Meta-Analysis of Incremental Rehearsal Using Phi Coefficients to Compare Single-Case and Group Designs

The current study meta-analyzed single-case design (SCD) and group research regarding incremental rehearsal (IR). We used phi to meta-analyze data from 19 IR studies. Data from the SCD studies resulted in a nonoverlap of all pairs (NAP) score of 98. 9 % (95 % CI = 97. 6-100 %), which resulted in a weighted phi of .77 (95 % CI = . 69-. 83). The group research data resulted in a median d of 1. 67 (95 % CI = 1. 43-1. 91), which converted to a weighted phi of . 63 (95 % CI = . 39-. 87). Results suggested that using NAP to compute a weighted phi from SCD research led to a coefficient that was not significantly different from the weighted phi based on group studies z = . 46, p > . 05, and there was not significant heterogeneity among the effect sizes (Q (k = 17, df = 16) = 11. 73, p > . 05, I ² = 0). Efficiency data resulted in a weighted phi of -. 22 (95 % CI = -. 79 to . 34), which suggested a small effect when comparing the efficiency to other approaches, but the studies used different data to determine efficiency. Suggestions for future research are included.