Reliability and Validity for Measures of Children’s Self- Efficacy for Walking to School David A. Rowe FACSM, Shemane Murtagh, David McMinn, Katherine L. Ord & Norah M. Nelson Physical Activity for Health Research Group, University of Strathclyde, Glasgow, UK Background  Many children do not meet current physical activity recommendations  Active commuting (e.g., walking, cycling) has been identified as a key target behavior for increasing physical activity in children  Self efficacy (SE) is a mediator of physical activity behaviors in children and accounts for the effect of intention on physical activity (Motl et al., 2002)  Need to measure self-efficacy within an active commuting intervention (Travelling Green) in order to investigate theoretical mechanisms for behavior change. Methods Data analyses Summary/Discussion  Participants: Children involved in an active commuting intervention (Travelling Green; McKee et al., 2007) and one of their parents  Children in Year 5 of primary (elementary) school (N = 165, 57% boys, 43% girls; age 8-9 yr)  Parents (N = 115)  Measures  Self-efficacy scales were part of a larger questionnaire developed using the conceptual framework of Panter et al (2008), to investigate determinants of active commuting  Developed specifically for the Travelling Green evaluation study, using methods described by Motl (2000) and Feltz (2008) for developing measures of SE  Self efficacy for overcoming barriers associated with active travel among children  14 question items (How sure are you that...?) relating to general barriers (e.g.,... you can walk to school?), social environment (e.g.,... you can ask your friend to walk to school with you?), weather (e.g.,... you can walk to school in bad weather?), traffic safety (e.g.,... you can cross difficult roads while walking to school?), personal safety (e.g.,... you can walk to school even if you are frightened of meeting strangers?), and planning (e.g.,... you can walk to school even if it takes a long time?).  3-point Likert response (Not sure, Kind of Sure and Very Sure).  Parents responded regarding their perceptions of child’s efficacy for overcoming the same 14 barriers (How confident are you that your child can...?)  5-point response format (Very Confident, Quite Confident, Somewhat Confident, Not Particularly Confident, Not at all Confident)  Procedures:  Child questionnaires completed on a Monday morning in small groups of 4 to 8, supervised by a team of researchers  Parent questionnaires were sent home with children with instructions for one parent to complete the questionnaire and return to school by the end of the week  To assess test-retest reliability of the C-SE, a subgroup of children (n = 23) completed the C-SE on two occasions, 7 days apart. Another sub-group (n = 72) completed the C- SE 8 weeks apart  All procedures were approved by the University of Strathclyde Ethics Committee  Informed consent obtained from parents and children  Response rate: 49% of parents and 67% of children involved in the Travelling Green intervention consented to participate in the evaluation study  Main conclusions  Reliability (internal consistency, 7-day, and 8-week test-retest) and validity (structural, convergent, and construct-related evidence) were acceptable to high for both parent and child questionnaire measures of children’s SE for walking to school  These instruments are suitable for use in investigations of active commuting in elementary school children  SE for walking to school may be better represented by a multidimensional model when measured in children rather than via a proxy (parent) measure  Recommendations  Questionnaire measures with children of this age should be administered in small groups  Future validity research will look at the role of SE and other psychological determinants in active commuting behavior and behavior change  Further research is needed into the dimensionality and invariance across child and parent populations for SE for active commuting in children i  Internal consistency:  Cronbach’s alpha (α) for total scale (14 items)  Test-retest reliability  Intraclass correlation coefficient (ICC) from 2-way ANOVA model, adjusted for a single test administration  Exploratory factor analysis  Common factor analysis extraction method, with oblique rotation  Known groups evidence  Children classified as walkers or non-walkers based upon a travel mode question in the questionnaire  Mean differences of walkers and non-walkers were compared using an independent groups t-test and Cohen’s d effect size (0.2 = small; 0.5 = moderate; > 0.80 = large; Cohen, 1998, pp )  SPSS version 16.0 (SPSS, Inc., Chicago, IL) used for all analyses Purpose  To assess the reliability and validity of children’s SE for walking to school as measured by child (C-SE) and parent (P-SE) questionnaires. References 1.Cohen, J. (1998). Statistical power analysis for the behavioral sciences (2nd ed.). Hillsdale, NJ: Lawrence Erlbaum. 2.Feltz, D., et al. (2008). Self-efficacy in sport: Research and strategies for working with athletes, teams and coaches. Champaign, IL: Human Kinetics. 3.McKee, R., et al. (2007). Promoting walking to school: Results of a quasi-experimental trial. Journal of Epidemiology and Community Health, 61, 818– Motl, R. W., et al. (2000). Factorial validity and invariance of questionnaires measuring social-cognitive determinants of physical activity among adolescent girls. Preventive Medicine, 31, Motl, R. W., et al. (2002). Examining social–cognitive determinants of intention and physical activity among black and white adolescent girls using structural equation modeling. Health Psychology, 21, 459– Panter, J.R., et al. (2008). Environmental determinants of active travel in youth: A review and framework for future research. International Journal of Behavioral Nutrition and Physical Activity, 5, 34. Acknowledgments: This study was part-funded by grants from the National Physical Activity Research Evaluation Group (provided by the Scottish Government), and Sustrans, a national charity dedicated to promoting sustainable transport MeasureNMeanSDSkewnessKurtosis Child SE Parent SE Results  Internal consistency:  Children SE α =.80  Parent SE α =.96  Test-retest reliability (children)  7-day: ICC =.86; Cohen’s d = 0.05 (p >.05)  8-week: ICC =.68; Cohen’s d = 0.14 (p >.05)  Exploratory factor analysis  First factor explained 33% of variance in children’s SE with 12 of 14 items loading >.30; three additional factors with eigenvalues marginally > 1.0 explained an additional 37% of item variance, but rotated factor loadings were mostly low on these factors  A single factor explained 66% of variance in parent SE with all items loading >.40)  Correlation between children’s SE and parent SE  Pearson r =.31 (p <.01)  Known groups evidence  SE for walkers was significantly (p <.01) and meaningfully higher than for non-walkers, as measured by the children’s scale (d = 0.60) and the parent scale (d = 1.06) T-R periodNMean (Pre)SD (Pre)Mean (Post)SD (Post) 7 days weeks Table 1. Baseline descriptive statistics Table 2. Test-retest reliability descriptive statistics