Criterion-referenced agreement of the FITNESSGRAM upper-body test of muscular strength and endurance.

Abstract

The purpose of the study was to investigate the percent agreement between the FITNESSGRAM push-up test (PSU) and the FITNESSGRAM alternate tests of upper-body strength and endurance. Further, the upper-body strength performances were compared across age groups and genders using survival analysis techniques.

Four hundred and three children, in grades three through six, from a local elementary school were recruited for the study. On the first day of data collection the children's height and weight were measured and the modified pull-up test (MPU) and the flexed-arm hang test (FAH) were administered. On the second day, children were administered the pull-up test (PU) and the push-up test.

The percent agreement indices for eight and eleven-year-old boys were moderate to high (.61 to .86). The PSU-MPU and PSU-FAH percent agreement indices were higher than the PSU-PU percent agreement index for eight, nine, and eleven-year-old boys. The kappa and modified kappa statistics for all three comparisons indicated a slight to substantial agreement (.28--.70 and .22--.72 respectively).

Eight to eleven-year-old girls yielded higher percent agreement indices for the PSU-PU comparisons (.67 to .82) than the PSU-MPU and PSU-FAH (.48 to .75) comparisons. The kappa and modified kappa statistics ranged from .09 to .55 and -.04 to .64 indicating a poor to moderate agreement indices than eight to eleven-year-old girls on all three comparisons. The same was indicated for kappa and modified kappa.

As for the survival analyses, the four tests of upper-body strength and endurance did not statistically differentiate (p > .05) the strength differences that are typically seen in boys and girls from age group to subsequent age group. On the other hand, strength and endurance levels between boys and girls were statistically different (p less than .05), with the boys' strength and endurance levels being higher. Those strength differences only held up for the push-up test across all age groups.

In conclusion, based on the large number of poor to moderate agreement indices, using the FITNESSGRAM alternative tests of upper-body strength and endurance will result in different healthy/unhealthy classifications for a high percentage of children, especially girls. Further, a longitudinal study needs to be conducted to compare survival curves across time to assess changes in children's muscular fitness performances.