An evaluation of the EASY instrument in a cross-sectional study

Abstract

Background: The purpose of this paper is to evaluate the impact of modifying the published scoring system to address identified potential weaknesses in the published scoring system for the Evaluation of Activity Surveys in Youth (EASY). A secondary purpose was to evaluate the EASY on children in Grades 1-5. The EASY is a self-report physical activity instrument for youth.

Methods: Original EASY survey results were collected at one time point from an online panel from participants across the United States as part of a larger cross-sectional University of Minnesota project looking at children's specific activity and sports participation between June and August 2019. Data was evaluated using three common scoring methods: simple summation, mean, and transformed summation. Data was compared by Grades 1-5 and 6-8.

Results: The summary statistics of the scores show that there is no statistically significant difference across the scoring methods by population. A paired t-test evaluation of the different scoring methods shows that while the scores are very similar within methodology (simple summation, mean, transformed sum) they are all statistically significantly different from one another, which demonstrates that for any given individual the specific scoring methodology used can result in meaningful differences. The transformed sum provided the strongest methodologic result. Analysis also concluded that administering the scale by proxy to children from grades 1-5 resulted in similar responses to those in Grades 6-8 broadening the appropriate populations able to use this scale.

Conclusion: The transformed sum is the preferred scoring method.

Trial registration: Not applicable.

Keywords: EASY; Evaluation of activity surveys in youth; Physical activity; Scoring methodology; Survey; Youth.

Fig. 1

(a) (simple sum) (b) (mean scoring) (c) (transformed scoring): Means for different scoring methods for the EASY instrument using box and whisker plots. The bars represent the upper and lower confidence limits with the dots representing the outliers for each analysis. The line in the box is the median and the diamond is the mean with the line connecting the 3 box plots demonstrating the change by condition

Fig. 2

Graphic compares the recommend scoring with the transformed sum; while many cases fall on the regression line (Rsq 0.98), there is a distinct group that deviates and have a higher activity level when the transformed sum is used and a large number in which the scoring is slightly lower

Fig. 3

Graphic demonstrates a similar pattern (school items capped at 5, conditional scoring) in which there is a distinct group with higher scores, but no scores which are substantially lower

Fig. 4

Graphic is the mean scoring method and it shows a different pattern in which low scores are fairly uniform, but as scores increase the number of cases diverging above and below increases as well