Validation of the James-Lowell racquetball test : a skills and ability test for use in instructional racquetball.

Abstract

The main purpose of the study was to validate the James-Lowell Racquetball Test (JLRBT), Form A: Knowledge, and Form B: Ability. A secondary purpose was to establish reliability and construct norms for the JLRBT, Form B.

Validation of Form A was accomplished by an item analysis utilizing the Flanagan Method as described by Scott and French. Validation of Form B occurred by correlating the instructors' subjective rating of each subject's overall racquetball ability and the initial test scores of the subjects on the JLRBT, Form B. Reliability was established by correlating the initial test scores and the retest scores of the subjects on the JLRBT, Form B. Norms were established utilizing the Sigma Scale. All correlation utilized the Pearson product-moment method of correlation.

The population for the study was taken from eleven racquetball activity classes at Middle Tennessee State University in the fall, 1980, semester. A total of 180 male and female students, aged 17 to 48 years, were administered the JLRBT, Form A, a fifteen-item racquetball knowledge test, prior to Form B, an ability test consisting of three consecutive thirty-second front wall rallies. Form B utilized a ten by twelve foot restraining rectangle and a line on the front wall five feet above the floor.

Findings. According to the standards set by Scott and French, the item analysis resulted in thirteen items (86.6%) being acceptable items for racquetball knowledge. Two items (13.4%) required revising.

Seven instructors teaching eleven racquetball classes subjectively rated their individual students' overall racquetball ability. The following rating scale was used: 5--Excellent, 4--Good, 3--Average, 2--Poor, 1--Very Poor. The subjective rating of the students occurred during the administration of Form B; however the instructors were unaware of the subjects' final test scores. Validity coefficients were obtained between the instructors' subjective ratings and the initial test scores. Coefficients for each coeducational class ranged from a low of r = .31 to a high of r = .82. When all instructor ratings were correlated with all subjects' initial test scores, r = .59 was produced. The initial instructor ratings obtained a positive correlation of r = .50 with the initial test scores of male subjects and r = .495 with the initial test scores of female subjects.

The test-retest method was used to establish reliability for Form B. The subjects consisted of 165 of the total population. When initial test scores of all 165 subjects were correlated with retest scores, r = .81 was produced. The initial test scores of male subjects obtained a positive correlation of r = .77 with male retest scores. The initial test scores of female subjects obtained a positive correlation of .6596 with female retest scores.

All correlations were significant at the .01 level.

Male subjects obtained a mean score of 17.53 on the initial test of Form B, while female subjects obtained a mean score of 10.0. As a group males scored higher than females on Form B.