Reliability and Convergent Validity of the National College Health Risk Behavior Survey Physical Activity Items

Purpose The purpose of this study was to assess the reliability and convergent validity of the NCHRBS physical activity items.

Methods Forty-five college students (25 females, 20 males) completed the physical activity items twice during the same day, with approximately 2 hours between administrations. In addition, 20 college students (9 females, age: 24.1 ± 3.5 y, BMI: 27.9 ± 6.4; 11 males, age: 24.0 ± 3.2 y, SMI: 26.4 ± 3.9) wore a Computer Science and Applications Inc. Actigraph Monitor (CSA), wore a Dig walker pedometer (DVV7OI), and recorded their daily physical activity in a diary for 7 days At the end of the 7 days, subjects completed the NCHRBS physical activity items.

Results The test-retest intraclass correlation coefficients (ICC) ranged from 0.97-0.99 for the four items. The VPA item was highly correlated with VPA recorded in the activity diary (r = 0.82) and number of days with 20 mins of VPA from the GSA (r = 0.60). The MPA item was moderately correlated with diary MPA (r = 0.66), number of days with 30 mins of MPA from the CSA (r = 0.61), and number of steps (r = 0.49) The FLEX item (r = 0.57) was moderately correlated and the MSE item (r = 0.88) was highly correlated with corresponding activities in the diary.

Conclusion The NCHRBS physical activity items have excellent test-retest reliability and validity indices similar to other self-report physical activity questions.