Skip to main content
Fig. 2 | Systematic Reviews

Fig. 2

From: Study results from journals with a higher impact factor are closer to “truth”: a meta-epidemiological study

Fig. 2

Relative deviation between individual study result and pooled estimate from corresponding meta-analysis depending on the (inflation-corrected) journal impact factor (JIF). The fit of a linear regression and a local polynomial regression (dotted line) together with their 95% confidence bands are drawn. Both regressions are adjusted for the weight that study results received in the meta-analysis, which is fixed at median value for plotting. The regressions are weighted with the JIF to account for heteroscedasticity

Back to article page