Figure 4.

Funnel Plot for Publication Bias for Two-Group, Post-Intervention Analysis

Lack of publication bias would be evident if the symmetrical distribution of studies (data points) around the mean effect size (center line) was present. This plot suggests that some publication bias may be present, evidenced by missing data from the lower left corner of the funnel. Larger studies would appear at the top of the funnel, and positive studies would appear to the right of the mean effect size. Thus smaller studies with negative findings are likely missing.