Two University of Georgia researchers studied the effect of social reinforcement on the duration of exercise in adolescents with moderate mental retardation (Clinical Kinesiology, Spring 1995). Eleven adolescents with IQs ranging from 32 to 61 were divided into two groups. All participated in a six-week exercise program. Group A (4 subjects) received verbal and social reinforcement during the program, while Group B (7 subjects) received verbal and social reinforcement and kept a self-record of individual performances. The researchers theorized that Group B subjects would exercise for longer periods than Group A. Upon completion of the exercise program, all 11 subjects participated in a run/walk "race" in which the goal was to complete as many laps as possible during a 15-minute period. The number of laps completed (to the nearest quarter lap) was used as a measure of the duration of exercise.

a. Specify the null and alternative hypotheses for a non parametric analysis of the data.

b. The Kruskal-Wallis H-test was applied to the data. The researchers reported the test statistic as H = 5.1429 and the observed significance level of the test as p-value = .0233. Interpret these results.

c. Are the assumptions for the test carried out in part b satisfied? If not, propose an alternative nonparametric method for the analysis.