ANOVA/TT: Analysis of variance teaching template for lotus 1-2-3

Abstract

Inferential statistics courses often require students to conduct multi factor analysis of variance (ANOVA) problems by hand and then employ a computer to check results. Many microcomputer programs perform twofactor ANOVAs (e.g., Anderson, 1984; Brown, 1984; Butler & Eamon, 1985; Coder, 1985, 1985; Collani & Waloszek, 1983; Corrigan, Bonelli, & Borys, 1980a, 1980b; Coulombe, 1983, 1984, 1985; Galla, 1981, 1984; Hacker & Angiolillo-Bent, 1981; Lane, 1981; Strang, 1980), and each has its unique strengths. However, most programs either (1) do not handle repeated measures (within-subjects) or mixed designs, (2) require too high a level of statistical knowledge for first-time users, (3) have cumbersome input procedures, (4) fail to reveal intermediate steps in calculations, (5) give no cues for the interpretation of the results, or (6) are not user friendly. We are aware of none that conduct two-factor ANOVAs, show intermediate steps, produce summary tables, indicate significant contrasts, conduct post hoc tests, and illustrate main and simple effects on a polygon, all on a spreadsheet with instant updates. Some of the recent microand mainframe computer packages (e.g., ANOVR, SAS, SPSSX, Minitab, and BMDP) have overcome many of these shortcomings, but there still remains a need for instructional tools to teach two-factor ANOVAs. The program herein described was developed to assist students in learning to analyze and interpret experimental data from entry to illustration. ANOVA/TT is a teaching template-a set of predesigned worksheet files-etc be used in conjunction with Lotus 1-2-3 (Lotus Development Corporation, 1985). It is user friendly and requires no previous computer experience to operate. It begins with a menu-driven interactive query session that directs the user to the proper ANOVA design. Once in the proper me, program flow is nearly identical to performing an ANOVA by hand. After data are entered into the cells of a matrix, a simple keypress initiates a macro that results in an ANOVA summary table, table of means and variances, post hoc analyses, and plots of main effects and interactions. ANOVA/TT is capable of conducting ANOVAs according to the following designs: (1) repeated