Related Experiment Video
Updated: Jun 8, 2026
05:59
Visualization of Intensity Levels to Reduce the Gap Between Self-Reported and Directly Measured Physical Activity
Published on: March 7, 2019
Developing measures on the perceptions of the built environment for physical activity: a confirmatory analysis
Jennifer L Gay1, Kelly R Evenson, Jessalyn Smith
1Division of Behavioral Sciences and Health Promotion, University of Texas School of Public Health, Brownsville, TX, USA. jennifer.l.gay@uth.tmc.edu.
Summary
This study validated a three-factor model of the built environment for physical activity, finding strong reliability and measurement invariance across diverse groups. These findings support using these constructs in broader research and settings.
Area of Science:
- Environmental psychology
- Public health research
- Measurement in social sciences
Background:
- Limited validity evidence for built environment scales hinders physical activity research.
- Existing scales require rigorous psychometric evaluation for diverse populations.
Purpose of the Study:
- To assess the test-retest reliability and measurement invariance of a three-factor built environment model.
- To evaluate the model across gender, race, geographic location, and physical activity levels.
Main Methods:
- A random sample of 1,534 adults completed a survey on neighborhood perceptions for physical activity.
- Test-retest reliability was assessed in 106 participants over two weeks.
- Fit indices, Cronbach's alpha, Mokken H, Spearman correlation coefficients (SCC), and intraclass correlation coefficients (ICC) were used.
Main Results:
- Strong construct test-retest reliability (ICC 0.90-0.93) was observed.
- Measurement invariance was supported across gender, race, geographic location, and physical activity levels.
- While configural invariance was strong, weak support for variance and covariance invariance was noted.
Conclusions:
- The three-factor built environment model demonstrates reliability and validity across diverse groups.
- This supports the use of these constructs in latent variable modeling and other settings.
- Findings suggest broader applicability of built environment assessments for physical activity research.
