Test-retest Reliability of Girl??s Perception of Environmental Factors and Transportation for Physical Activity

2274 Preliminary evidence suggests that the physical environment and access to transportation affects physical activity levels of youth. One of the first steps in studying this topic is to establish reliable instruments with the population under study. PURPOSE: The objective of this study was to examine the test-retest reliability of a survey designed for youth to assess perceptions of environmental factors and access to transportation for physical activity. METHODS: Test and retest surveys were conducted on average 11 days apart with 480 sixth and eighth grade girls in or near Tuscon, AZ, San Diego, CA, New Orleans, LA, Baltimore, MD, Minneapolis, MN, and Columbia, SC. The instrument consisted of 24 questions on safety and aesthetics of the perceived environment and access to transportation and related facilities. Additionally, girls were asked if they were aware of 14 different recreational facilities (structured and unstructured activities), and if so, would they go there and was it easy to get there. Test-retest reliability was determined using unweighted (2 level items) and weighted (>=3 level items) kappa coefficients, overall and separately by grade. RESULTS: Item-specific reliability for 8 items assessing safety of the perceived environment ranged from 0.37-0.52, with no differences by grade. Four items assessing aesthetics ranged from 0.31-0.39, and 9 items assessing access to transportation ranged from 0.34-0.55. For some of these items, reliability was lower among 6th graders compared to 8th graders. Three items assessing access to facilities ranged from 0.42-0.58, with no differences by grade. Awareness of and interest in going to the 14 recreational facilities ranged from 0.47-0.64, and reports on being able to get to those locations ranged from 0.39-0.63. CONCLUSION: Test-retest reliability of this survey to assess perceptions of environmental factors and access to transportation for physical activity indicated fair to moderate agreement for most items among 6th and 8th grade girls. Further work is needed to improve reliability of these self-report items and to examine the association of these factors with physical activity. Supported by NIH/NHLBI.