ ___ ___ ___ ____ TITL E: Env ironm ental Attitudes and Travel Behavio r DATA:     FILE IS Env Att.txt ; VARIABL E:     NAMES ARE 180    CollID  StudID x1 -x41 Walk Bike Tran Bus Train     Taxi Driver Pass x50 -x61     Male Grade Ho useCar OwnCar Income Ho useP pl     Children Ho useL ic DriLic Dwe lling     Tenure YearHo me     Kno w GenKno w TranKnow     SchDri SchPass SchTran SchBus SchTaxi     SchWalk SchBike SchOth SchCar SchTB SchWB     AftDri AftP ass AftT ran AftBus AftTaxi     AftWalk AftBike AftOth AftCar AftTB AftWB;         Missing are All (-9 99 9 );     USEVARIABLES are     x3 x4 x5 x6 x7 x8 x9      x10 x12 x13 x14      x15 x16 x17     Male Grade Ho useCar OwnCar HouseP pl     Children Ho useL ic DriLic YearHome SchWB SchTB     Duplex Townho us LowRise HighRise Rent       LowIn HighIn ;      GROUP ING Is x5 0 (1=st udent 2=parent );  DEFINE:           x5=6 -x5;           x6=6 -x6;           x8=6 -x8;           x10=6 -x1 0;           x13=6 -x1 3;           x15=6 -x1 5;           Duplex = (Dwe lling == 2);           Townho us = (Dwe lling == 3);           LowRise = (Dwelling == 4);           HighRise = (Dwelling == 5);           Rent = (Tenure == 1);           LowIn = (Incom e <=3);           HighIn = (Incom e >=5 AND Incom e <=6 );   181ANALY SIS:         PROCESSORS=8 ; MOD EL:         EnvAtt BY x3 x6 x10 x13 x5 x16 x17 x8 x9 x12 x15 x4 x7 x14 ;         EnvAtt SchWB SchTB ON Male Grade Ho useCar OwnCar           HouseP pl Children HouseL ic DriLic YearHom e Duplex To wnho us Lo wRise          HighRise Rent LowIn HighIn;          EnvAtt ON SchWB SchTB;         x7 with x4 x9;         x6 with x5;         x13 with x1 5 x8;         x16 with x1 7;         x10 with x4 x7 x16;         x3 with x12 x4;  MOD EL IND IRECT:       EnvAtt IND Male;        EnvAtt IND Grade;        EnvAtt IND HouseCar;        EnvAtt IND OwnCar;       EnvAtt IND HouseP pl;       EnvAtt IND Children;       EnvAtt IND HouseL ic;       EnvAtt IND DriL ic;       EnvAtt IND YearHom e;       EnvAtt IND Duplex;       EnvAtt IND To wnHo us;       EnvAtt IND Lo wRise;       EnvAtt IND HighRise;       EnvAtt IND Rent;       EnvAtt IND Lo wIn;       EnvAtt IND HighIn;  OUTPUT:            TECH4 MOD IND ICES SAM P STAT STAND ARD IZED; ______ ___ ___ ___ _____ ____ ____ ___ ___ ___ _____ ____ ____ ___ ___ ___ _____ ____ ____ ___ ___ ___ ____ 182TITL E: Env ironm ental Kno wledge, Attitudes and Travel Behavio r DATA:     FILE IS Env Att .txt ; VARIABL E:     NAMES ARE     CollID  StudID x1 -x61     Male Grade Ho useCar OwnCar Income Ho useP pl     Children Ho useL ic DriLic Dwe lling     Tenure YearHo me     Kno w GenKno w TranKnow     SchDri SchPass SchTran SchBus SchTaxi     SchWalk SchBike SchOth SchCar SchTB SchWB     AftDri AftP ass AftT ran AftBus AftTaxi     AftWalk AftBike AftOth AftCar Af tTB AftWB;        Missing are All (-9 99 9 );     USEVARIABLES are     x3 x4 x5 x6 x7 x8 x9      x10 x12 x13 x14      x15 x16 x17 Kno w     Male Grade Ho useCar OwnCar HouseP pl     Children Ho useL ic DriLic YearHome     SchTB SchWB     Duplex Townho us LowRise HighRise Rent       LowIn HighIn ;  DEFINE:           x5=6 -x5;           x6=6 -x6;           x8=6 -x8;           x10=6 -x1 0;           x13=6 -x1 3;           x15=6 -x1 5;           Duplex = (Dwe lling == 2);           Townho us = (Dwe lling == 3);           LowRise = (Dwelling == 4);           HighRise = (Dwelling == 5); 183          Rent = (Tenure == 1);           LowIn = (Incom e <=3);           HighIn = (Incom e >=5 AND Incom e <=6 );  ANALY SIS:         PROCESSORS=8 ; MOD EL:         EnvAtt BY x3 x6 x10 x13 x5 x16 x17 x8 x9 x12 x15 x4 x7 x14 ;         EnvAtt SchTB SchWB Kno w ON Male          Grade Ho useCar OwnCar Ho useP pl Children Ho useL ic          DriLic YearHom e Duplex To wnho us LowRise HighRise Rent LowIn HighIn;          SchTB SchWB on EnvAtt ;          EnvAtt on Know;         x7 with x4 x9;         x6 with x5;         x13 with x1 5 x8;         x16 with x1 7;         x10 with x4 x7 x16;         x3 with x12 x4;  OUTPUT:            TECH4 MOD IND ICES SAM P STAT STAND ARD IZED; 184

Cite

Citation Scheme:

    

Usage Statistics

Country Views Downloads
China 38 0
Canada 22 5
United States 16 1
Germany 6 2
Greece 6 0
India 4 0
France 4 0
Hong Kong 2 0
Rwanda 2 0
Malaysia 2 0
Philippines 2 0
Singapore 2 0
Taiwan 1 1
City Views Downloads
Beijing 23 0
Unknown 22 4
Shenzhen 10 0
Vancouver 6 5
Piraeus 6 0
Hamilton 5 0
Edmonton 4 0
Washington 4 0
Halifax 4 0
Guangzhou 3 0
Srinagar 3 0
Ashburn 2 0
Ottawa 2 0

{[{ mDataHeader[type] }]} {[{ month[type] }]} {[{ tData[type] }]}
Download Stats

Share

Embed

Customize your widget with the following options, then copy and paste the code below into the HTML of your page to embed this item in your website.
                        
                            <div id="ubcOpenCollectionsWidgetDisplay">
                            <script id="ubcOpenCollectionsWidget"
                            src="{[{embed.src}]}"
                            data-item="{[{embed.item}]}"
                            data-collection="{[{embed.collection}]}"
                            data-metadata="{[{embed.showMetadata}]}"
                            data-width="{[{embed.width}]}"
                            async >
                            </script>
                            </div>
                        
                    
IIIF logo Our image viewer uses the IIIF 2.0 standard. To load this item in other compatible viewers, use this url:
http://iiif.library.ubc.ca/presentation/dsp.24.1-0073322/manifest

Comment

Related Items