Skip to main content
. 2012 Oct 17;3:390. doi: 10.3389/fpsyg.2012.00390

Figure 3.

Figure 3

Comparison of overall meta-analytic effect sizes between studies with quality scores above and below the median. Dark bars show the overall effect size under the assumptions of the fixed-effect model; light bars indicate assumptions were those of the random-effects model. Error bars show 95% confidence intervals.