The formula used to calculate the fail-safe number is k0 ¼ kð d dcÞ=dc , in which k0 is the number of studies with null results needed to render the average effect as negligible, k the number of studies in the synthesis, d the overall average effect size, and dc is the effect size at a negligible level, where in this analysis it was set at 0.01 [Hedges and Olkin, 1985]. DSTAT and MetaWin do not include the procedure for calculating the planned comparison statistics, therefore quality-weighted effect size, and so an Excel program developed by the first author was used based on formulas provided by Shadish and Haddock [1994]. Similarly, an Excel program developed by the first author was used to calculate the planned comparison statistics using formulas presented by Rosenthal et al. [2000]. Weighting effect sizes for quality of research studies and their reports To address the variability in the quality of studies and their reports, they were assigned quality weightings. These weightings form an index of the confidence that can reasonably be placed in the findings from each study and from the overall synthesis [Orwin, 1994; Wortman, 1994].