Battery Reliability of the President's Challenge

The FITNESSGRAM (Cooper Institute, 2004) and The President's Challenge (President's Council on Physical Fitness and Sport, 2003) are commonly used physical fitness assessment batteries among school-aged children. Although empirical measurement evidence exits for the FITNESSGRAM, less reliability and validity studies exist for the President's Challenge (Baumgartner, Jackson, Mahar, & Rowe, 2007). The purpose of this study, therefore, was to examine the battery reliability of the President's Challenge Physical Fitness Test Battery under two scoring conditions: 1) test items scored by administrators, and 2) test items scored by student partners. Parental assent and IRB approval were obtained prior to the study. Middle-school students (n = 265), ages 12 to 15, were administered the recommended items of the President's Challenge during the spring. Students were tested on sit-up (SU), pull-up (PU), v-sit (VSIT), and shuttle (SHUTTLE) by 8 administrators and retested one week later by the same administrators. Following a week of rest, students were tested on the one-mile run (MILE) by the same administrators and retested one week later. Students scored partners simultaneously with administrators on all items except SHUTTLE and recorded scores in a separate location. Canonical correlation analysis was used to examine battery reliability of the President's Challenge for both the administrator scores and student scores (administrator scores for SHUTTLE were included in the reliability analysis for student scores to keep the number of test items constant). Intraclass correlation coefficients (ICC₁) were also conducted on individual test-retest items. The President's Challenge demonstrated acceptable reliability based on both instructor scores (R_c1 = .95, R _{d Total} = .69) and student scores (R_c1 = .95, R _{d Total} = .73). Shared variance between administrations (R_{d Total}) was similar to findings reported for middle-school students on a precursor to the FITNESSGRAM, namely the Health-Related Physical Fitness Test (AAHPERD, 1980; Safrit & Wood, 1987) but lower than findings reported for college students on the same battery (DiNucci, McCune, & Shows, 1990). ICC values were also acceptable for instructor scores (SU = .86, PU = .95, SHUTTLE = .84, VSIT = .93, MILE =.84) and student scores (SU = .91, PU = .95, VSIT = .90, MILE =.82). The President's Challenge recommends partner assistance in test administration and evidence from this study demonstrates similar reliability evidence under administrator and student scoring conditions.
Keyword(s): measurement/evaluation, physical education PK-12, research