Related Experiment Videos
Development and testing of a short physical activity recall questionnaire
Charles E Matthews1, Barbara E Ainsworth, Cara Hanby
1Division of General Internal Medicine, Department of Medicine, Vanderbilt-Ingram Cancer Center, Vanderbilt University School of Medicine, Nashville, TN 37232, USA. charles.matthews@vanderbilt.edu
Medicine and Science in Sports and Exercise
|June 11, 2005
Summary
Two short telephone activity recall questionnaires showed good reliability for assessing physical activity. While individual errors occurred, they effectively captured overall activity patterns in the study population.
Area of Science:
- Physical activity epidemiology
- Behavioral science
- Health measurement
Background:
- Accurate assessment of physical activity is crucial for public health research.
- Short questionnaires are needed for efficient data collection in large populations.
- Telephone-based recalls offer a feasible method for physical activity assessment.
Purpose of the Study:
- To develop and validate two short telephone activity recall (STAR) questionnaires.
- To compare closed-ended versus open-ended response formats for assessing moderate and vigorous physical activity.
- To evaluate the reliability and validity of these instruments against objective measures.
Main Methods:
- 104 participants completed a 3-day test-retest reliability study.
- 88 participants underwent 10-14 24-hour physical activity recalls (24PAR).
- Objective physical activity was monitored using Actigraph for at least 7 days over a 28-day period.
Main Results:
- High test-retest consistency (65-92%) and moderate kappa values (0.50-0.75) were observed.
- Correlations with moderate-intensity activity duration ranged from 0.30 to 0.40.
- Agreement with 24PAR for meeting recommendations was 60-70%, with lower kappa values for men.
- Sensitivity ranged from 50-90% and specificity from 63-84% against 24PAR.
- Low kappa values (<0.20) were found when comparing with Actigraph.
- Overall classification into meeting recommendations correlated with higher total 24PAR activity and Actigraph measures (P < 0.08).
- Open-ended format showed better performance for moderate activity, closed-ended for vigorous activity.
Conclusions:
- The STAR questionnaires demonstrate reasonable reliability and validity for assessing physical activity.
- These instruments can differentiate overall physical activity levels within a population.
- Substantial individual classification errors necessitate cautious interpretation of results.