UBC Faculty Research and Publications

Conceptualizing physical activity parenting practices using expert informed concept mapping analysis Mâsse, Louise C; O’Connor, Teresia M; Tu, Andrew W; Hughes, Sheryl O; Beauchamp, Mark R; Baranowski, Tom Jun 14, 2017

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

Item Metadata


52383-12889_2017_Article_4487.pdf [ 622.51kB ]
JSON: 52383-1.0361132.json
JSON-LD: 52383-1.0361132-ld.json
RDF/XML (Pretty): 52383-1.0361132-rdf.xml
RDF/JSON: 52383-1.0361132-rdf.json
Turtle: 52383-1.0361132-turtle.txt
N-Triples: 52383-1.0361132-rdf-ntriples.txt
Original Record: 52383-1.0361132-source.json
Full Text

Full Text

RESEARCH ARTICLE Open AccessConceptualizing physical activity parentingpractices using expert informed conceptmapping analysisLouise C. Mâsse1*, Teresia M. O’Connor2, Andrew W. Tu1, Sheryl O. Hughes2, Mark R. Beauchamp3,Tom Baranowski2 and Physical Activity Parenting Expert GroupAbstractBackground: Parents are widely recognized as playing a central role in the development of child behaviors suchas physical activity. As there is little agreement as to the dimensions of physical activity-related parenting practicesthat should be measured or how they should be operationalized, this study engaged experts to develop anintegrated conceptual framework for assessing parenting practices that influence multiple aspects of 5 to 12 yearold children’s participation in physical activity. The ultimate goal of this study is to inform the development of anitem bank (repository of calibrated items) aimed at measuring physical activity parenting practices.Methods: Twenty four experts from 6 countries (Australia, Canada, England, Scotland, the Netherlands, & UnitedStates (US)) sorted 77 physical activity parenting practice concepts identified from our previously publishedsynthesis of the literature (74 measures) and survey of Canadian and US parents. Concept Mapping software wasused to conduct the multi-dimensional scaling (MDS) analysis and a cluster analysis of the MDS solution of theExpert’s sorting which was qualitatively reviewed and commented on by the Experts.Results: The conceptual framework includes 12 constructs which are presented using three main domains ofparenting practices (neglect/control, autonomy support, and structure). The neglect/control domain includes twoconstructs: permissive and pressuring parenting practices. The autonomy supportive domain includes fourconstructs: encouragement, guided choice, involvement in child physical activities, and praises/rewards for theirchild’s physical activity. Finally, the structure domain includes six constructs: co-participation, expectations,facilitation, modeling, monitoring, and restricting physical activity for safety or academic concerns.Conclusion: The concept mapping analysis provided a useful process to engage experts in re-conceptualizingphysical activity parenting practices and identified key constructs to include in measures of physical activityparenting. While the constructs identified ought to be included in measures of physical activity parenting practices,it will be important to collect data among parents to further validate the content of these constructs. In conclusion,the method provided a roadmap for developing an item bank that captures key facets of physical activity parentingand ultimately serves to standardize how we operationalize measures of physical activity parenting.Keywords: Physical activity, Parenting practices, Children, Concept mapping, Measurement* Correspondence: lmasse@bcchr.ubc.ca1BC Children’s Hospital Research Institute, School of Population and PublicHealth, University of British Columbia, F508-4480 Oak Street, Vancouver, BCV6H 3V4, CanadaFull list of author information is available at the end of the article© The Author(s). 2017 Open Access This article is distributed under the terms of the Creative Commons Attribution 4.0International License (http://creativecommons.org/licenses/by/4.0/), which permits unrestricted use, distribution, andreproduction in any medium, provided you give appropriate credit to the original author(s) and the source, provide a link tothe Creative Commons license, and indicate if changes were made. The Creative Commons Public Domain Dedication waiver(http://creativecommons.org/publicdomain/zero/1.0/) applies to the data made available in this article, unless otherwise stated.Mâsse et al. BMC Public Health  (2017) 17:574 DOI 10.1186/s12889-017-4487-1BackgroundParents are widely recognized as playing a central role inthe development of child behaviors that influence risk ofchildhood obesity [1–4]. Interventions aimed at reducingchildhood obesity, have shown that the familial environ-ment plays a significant role in enabling healthy behav-iours and in influencing children’s physical activity [5, 6].While levels of physical activity in children are known tobe influenced by a range of individual, social and envir-onmental factors, parents play a critical role in socializ-ing their children to be physically active through theirchildrearing parenting styles and practices [7].Parenting styles and parenting practices are the termsused to describe how parents communicate with theirchild about their behaviors. Parenting styles highlightthe overarching affective childrearing behaviors thatparents use to interact with their child across differentcontexts [8]. In contrast, parenting practices refer tocontent and context specific childrearing approachesparents use to bring about certain childrearing outcomessuch as child engagement in physical activity [9]. Parent-ing practices are thought to be important for influencingchild behaviors such as physical activity. Recent reviews[10–13] identified parental encouragement, modeling,co-participation, and logistic support to be associatedwith children’s physical activity. However, the findingsacross these reviews are inconsistent. Currently, littleagreement exists in terms of the dimensions of physicalactivity-related parenting practices that should be mea-sured or how they should be operationalized. This lackof consistency has made it difficult to compare resultsacross studies [14] thereby limiting our ability to fullyunderstand how parents influence children’s physicalactivity behaviors and how interventions may mostefficiently and effectively be developed to positivelyinfluence parenting practices.To improve comparisons across studies, one solutionis to utilize Item Response Modeling (IRM) itembanking which creates a repository of calibrated items.Item banking has been used to address some of themeasurement challenges faced in other fields (i.e.,patient-reported outcomes) including being able tocompare results across studies when researchers usedifferent measures and reducing participant burdenamong others [14–16]. IRM item banking supplementedwith Computerized Adaptive Testing allows researchersthe flexibility to select which items to include in a studywhile maintaining the ability to compare scores for aspecific dimension across studies [15–17]. A physicalactivity parenting item bank requires that a conceptualframework guide the operationalization of underlyingdimensions. As there is little agreement on how mea-sures of physical activity parenting should be operation-alized, [14] this study engaged researchers who haveexpertise in physical activity parenting practices todevelop an integrated conceptual framework for asses-sing the practices that influence multiple aspects of 5 to12 year old children’s participation in physical activity.MethodsParticipants/expertsEach expert recruited to develop the physical activityconceptual framework had to be a leading authority in: 1)developing family interventions aimed at treating or pre-venting childhood obesity and/or modifying health behav-iors associated with obesity; and/or 2) studying the etiologyof children’s obesity from the perspective of parenting andfamilies. Experts were identified by: 1) reviewing themembership list of the International Society of BehavioralNutrition and Physical Activity (ISBNPA) and the list ofattendees of the 2012 pre-ISBNPA meeting, as it focusedon improving measures of physical activity and food-relatedparenting practices and general parenting styles; 2) review-ing citations within recent reviews published on this topic;[11, 18] 3) conducting searches on PubMed, ERIC,PsycINFO, and ScienceDirect; and 4) talking to ournetwork of researchers. Thirty scientists from the expertsearch were invited, of whom 20 participated (67% responserate) and were remunerated for devoting a day to thisinitiative. Four members of the research team (MRB, TB,TMO, SOH) also participated. In total, 24 experts from sixcountries (Australia, Canada, England, Scotland, theNetherlands, and US) provided input in conceptualizingthe measures of physical activity parenting practices.ProceduresIdentification of physical activity parenting practicesConcept mapping procedures traditionally involve ex-perts in brainstorming to identify the specific constructs(in this case, parenting practices) to be included in theconceptual framework [19–21]. Given the extensivework previously conducted in this area, the physical ac-tivity parenting practices were identified by: 1) conduct-ing a review of published measures of physical activityparenting practices and 2) collecting qualitative datathrough semi-qualitative interviews from 134 parents of5 to 12 year old children to identify the practices theyself-report to enable their children to be more physicallyactive. These processes are fully described in a previouspaper that examined whether current measures includepractices that parents self-reported using [22]. Briefly,step one identified 74 measures which included a totalof 608 items that measured physical activity parentingpractices [22]. Step two identified 1378 parent responsesthat were coded as specific physical activity-relatedparenting practices [22]. In preparation for the Conceptmapping procedures, this data was collected as part ofthe larger study and published elsewhere [22].Mâsse et al. BMC Public Health  (2017) 17:574 Page 2 of 11Condensing the list of physical activity parenting practicesTo condense the list of physical activity parentingpractices identified from the literature search and self-reported by the parents, we followed the binning andwinnowing process developed by the National Institutesof Health PROMIS initiative [23]. The binning processconsisted of assigning primary and secondary codes toidentify items or parent responses that measured thesame construct. A list of 14 primary codes (autonomysupport, co-participation, encouragement, expressingnegative/positive emotions, lack of parental control,logistic support/facilitation, modeling, monitoring, pres-sure to be active, restriction, rewards and discipline,structure of the environment, teaching & reasoning) andbetween 1 and 5 secondary codes (for example logisticsupport and facilitation included the following 5 second-ary codes enrollment as well as providing equipment,financial, transportation, and general support; autonomysupport included the following two secondary codes:child choosing and negotiation) per primary code weredeveloped for this purpose [22]. Previous work by ourgroup reported how, the 608 items from the publishedliterature and 1378 parent responses were initially con-solidated to 126 unique key parenting practices [22].Additional work by our group to further reduce thenumber of physical activity parenting practice conceptsfor the Experts to review reduced these to 77 key phys-ical activity parenting practices. We used a consensusprocess to reduce the pool of parenting practices fromthe literature and from parent responses where two re-searchers and two members of the investigative teamhad to agree on all decisions made in reducing the data.These key parenting practices were not measurementitems, but instead captured the content from similaritems that measured a similar practice. For example, oneitem representing a physical activity parenting practicewas “Restrict [activity type] inside the house” with an ex-ample of activity type being active play, ball games, run-ning, riding tricycle/scooter. In the literature there couldhave been five items linked to this parenting practice,but for the sorting task the experts were only providedthe key practice and not the actual measurement items.Sorting the list of physical activity parenting practicesUsing the online Concept Mapping software (CS GlobalMax version from Concept Systems Inc., Ithaca, NewYork), experts were asked to sort the 77 key parentingpractices identified from the published literature andparents’ responses into groups that made sense to them(i.e., aggregating similar constructs together), and nameeach grouping. The only restriction was to not include amiscellaneous or “junk” group, but rather to put prac-tices that were not perceived to fit into a group or con-cept into its own unique group. As part of the invitation,experts reviewed the consent form for the study andwere provided access to the software only if they agreedto participate which was achieved by clicking on a box.AnalysisConcept mapping methods developed by Kane andTrochim [24] were used to: 1) have experts sort thephysical activity parenting practices identified both fromthe literature and parent semi-structured qualitativeinterviews, 2) quantitatively analyze the sorting usingMultidimensional Scaling (MDS) Analysis followed by acluster analysis, and 3) inductively develop the concep-tual framework by integrating the results from theconcept mapping analysis with expert input.Groupings developed by the experts were preliminarilyreviewed to determine whether the sorting was com-pleted and whether the experts followed the instructions.One expert grouped three single parenting practices intoa miscellaneous category. This group was subsequentlysplit into three singleton groups before proceeding withthe analyses as such miscellaneous groupings cannot beincluded in the analyses. The expert sorting was ana-lyzed with a non-parametric MDS analysis that extracteda two-dimensional solution. The selection of a two-dimensional solution was determined a-priori as the in-tent of this initial processing of the data was to create asimilarity matrix that could be visually displayed andinterpreted. In addition, Kruskal and Wish [25] foundthat when MDS is combined with a cluster analysis, atwo-dimensional solution is preferred. The MDSassigned an x/y coordinate to each physical activity par-enting practice which was displayed on a point map andqualitatively interpreted. Parenting practices located neareach other on the point map represent those that weregrouped together often by the experts and thus likelymeasure a similar construct. Overall fit of the MDSsolution is assessed by evaluating the stress value whichranges between 0 and 1. Acceptable stress valuestypically range from 0.205 to 0.365 when MDS is usedto develop a conceptual framework [26] (as opposed tobeing used in controlled psychometric evaluations wherelower stress values are expected) [25].A cluster analysis of the MDS solution was then per-formed. Trochim’s [24] procedures were followed toidentify the number of clusters retained. This iterativeprocess started with more clusters than anticipated andsequentially reduced the number of clusters by one toexamine whether the two combined clusters were con-ceptually similar. The procedure was stopped when itdid not make conceptual sense to further combine clus-ters. We arbitrarily started with a 28-cluster solution asthe starting point to examine all solutions with fewerclusters all the way to a 2-cluster solution. Determin-ation of the number of clusters retained integrated theMâsse et al. BMC Public Health  (2017) 17:574 Page 3 of 11results of the cluster analysis with a qualitative analysisof the concept map, which resulted in sharpening theshape of the clusters. This process involved examiningthe content of each cluster and determining whetherborderline parenting practices should remain in thecluster, be re-assigned to a nearby cluster, or should bemoved based on conceptual knowledge. While theanalytical process is quantitatively informed, there isstrong qualitative component to this type of analysis asdescribed by Trochim [24]. Importantly, the quantitativeresults provide the foundation for initiating the discus-sion among the experts but the resulting solution isqualitatively derived. Given the exploratory nature of thecluster analyses, the results aimed to identify the numberof concepts to include in our conceptual framework;however, further sharpening of the content within eachconstruct is likely to occur at the item creation stage.We a-priori opted to present the main constructs identi-fied by the experts under the three main parentingdomains of control, autonomy support, and structure asa recent expert paper highlighted its utility in the area ofphysical activity [14] and it utilizes a nomenclature usedin the nutrition field and by developmental psychologists[9, 27–29].Three members of the research team (AWT, LCM, &TMO), of which two were not involved in the sortingprocess (AWT & LCM) independently completed thisprocess, discussed their solutions, and iteratively reviewedtheir solutions until the three members agreed on an ini-tial solution. This initial solution was presented to thelarger team of investigators who suggested further modifi-cation. The modified initial solution was then presented tothe expert group to receive further input and ensure theexperts agreed with the final solution. Having the expertsreview and endorse the decisions lend further credibilityto this process. At all levels of the analyses, LCM andAWT took the lead in integrating the feedback receivedby the research team and the larger group of experts sincethey did not take part in the sorting. This process wasfollowed to reduce the possibility of having a specifictheory influence the selection of the clusters, although ourcollective knowledge of the current literature could havebiased some of this process.ResultsClustering of the physical activity parenting practicesThe stress value for the MDS solution was 0.285 (stand-ard deviation of 0.04)– within the range consideredacceptable for solutions used to develop conceptualframeworks [26]. The preliminary analysis conducted bythree members of the research team (AWT, LCM, &TMO), initially identified seven to eight potentialconstructs. This preliminary solution was presented anddiscussed with other members of the research team(MRB, TB & SOH). Together the research teamconverged on a solution with 12 constructs addressingvarious physical activity parenting practices. After the 12construct solution was reviewed by the expert group,further refinement and reshaping of the solution occurred.Figure 1 shows both the conceptual solution agreed upon(12 constructs are line shaped) and the 11-cluster MDSsolution (shaded shapes), to show discrepancies betweenthe MDS and conceptual solutions. We opted to map theconceptual solution onto the 11-cluster solution forcomparison purposes as it aligned best with the concep-tual solution. The statements for each of the constructslisted in Fig. 1 are found in Table 1.In comparison to the MDS solution, the conceptualsolution produced by the experts modified the boundar-ies of some clusters as the content of some statementsFig. 1 Two-dimension point map showing the 77 physical activity parenting statements clustered into 12 constructs (outlined in blue in theconceptual solution) superimposed onto the 11-cluster Multi-Dimensional Scaling solution (shaded in grey)Mâsse et al. BMC Public Health  (2017) 17:574 Page 4 of 11Table 1 Conceptual solution of the physical activity parentingpractice statements (n = 77) sorted by the experts (n = 24)Construct Number Abbreviated statementNeglect/ControlPermissive 15 Child has a television in bedroom*40 Allow child to watch TV or playvideo/computer games whenevers/he wants to*49a Allow child to be less active whenon vacation*77 Allow child to skip physical activityor sports when s/he wants toPressuring 4 Complain to or make child feelbad for not exercising9 Tell child that spending time infront of a screen is not good forhis/her health/eyesight/weight*23 Pressure child to try harder atsports or his/her physical activity26a Punish child if s/he is sedentaryinstead of being active (i.e. nosnacks or take away TV/computerprivileges)*34 Tell my child that s/he needs toexercise so that s/he can loseweight36 Show child that you are angrywhen s/he does not participate inregular physical activity or exercise42 Remind/nag child to exercise or bephysically active43 If child says ‘I don’t feel like walkingor bicycling there,’ try to get him/her to do this anyway51 Criticize or tell child s/he is notgood when doing certain sports61a Have a rule ‘If you want a treat,you need to exercise’76 Punish child by not allowinghim/her to take part in physicalactivity or sportsAutonomy SupportEncouragement 2a Tell child that physical activity willmake him/her look good5 Tell child that physical activity orvigorous exercise is good forhis/her health and will makehim/her feel good7a Show child examples of rolemodels (i.e. people who are active)to encourage him/her to be active12 Encourage child to participate inphysical activity or play sports(./in his/her free time)16 Encourage child to be less sedentary*27 Encourage child to be active forat least 60 min per dayTable 1 Conceptual solution of the physical activity parentingpractice statements (n = 77) sorted by the experts (n = 24)(Continued)29a Try to encourage child to dophysical activities by telling s/hewill make new friends41a Encourage active video games asa way to be active indoors54 Encourage child to walk, bike oruse resources (park or communitycenter) in neighborhood to be active55 Tell child s/he is doing well inphysical activities or sports65a Get child to be physically activeby telling how much fun theactivity isGuided choice 10 Allow child to choose whether s/heparticipates in sports or vigorousphysical activity in free time11 Negotiate with child on howmuch physical activity/sportss/he does28 Negotiate with child on howmuch TV/video/DVD s/he isallowed to watch*45 Provide child with physical activityoptions from which my childcan chooseInvolvement 8 Show an interest in child’s sportsby talking about his/her activities17 Involve child in active chores andyard work around the house25 Go to child’s sports or physicalactivities and watch childparticipate31 Watch sports with child, talk aboutsports with child, and take child tosports games, to encourageparticipation in physical activity50 Involved in child’s activities(e.g., coaching activities, watchingchild play)52 Spend time teaching child how toplay a sport or do certain physicalactivitiesPraises / Rewards 14a Tell child that you like it whens/he is physically active20 Reward child for exercising75 Praise child for being physicallyactive or for participating in sportsStructureCo-participation 6 Practice active habits with child(e.g. parking far from the door,taking the stairs)24 Play sports or active games withchild48 Invite child to join your exerciseor do something active with youMâsse et al. BMC Public Health  (2017) 17:574 Page 5 of 11seemed more conceptually related to nearby constructs.Such examples can be found in constructs E (co-participation),F (involvement), G (facilitation), H (expectations), J(pressuring), and K (praises/rewards). While it is importantTable 1 Conceptual solution of the physical activity parentingpractice statements (n = 77) sorted by the experts (n = 24)(Continued)57 Go for walks with child71 Use sport/physical activity as aform of family recreation(e.g., going on bike rides together,hiking, skating)Expectations 35 Make sure child uses activetransportation when going placesclose to home (e.g. walking, biking)58 Limit the amount of time childspends [sedentary activity] onweekend/weekday [playingcomputer games, watching TV,watching videos, electronic games,video games, on the phone]*64a Make child responsible for takingthe dog for a walk and/or playingwith the dog68 If the weather is nice, child knowsthat s/he is expected to playoutside73 Make sure child is physically activeat least 60 min per day74 Have a rule that child mustparticipate in active sports orphysical activitiesFacilitation 18 Buy/provide physical activity orsports equipment for child22 Take child to the park, playground,or places that s/he can bephysically active39 Help child find ways to reducehis/her sedentary habits47 Store child’s active toys/sportsequipment in a place that is easilyaccessible63 Enroll child in sports and physicalactivity programs67 Try to make physical activity into afun game to get child more active69 Arrange for child to be with friendsin order to be active with them70a Encourage competition or setchallenges (e.g., walking a certaindistance) during activities to getchild more activeModeling 46 Child sees you being sedentary oris sedentary with you*60 If you would like to watchTV/video/DVD, you restrainyourself because of the presenceof child*72 Use own active behavior toencourage child to be physicallyactiveMonitoring 44 Keep track of the amount of physicalactivity or exercise child getsTable 1 Conceptual solution of the physical activity parentingpractice statements (n = 77) sorted by the experts (n = 24)(Continued)53 Keep track of the amount of timechild spends in front of screens(e.g television, computer)*Restriction forsafety/academic concerns1 Restrict child’s outdoor activitiesbecause neighborhood is not safe13 When child plays outside, s/hemust be supervised21 Don’t allow child to play oncommunity or sports teams(./so s/he can concentrate onschoolwork)30 Have rules that child is notallowed to walk to theneighborhood park alone37a Have a rule that child must dohomework before s/he is able toexercise or be physically active38 Don’t allow child to play outsidein the street after dark or after acertain time56a Prohibit child from playing certainsports59 Restrict some physical activitiesbecause afraid child will be hurt62 Restrict the amount of time childspends playing outside66 Restrict [activity type] inside thehouse [active play, ball games,running, riding tricycle/scooter]Drop 32 Reward child for good behaviorwith TV, DVD, or computer time– Drop because the behavior isunspecified and need practicesthat are more specific)*19 Do not enroll child in physicalactivities that are too expensive– Drop because (Capture socio-economic issues which does notfit with other items dropped fromclustering but can be included asa single item)3 Enroll child in too many activitiesleaving no time for free play(not specific to physical activity)33 Try to get child to be active(e.g. playing tag, biking, dancing)instead of watching TVor playing video games (Dropbecause how parent achieve thisis unspecified)*Practices related to sedentary behaviors were omitted from the operationaldefinition as the focus was on physical activityaKey practices that were identified from parent responses [22].Mâsse et al. BMC Public Health  (2017) 17:574 Page 6 of 11to consider the proximity of the statements, reshaping ofthe statistical clusters occurred when it made theoreticalsense to do so. Reshaping predominantly occurred when aparenting practice statement was thought to better fit con-ceptually with the operational definition of a neighboringcluster and both the larger team of investigators and ex-perts agreed with this decision. For example, the parentingpractice statement number 75 “Praise child for beingphysically active or for participating in sports” wasmerged with the praises/rewards construct as theconstruct already included a practice statement related topraise, namely number 14 “Tell child you like it when s/heis physically active”.In the MDS statistical solution, the monitoring andpermissive constructs were combined (see clusters Aand B, respectively); however, after examining thecontent of this combined cluster it was determined thesetwo concepts should not be combined. Although mer-ging of clusters A (monitoring) and B (permissive)appeared early in the clustering process (i.e., at the 21stcluster solution), it did not make conceptual sense toregroup them. As a result the cluster was separated.Construct G, labelled facilitation for physical activityemphasizes the ways in which parents support participa-tion in physical activity including financial assistance,provision of material goods, and planning and facilitat-ing physical activity. However, at the bottom of clusterG there are three practices which were not added to thiscluster as they fit better with nearby clusters (8 and 25fit better with involvement and 7 with encouragement).As they fall in between these two clusters, it suggestsless agreement as to what fits with the parental involve-ment and encouragement clusters.The MDS12-cluster solution shows four parenting prac-tice statements grouped as a thin and long cluster (state-ments 9, 14, 20, and 34 located on the right side of themap). Grouping of these four statements occurred at the13-cluster solution and in prior solutions, statements 14and 20 were grouped together and statements 9 and 34were grouped together. Evaluation of the map revealed thatstatements 14 “Tell child that you like it when s/he is phys-ically active”, 20 “Reward my child for exercising”, and 75“Praise child for being physically active or for participatingin sports” refer to praises/rewards and were regroupedunder construct K. In contrast, statements 9 “Tell child thatspending time in front of a screen is not good for his or herhealth/eyesight/weight” and 34 “Tell child that s/he needsto exercise so that s/he can lose weight” relate to pressuringthe child to be active for health concerns and were groupedwith construct J (pressuring).Although some practices regrouped into some clusters,we opted to delete some of them because they were eithertoo vague or they measured irrelevant concepts (state-ments 3, 19, and 32 were dropped). Statement 3 “enrollmy child in too many activities leaving no time for freeplay” was deleted from construct H (expectations) as thisconcept can have both positive and negative implicationsfor a child’s sustained participation in physical activity(increase their physical activity but decrease their sense ofautonomy). Statement 19 “Do not enroll child in physicalactivities that are too expensive” was deleted from con-struct I (restriction for safety or academic concerns) be-cause it captures socio-economic issues. While cost is arestriction to enrollment and participation in physicalactivity, it is preferable to measure this aspect separatelyas it affects only certain families. Finally, statement 32“reward my child for good behavior with TV, DVD, orcomputer time” was deleted from construct H (expecta-tions) as the “good behavior” was not specified and itemsthat capture this concept should be regrouped underconstruct K (praises/rewards).Finally, while most experts endorsed the conceptualsolution which includes 12 constructs, some identifiedpotential overlaps with some constructs – includingco-participation with involvement and encouragementwith praises/rewards. We opted to keep these constructsseparate at this stage, while acknowledging that furtherpsychometric work will enable us to shed light onwhether such an operationalization is supported, orwhether further refinement is needed. In addition, manyexperts indicated that the statements related to seden-tary behaviors should be excluded from the physical ac-tivity parenting practice item bank. These statements arehighlighted in Table 1 and our construct definitions haveeliminated these statements from the operationaldefinitions.Categorizing the physical activity parenting practices intoa recent parenting taxonomyFigure 2 shows how each of the physical activity parent-ing practice constructs can be grouped into higher-orderdomains of parenting: neglect/control, autonomy sup-port, and structure. Davison et al.’s [14] and Vaughn etal.’s [30] papers informed this categorization. Definitionsfor these domains are shown in Fig. 2 with Table 2 pro-viding the operational definitions for each construct.DiscussionTo minimize inconsistencies in measures of physicalactivity parenting practices, [11, 14, 16, 18] this studyutilized concept mapping methods to aggregate inputfrom experts to identify constructs of parenting prac-tices. Expert sorting of 77 parenting practice statementsidentified from a review of the literature and from inter-views with parents, resulted in identifying 12 constructswhich are presented using the three main domains ofgeneral parenting practices, namely neglect/control, auton-omy support, and structure. The neglect/control domainMâsse et al. BMC Public Health  (2017) 17:574 Page 7 of 11includes two constructs: permissive and pressuringparenting practices. The autonomy support domain in-cludes four constructs: encouragement, guided choice,involvement in child physical activities, and praises/rewards to encourage their children’s physically activ-ity. Finally, the structure domain includes six con-structs: co-participation, expectations, facilitation, modeling,monitoring, and restricting physical activity for safety oracademic concerns.The constructs were categorized under the three maindomains of parenting which have recently been used toclassify food and physical activity parenting practices[14, 30] and integrate terms from developmentalpsychology to characterize parenting more broadly [9,28, 29]. The neglect/control domain partially aligns withBaumrind’s definition of control which reflects the “claimsthat parents make on children to become integrated intosociety through behavior regulation, direct confrontation,and maturity demands (behavioral control), and supervi-sion of children’s activities” [28]. In our framework, theneglect/control domain includes the coercive componentsof Baumrind’s definition with some aspects of controlclassified under structure to highlight that some level ofcontrol is necessary to set the proper environment forchildren to be physically active. As a result, both expecta-tions and monitoring were classified under structure asthey provide necessary structure and boundaries to thechild. This aligns with Grolnick and Pomerantz’s [31]conceptualization of control which suggests regroupingthe dominating and pressuring parenting practicesunder control; whereas, control practices that offerguidance to the child should be regrouped understructure.Darling and Steinberg’s operationalized control (de-mandingness) in terms of “the parent’s willingness to actas a socializing agent” [9]. This prompted us to classifybeing permissive with control to capture a lack of“willingness to act as a socializing agent” as measuringthe opposite end of this continuum (neglect/control).Both constructs are independent of each other but areregrouped together as they capture less desirableparenting practices.The autonomy support domain aligns with Baumrind’sdefinition of responsiveness which includes “the extentto which parents foster individuality and self-assertionby being attuned, supportive, and acquiescent tochildren’s requests: it includes warmth, autonomy sup-port, and reasoned communication” [9, 28]. Finally, thestructure domain aligns with current definitions thatfocus on structuring the child’s environment [29] toachieve specific childrearing outcomes.Our classification differs slightly from Davison’s [14]physical activity and Vaughn’s [30] food parentingpractice classifications as: 1) we utilized different termin-ology to refer to one of the domains of parenting, whereour classification refers to what others have termeddemandingness or control as neglect/control; and 2) weclassified expectations under structure which followsVaughn’s classification and Grolnick and Pomerantzconceptualization of control [30, 31]. ImportantlyFig. 2 Categorizing the physical activity (PA) parenting practices into current parenting taxonomyMâsse et al. BMC Public Health  (2017) 17:574 Page 8 of 11regrouping the constructs into the three broad domainsof parenting shown in Fig. 2 does not imply that theconstructs measure a higher order factor. For example,the extent to which the constructs within the autonomysupport domain should be examined together or separ-ately will depend on the psychometric properties of thisdomain – whether the four constructs measure a higherorder factor or not. Based on our concept mappingresults, the encourage and praises/rewards constructsmay turn out to be highly correlated as they are proxim-ally located on the point map (see Fig. 1). However, theremaining constructs, namely involvement and guidedchoices, will likely measure independent dimensions.Instead, the classification is useful as it highlights theneed to examine overall profile of parenting and thatthis complexity needs to be accounted when one exam-ine their impact on children health behaviors.The exploratory nature of this analysis means that it ispossible for some statements to be misclassified.However, any misclassified statements did not result inidentifying new constructs which provided some validityfor the concept mapping results. Importantly, the analyt-ical process used in this paper identified the mainconstructs and provided some examples of parentingpractices that fit under these constructs. In the creationof the item bank, it will be important to maintain itemsthat match these operational definitions, but to considerhow theories and models of health behavior can informthe operationalization of these constructs (e.g., socialsupport models, [32] self-determination theory, [33]social cognitive theories, [34] socio-ecological models,[35] among others). Importantly, once we collect dataamong parents, we will be able to refine measures ofthese constructs and analyze whether all the constructsare independent or whether there is some overlapamong them as suggested by some of the experts.Enabling comparisons across studies is an essentialstep to elucidate the mechanisms through which parentscan influence children’s physical activity. This study willprovide the foundation for operationalizing measuresof physical activity parenting practices which can beused in observational and/or intervention studies. Theconstructs identified from the concept mapping analysiswill provide the foundation for developing an item bankcalibrated with Item Response Modeling [36] supportedwith computerized adaptive testing which will standardizethe measurement of parenting practices while allow-ing researchers some flexibility in selecting items ofinterest [15]. Specifically, utilizing the item bank withcomputerized adaptive testing will allow physicallyactivity researchers to tailor the measurement ofparenting practices and reduce the burden of com-pleting lengthy questionnaires. This process works byfirst having participants answer select items for a specificphysical activity parenting construct, with their responsesdetermining which items they receive next. The computerstops administering items for a specific physical activityparenting construct when the parental score on a givenconstruct can be estimated with enough precision. Withinintervention studies, baseline assessments of certainconstructs which are not often used by parents, suchTable 2 Definition of physical activity parenting practiceconstructsDomain/Construct DefinitionNeglect/Control• Permissive Parent does not guide their child’sbehaviors and allows them to decidewhether they engage in physical activity• Pressuring Parent criticizes, nags, forces, pressures,punishes, or uses threats to get theirchildren to be physically activeAutonomy Support• Encouragement Parent suggests or encourages child to bephysically active by explaining the reasonsfor being active, highlighting role modelsor provides positive verbal reinforcementfor doing so• Guided Choice Parent promotes independence indecisions related to physical activity byproviding child with options or bynegotiating with the child• Involvement Parent demonstrates an interest in thechild’s participation in physical activity orsports by watching child participate in his/her physical activity or sports, talking abouthis/her physical activities, teaching childnew skills, and volunteering/coaching inchild physical activity or sports• Praises/Rewards Parent positively reinforces participation inPA by verbally praising their child oracknowledging their participation withoutcoercing their participationStructure• Co-participation Parent engages in physical activity withtheir child• Expectations Parent sets clear expectations aboutphysical activity as to when and howmuch physical activity the child should do• Facilitation Parent positively supports child physicalactivity by getting them involved inactivities through enrollment or takingthem to places to be active, and bysupporting their physical activity (financialassistance, provision of equipment, servicessuch as transportation and planningphysical activities)• Modeling Parent models an active lifestyle• Monitoring Parent tracks child involvement in physicalactivity• Restriction for safety/academic concernsParental concerns about safety andacademic performance results in limitingchild involvement in physical activityMâsse et al. BMC Public Health  (2017) 17:574 Page 9 of 11as guided choices, [22] could be measured with feweritems at baseline and presumably with more items inthe follow-up if the physical activity intervention fo-cused on this aspect of parenting. Interestingly, theprocedure allows some flexibility of adding new itemsfor a given physical activity parenting construct, andas long as the researcher utilizes some of the itemsthat have been pre-calibrated in the item bank, theywill be able to compute a score for a given constructthat can be compared across studies. While the fieldof physical activity has not yet taken advantage ofthese advanced psychometric methods, there are suc-cessful examples in the fields of outcomes research(e.g., to measure quality of life) which can serve as modelsfor the field of physical activity (see www.nihpromis.org/NIH PROMIS initiative) [36].The concept mapping analysis identified a number ofcentral constructs that ought to be included in measuresof physical activity parenting practices. It is possible thatconstraining the number of statements provided to theexperts as well as the selection of specific statementscould have biased the types and numbers of constructsidentified. Furthermore, the statements provided to theexperts were in some instances more generic than theoriginal items or parent responses. As a result, thesesmall nuances were not captured and likely yieldedbroader constructs or can explain why some statementswere not clearly located in the clusters to which theyconceptually belong. While it is likely that other relevantconstructs have not been captured through our conceptmapping analysis, the ones identified likely need to beincorporated in future research and provide a basis formeasuring physical activity parenting practices.ConclusionsThe concept mapping analysis engaged experts in re-conceptualizing measures of physical activity parentingwhich provided an initial roadmap for developing anitem bank that captured 12 key physical activity parent-ing constructs.AbbreviationsAWT: Andrew W. Tu; DVD: Digital Video Disc; IRM: Item response modeling;ISBNPA: International Society of Behavioral Nutrition and Physical Activity;LCM: Louise C. Mâsse; MDS: Multidimensional Scaling; MRB: Mark R.Beauchamp; PA: Physical Activity; PROMIS: Patient Reported OutcomesMeasurement Information System; SOH: Sheryl O. Hughes; TB: TomBaranowski; TMO: Teresia M. O’Connor; TV: Television; US: United StatesAcknowledgementsThis work was supported by the Canadian Institutes of Health Researchgiven to LCM (201109MOP-2585565-PH2-CAAA-143786). Maria Valentehelped coordinate the data collection for this project. The Physical ActivityParenting Expert Group conducted the sorting task, provided input into thesolution, and critically reviewed the final manuscript.The Physical Activity Parenting Expert Group includes:• Maria Bryant PhD, University of Leeds, Leeds, UK, m.j.bryant@leeds.ac.uk• Kirsten Davison, PhD, Harvard University, Boston, Massachusetts, US,kdavison@hsph.harvard.edu• David A. Dzewaltowski, PhD, Kansas State University, Manhattan, Kansas US,dadx@ksu.edu• Wendy Grolnick, PhD, Clarke University, Worcester, MA, USA,wgrolnick@clarku.edu• Trish Gorely, PhD. University of Stirling – Highland Campus, Inverness, UK,trish.gorely@stir.ac.uk• Kylie D. Hesketh, PhD, Deakin University, Burwood VIC, Australia,kylieh@deakin.edu.au• Nicholas L. Holt, PhD, University of Alberta, Edmonton, AB, Canada,nick.holt@ualberta.ca• Russell Jago, PhD, University of Bristol, Bristol, UK, russ.jago@bristol.ac.uk• Leslie A. Lytle, PhD, The University of North Carolina at Chapel Hill, ChapelHill, NC, US, llytle@email.unc.edu• Jason A. Mendoza, MD, MPH, University of Washington, Seattle, WA, US,jason.mendoza@seattlechildrens.org• Philip Morgan, PhD, University of Newcastle, Callaghan NSW, Australia,Philip.Morgan@newcastle.edu.au• Katie L. Morton, PhD, University of Cambridge School of Clinical Medicine,Cambridge, UK, km576@medschl.cam.ac.uk• Ryan E. Rhodes, PhD, University of Victoria, Victoria BC Canada,rhodes@uvic.ca• Julie Saunders. PhD, The University of Western Australia, Crawley, WA,Australia, julie.saunders@uwa.edu.au• Ester FC Sleddens, PhD Maastricht University Medical Center+, Maastricht,The Netherlands, Ester.Sleddens@maastrichtuniversity.nl• Sandra C. Soto, MPH San Diego State University, San Diego, CA, US,sandra.soto@mail.sdsu.edu• Stewart G. Trost, PhD Queensland University of Technology, Brisbane, QLD,Australia, s.trost@qut.edu.au• Anna Timperio, PhD, Deakin University, Geelomg, VIC Australia,anna.timperio@deakin.edu.au• Gregory J. Welk, PhD, Iowa State University, Ames, Iowa, US,gwelk@iastate.edu• Dawn K. Wilson, PhD, University of South Carolina, Barnwell College,Columbia, SC, US, wilsondk@mailbox.sc.eduFundingLCM received salary support from the BC Children’s Hospital Research Institute(BCCHR). The funding agency conducted a peer review evaluation of the grantand was not involved in designing the study or in conducting the research.AWT received post-doctoral support from BCCHR and the Michael SmithFoundation for Health Research. SOH, TB, and TMO were funded in part byinstitutional support from the US Department of Agriculture, AgriculturalResearch Service (USDA/ARS) (Cooperative Agreement no. 58–6250–0-008).The content of this publication do not necessarily reflect the views or policiesof the USDA, nor does mention of organizations imply endorsement from theUS government. KDH is supported by an Australian Research Council FutureFellowship (FT130100637) & Honorary National Heart Foundation of AustraliaFuture Leader Fellowship (100370).Availability of data and materialsThe de-identified datasets used in the current study are available from thecorresponding author on reasonable request.Authors’ contributionsLCM, TMO, SOH, MRB and TB designed the study. LCM oversaw the datacollection. The Physical Activity Parenting Expert Group and TMO, SOH, MRB,and TB conducted the sorting of the parenting practices. LCM, TMO andAWT performed the analyses. SOH, MRB, and TB provided initial in the analyses.The Physical Activity Parenting Expert Group reviewed the solution andprovided in-depth input. LCM drafted the manuscript with help from AWT.TMO, SOH, MRB, and TB critically assessed draft of the manuscript.All authors read and approved the final manuscript.Competing interestsThe authors declare that they have no competing interests.Consent for publicationThis manuscript contains no individual person data, so not applicable.Mâsse et al. BMC Public Health  (2017) 17:574 Page 10 of 11Ethics approval and consent to participateThis research protocol was approved by the Research Ethics Board at theUniversity of British Columbia and received Institutional Review Board approvalfrom Baylor College of Medicine. All participants from the web-based panelprovided consented to be part of the panel. All experts provided consent priorto participating. Experts who expressed an interest in the study receivedinformation about the study and a copy of the consent form. When expertsreceived the invite to complete the task for this study they had to again reviewthe written consent form and clicked whether they agreed to participate in thisstudy to proceed any further.Publisher’s NoteSpringer Nature remains neutral with regard to jurisdictional claims inpublished maps and institutional affiliations.Author details1BC Children’s Hospital Research Institute, School of Population and PublicHealth, University of British Columbia, F508-4480 Oak Street, Vancouver, BCV6H 3V4, Canada. 2USDA/ARS Children’s Nutrition Research Center, BaylorCollege of Medicine, CNRC-2034 1100 Bates St, Houston, TX 77030, USA.3University of British Columbia, Rm 122 War Memorial Gymnasium, 6081University Boulevard Vancouver, Vancouver, BC, Canada.Received: 6 December 2016 Accepted: 1 June 2017References1. Barlow SE. Expert committee recommendations regarding the prevention,assessment, and treatment of child and adolescent overweight and obesity:summary report. Pediatrics. 2007;120(Suppl 4):S164–92.2. Faith MS, Van HL, Appel LJ, Burke LE, Carson JA, Franch HA, et al. Evaluatingparents and adult caregivers as "agents of change" for treating obesechildren: evidence for parent behavior change strategies and research gaps:a scientific statement from the American Heart Association. Circulation.2012;125:1186–207.3. Waters E, de Silva-Sanigorski A, Hall BJ, Brown T, Campbell KJ, Gao Y, et al.Interventions for preventing obesity in children. Cochrane Database Syst Rev.2011; CD001871.4. National Academies of Sciences EaM. Parenting matters: Supporting parentsand children ages 0–8. Washington, DC: The National Academies Press;2016.5. Loveman E, Al-Khudairy L, Johnson RE, Robertson W, Colquitt JL, Mead EL,et al. Parent-only interventions for childhood overweight or obesity inchildren aged 5 to 11 years. Cochrane Database Syst Rev. 2015; CD012008.6. Oude LH, Baur L, Jansen H, Shrewsbury VA, O'Malley C, Stolk RP et al.Interventions for treating obesity in children. Cochrane Database Syst Rev.2009; CD001872.7. O'Connor TM. Informing parent targeted interventions to promoteincreased physical activity among youth. J Appl Res Child. 2014;5:15.8. Baumrind D. Current patterns of parental authority. Dev Psychol.1971;4:1–103.9. Darling N, Steinberg L. Parenting style as context: an integrative model.Psychol Bull. 1993;113:487–96.10. Verloigne M, Van LW, Maes L, Brug J, De BI. Family- and school-basedcorrelates of energy balance-related behaviours in 10-12-year-old children:a systematic review within the ENERGY (EuropeaN ENERGY balance researchto prevent excessive weight gain among youth) project. Public Health Nutr.2012;15:1380–95.11. Trost SG, McDonald S, Cohen A. Measurement of general and specificapproaches to physical activity parenting: a systematic review. Child Obes.2013;9(Suppl):S40–50.12. Xu H, Wen LM, Rissel C. Associations of parental influences with physicalactivity and screen time among young children: a systematic review. J Obes.2015;54692513. Kaushal N, Rhodes RE. The home physical environment and its relationshipwith physical activity and sedentary behavior: a systematic review.Prev Med. 2014;67:221–37.14. Davison KK, Mâsse LC, Timperio A, Frenn M, Saunders J, Mendoza JA, et al.Physical activity parenting measurement and research: challenges,explanations, and solutions. Childhood Obesity. 2013;9:S103–9.15. Mâsse LC, Wilson M, Baranowski T, Nebeling L. Improving psychometricmethods in health education and health behavior research. Health Educ Res.2006;21(Suppl 1):i1–3.16. Mâsse LC, Watts AW. Stimulating innovations in the measurement ofparenting constructs. Child Obes. 2013;9(Suppl):S5–13.17. Mâsse LC, Heesch KC, Eason KE, Wilson M. Evaluating the properties of astage-specific self-efficacy scale for physical activity using classical testtheory, confirmatory factor analysis and item response modeling. HealthEduc Res. 2006;21(Suppl 1):i33–46.18. Sleddens EF, Kremers SP, Hughes SO, Cross MB, Thijs C, de Vries NK, et al.Physical activity parenting: a systematic review of questionnaires and theirassociations with child activity levels. Obes Rev. 2012; 13:1–15-1033.19. Trochim W. An introduction to concept mapping for program planning andevaluation. Evaluation and Program Planning. 1989;12:1–16.20. Burke JG, O'Campo P, Peak GL, Gielen AC, McDonnell KA, Trochim WM.An introduction to concept mapping as a participatory public healthresearch method. Qual Health Res. 2005;15:1392–410.21. Trochim W, Kane M. Concept mapping: an introduction to structuredconceptualization in health care. Int J Qual Health Care. 2005;17:187–91.22. Mâsse LC, O'Connor TM, Tu AW, Watts AW, Beauchamp MR, Hughes SO,et al. Are the physical activity parenting practices reported by US andCanadian parents captured in currently published instruments? J Phys ActHealth. 2016;13:1070–8.23. DeWalt DA, Rothrock N, Yount S, Stone AA. Evaluation of item candidates:the PROMIS qualitative item review. Med Care. 2007;45:S12–21.24. Kane M, Trochim WMK. Concept mapping for planning and evaluation.Thousand Oaks, CA: Sage Publications; 2007.25. Kruskal JB, Wish M. Multidimensional scaling. Beverly Hills, CA: SagePublications; 1978.26. Trochim WMK. The reliability of concept mapping. http://www.socialresearchmethods.net /research/Reliable/reliable.htm. Accessed 15 Aug 2016.27. O'Connor T, Watson K, Hughes S, Beltran A, Hingle M, Baranowski J, et al.Health professionals' and dietetics practitioners' perceived effectiveness offruit and vegetable parenting practices across six countries. J Am Diet Assoc.2010;110:1065–71.28. Baumrind D. Patterns of parental authority and adolescent autonomy.New Dir Child Adolesc Dev. 2005;108:61–9.29. Slater MA, Power TG. Multidimensional assessment of parenting in single-parent families. In: Vincent JP, editor. Advances in family intervention,assessment and theory. Greenwich, CN: JAI Press; 1987. p. 197–228.30. Vaughn AE, Ward DS, Fisher JO, Faith MS, Hughes SO, Kremers SP, et al.Fundamental constructs in food parenting practices: a content map toguide future research. Nutr Rev. 2016;74:98–117.31. Grolnick WS, Pomerantz EM. Issues and challenges in studyin parentalcontrol: toward a new conceptualization. Child Dev Perspect. 2009;3:165–70.32. Holt-Lundstad J, Uchino BN. Social support and health. In: Glanz K, Rimer BK,Viswanath K, editors. Health behavior: Theory, research, and practice. 5thedition edition. San Francisco, CA: Jonh Wiley & Sons, Inc; 2015:183–185.33. Deci EL, Ryan RM. The "what" and "why" of goal pursuits: human needs andthe self-determination of behavior. Psychol Inq. 2012;11:227–68.34. Baranowski T, Perry CL, Parcel GS. How individuals, environments, and healthbehavior interact: social cognitive theory. In: Glanz K, Rimer BK, Lewis FM, editors.Health behavior and health education. 3rd ed. San Francisco, CA: Jossey-Bass;2002. p. 165–84.35. Bronfenbrenner U, Evans GW. Developmental science in the 21st century:emerging questions, theoretical models, research designs and empiricalfindings. Soc Dev. 2000;9:115–25.36. Revicki DA, Cella DF. Health status assessment for the twenty-first century:item response theory, item banking and computer adaptive testing.Qual Life Res. 1997;6:595–600.Mâsse et al. BMC Public Health  (2017) 17:574 Page 11 of 11


Citation Scheme:


Citations by CSL (citeproc-js)

Usage Statistics



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"
                            async >
IIIF logo Our image viewer uses the IIIF 2.0 standard. To load this item in other compatible viewers, use this url:


Related Items