PICTURING REPEATED-MEASURES - COMMENTS ON LOFTUS, MORRISON, AND OTHERS

Loftus's (1993b, 1995) recommendation that investigators routinely inc lude plots of appropriate means along with 95% confidence intervals or some other indication of variability has considerable merit, yet we a gree with Morrison and Weaver (1995) that such plots can supplement bu t not supplant the usual reporting of analysis of variance results. Pr oviding them may be easier than Loftus and Masson (1994) indicated, es pecially when error bars are understood as supplemental descriptive de vices. We suggest a general, unified approach that applies to the expl ication of both between- and within-subjects effects. Variability is e stimated separately for each group of scores identified as different b y analysis because this serves description better. Raw scores are used for between-subjects effects, scores adjusted for between-subjects va riability for within-subjects effects. All computations and figures ar e easily effected using common spreadsheet programs.