Software Reviews : NCSS: Power Analysis and Sample Size (PASS)Publisher: Jerry L. Hintze,329 N. 1000 East,Kaysville,UT 84037;8oi- 546-0445;FAX [phone] Year of Publication: 1991 Version Reviewed: i.o Materials: 248-page manual,4 5.25" disks Price: $149 Machine Specificity: MS-DOS computers System Requirements: 512K RAM,2 floppy drives or a hard disk with 2 to 5MB free;graphics printer and/or plotter highly recommended Effectiveness: Excellent User-Friendliness: Good Documentation: Good

Abstract

PASS is the Power Analysis and Sample Size program. It analyzes t-tests (one mean; two means), F-tests of several means, chi-square tests (goodness of fit; contingency table), tests of proportions (one; two), log-rank survival tests, correlations (single; two correlations), ANOVA (fixed effects; randomized block; repeated measures), log-rank tests, regression (multiple; logistic), and bioequiva- lence procedures. It then produces AScii and graphics files reporting power levels. It can also be used for determination of needed sample size to achieve a given power level with a given procedure. Rank- and-file social scientists who merely want to estimate sample size during the design process will find PASS too sophisticated and non- user-friendly for their needs. Methodologists who understand power analysis, however, will find PAss an extremely useful and effective tool.