Figure 3. Forest Plot Depicting the Results of a Random-Effects Meta-analysis.
Results are reported as Cohen κ agreement between prospective and retrospective measures of childhood maltreatment. When studies reported multiple effect sizes for different maltreatment types, the mean of the κs was calculated to compute 1 overall effect size per study (κ = 0.19; 95% CI, 0.14-0.24; P < .001; I2 = 93%). Diamond marker indicates overall effect size and its variation; different sizes of markers are a function of the standard error for κs in individual studies in the random-effects model.