Reliability and Validity of the FITNESSGRAM®: Quality of Teacher Collected Health Surveillance Data

Abstract

Texas enacted legislation (SB530) requiring public schools to conduct health-related physical fitness evaluations on students in grades 3-12 each year using the FITNESSGRAM® with students classified as achieving or not achieving a Healthy Fitness Zone (HFZ). PURPOSE: To investigate the reliability and validity of the FITNESSGRAM® assessments in typical school settings. Test-retest reliability of teacher administered fitness tests was evaluated and compared to expert testers using the same protocol whereas validity was evaluated by comparing teacher administered scores to the trained expert testers. METHODS: 1010 children (Boys: 439, Girls: 569, Non-reporters: 2; 58% 3rd and 5th grade, 42% ≥ 6th grade) were administered the FITNESSGRAM® with HFZ status recorded for each test. The test conditions were 492 (49%) Teacher-Teacher, 219 (22%) Expert-Expert, 202 (20%) Teacher-Expert, and 97 (10%) Trained Teacher-Expert. RESULTS: HFZ achievement results follow a similar pattern for cardiorespiratory fitness (CRF), body composition (BMI), and total of 6 HFZs achievement to those reported with the 2008 Texas statewide testing. There is a general decline in achieving the HFZ with increasing grade level and little change across grade level for achievement of the BMI HFZ. There is also a decline across grade levels for achievement of all 6 HFZs. Reliabilities are generally highest with BMI (rxx'≈.80), slightly less reliable for CRF fitness (rxx'≈.70), and lower for musculoskeletal fitness (rxx'≈.60). Reliabilities for teachers are adequate or acceptable for all fitness items. Reliabilities for expert testers are better than those for teachers. Importantly, the reliabilities are substantially better with expert testers (rxx'≈.90,.80, &.70, respectively). Logistic Regression analyses indicate that the odds of agreement between test administrations go up significantly when the repeat testing is conducted by an expert tester. CONCLUSIONS: Teachers adequately assess health-related physical fitness but results are increased when experts are involved. Sufficient time for training teachers to complete testing will likely increase the quality of data collection. This is particularly true, and important, for the PACER test, a widely used assessment of CRF. Robert Wood Johnson Foundation, Project No. 64693.