Each students raw score on each test item was converted to a performance percentile as per norms available from the Presidents Council on Physical Fitness and Sports (1992). Physical education teachers at each of the schools administered the tests during the academic year on a voluntary basis, students participated on a voluntary basis, and teachers submitted scores to the district administration on a voluntary basis. Teachers administered the tests at any time during the academic year they deemed appropriate for their academic program and curriculum.
Iowa Tests of Basic Skills The Iowa Tests of Basic Skills (ITBS) were used to measure academic achievement. This is a frequently used achievement test and has been validated by its developers (Riverside Publishing, Houghton Mifflin Company). The ITBS assessment is norm referenced and consists of multiple forms (forms A, K, L and M), subjects were administered Form M. Third and sixth graders are administered the basic skills version, while fifth and eighth graders are given the integrated writing version.
The test battery and administration are standardized. The test was administered by classroom teachers to all students in the classroom during the first three weeks of March during the academic year. Data Collection ITBS and Presidents Challenge test scores were received by the Office of Research, Evaluation and Assessment of the Seattle School District. Scores were then matched according to student number, and the student number was replaced by a code number.
Permission to use, and access to, this pre-existing data set was given to the researchers by the Office of Research, Evaluation and Assessment of the Seattle School District, and use of this data followed guidelines established by the Western Washington University human subjects review committee. Statistical Analysis For each subject, the percentile scores for each of the three academic tests (reading, language arts, and mathematics) were averaged, to give an average academic percentile score.
Similarly, the percentile scores for each of the five fitness tests were averaged, to give an average physical percentile score. The relationship between average academic and physical percentile scores was plotted, a Pearson coefficient of correlation, and the significance of the correlation were calculated (Vincent, 1999). Statistical analysis was performed using Excel 2002 (Microsoft Corp. , Redmond, Washington) and a p-value <. 05 was considered significant. Results.
The relationship between physical fitness, measured by an average score on the Presidents Challenge test battery, and academic performance, determined by the Iowa Tests of Basic Skills, in Seattle school children was examined. Data from 5847 students in grades three (n = 2049), five (n = 2169), six (n = 948) and eight (n = 681) were analyzed. The Pearson coefficient of correlation between mean fitness and academic percentile scores was 0. 19, p <. 05. A Pearson coefficient of correlation is calculated by fitting a straight line of best fit to the relationship between two variables to estimate the strength of the relationship.
In some cases the line that best describes the relationship between two variables is not a straight line, and if so using a straight line to describe the relationship results in an underestimation of the strength of the relationship. To determine if the low relationship between the variables was due, in part, to the improper use of a straight line to describe the relationship, various curved lines (logarithmic, exponential, and power) were also fitted to the data. Each of these curved lines was found to not fit the data as well as a straight trend line. Hence, the relationship was described by the Pearson coefficient of correlation.