Statistical Comprehension for Graduate Students in Physical Education: Test and Norms

Abstract

The purpose of this study was to construct for use by graduate students in physical education valid, reliable, and parallel forms of a test of statistical comprehension, and to develop norms to measure their level of understanding of the statistical techniques and methods that frequently appear in the RESEARCH QUARTERLY. Procedures utilized in constructing the final forms of the test were: (a) using members of the AAHPER Research Council to establish content specifications and to select specific concepts to be included in the test, (b) developing 180 experimental test items dealing with understanding of 60 statistical concepts, (c) administering three tryout test forms to 196 graduate students in physical education at 10 colleges or universities, (d) selecting, based on the tryout test data, and the table of specifications, 50 test items for each of two final test forms. Data used to evaluate the test forms and to establish norms were obtained from the administration of both forms of the test to 1,013 graduate students at 51 colleges and universities. For graduate students who have five or more semester hours of statistics and/or measurement courses, or more than 30 semester hours beyond the master's degree, the test has acceptable reliability (.79 to .91) and validity.