A Fortran Extended program which computes Goodman and Kruskal's Tau-b, its asymmetrical counterpart, Tau-a, and three sets of confidence limits for each coefficient under full multinomial and proportional stratified sampling is presented. A correction of an error in the calculation of the large sample standard error of Tau-b is discussed.…

A large class of rank tests which includes the "sign test" and Wilcoxon signed-ranks test is described. These distribution free tests provide a basis for testing various research hypotheses. Exact small sample and approximate large sample procedures are considered. Applications are presented including numerical examples. (Author/HG)

In traditional null hypothesis testing, researchers use critical values of various test statistics in order to minimize the risk of making Type I errors. These critical values are associated with common alpha levels (e.g., 0.01, 0.05) that indicate the probability of a Type I error. Alpha values are set at conservative levels such that the Type I…

An examination of statistical weaknesses and test limitations of performance contracting. (Author)

College Ss were twice administered both on IQ test and a test opinion scale in counterbalanced orders. The Ss received differential feedback before the second administration. Feedback significantly affected subsequent retest IQ test scores and opinion scale scores. It was concluded that when the testing is viewed as a complex interaction of…