Related Experiment Video
Updated: Jun 7, 2026
05:59
Visualization of Intensity Levels to Reduce the Gap Between Self-Reported and Directly Measured Physical Activity
Published on: March 7, 2019
Identifying group-sensitive physical activities: a differential item functioning analysis of NHANES data
1Department of Kinesiology, Boise State University, Boise, ID 83725, USA. yonggao@boisestate.edu
Medicine and Science in Sports and Exercise
|October 22, 2010
Summary
Differential item functioning (DIF) analysis identified physical activities (PA) that vary by demographic subgroups. This information can help tailor interventions for specific populations to improve physical activity engagement.
Area of Science:
- Public Health
- Exercise Science
- Biostatistics
Background:
- Physical activity (PA) is crucial for health, but participation varies significantly across demographic groups.
- Understanding these variations is key to developing effective, targeted public health interventions.
- Differential item functioning (DIF) analysis offers a statistical approach to identify measurement bias in PA questionnaires across subgroups.
Purpose of the Study:
- To identify subgroup-sensitive physical activities (PA) using differential item functioning (DIF) analysis.
- To pinpoint specific leisure-time moderate and/or vigorous PA (MVPA) items that function differently across various demographic groups.
- To provide data for designing targeted health interventions for diverse populations.
Main Methods:
- Utilized a sub-unweighted sample of 1857 participants from the 2003-2004 National Health and Nutrition Examination Survey (NHANES) PA questionnaire data.
- Applied Mantel-Haenszel, simultaneous item bias test, and ANOVA DIF methods to analyze 33 leisure-time MVPA items.
- Examined DIF across race/ethnicity, gender, education, income, and age categories.
Main Results:
- Numerous leisure-time MVPA items exhibited significant DIF, indicating subgroup-specific participation patterns.
- Significant differences were observed in participation for activities like basketball, dance, golf, hiking, soccer, and walking across racial/ethnic, gender, education, income, and age groups.
- For instance, non-Hispanic Blacks were more likely to engage in basketball and dance compared to non-Hispanic Whites, who were more likely to participate in golf and hiking.
Conclusions:
- DIF analysis effectively identifies PA items that are sensitive to specific demographic subgroups.
- This methodology provides valuable insights for creating group-specific, targeted interventions to address disparities in PA.
- The findings support the use of DIF in public health research to enhance the cultural and demographic relevance of PA assessments and interventions.
