Funnel plot to assess for publication bias. In this plot the result of the individual study is plotted on the horizontal axis (in this case the risk ratio [RR]) against a measure of the precision of the data (either the spread of the data or the size of the study) on the vertical axis (this graph uses spread of data measured by the standard error of the log of the relative risk [SE(logRR)]), the smaller the spread of data or the greater the study size the further up the vertical axis. Individual studies are represented by the small squares. From the overall result of the meta-analysis the central estimate is plotted (the vertical dashed line), and the 95% confidence intervals are drawn (the diagonal dashed lines) to form the funnel or inverted V. The assumption is that the larger the study (or the study with a smaller spread of data) then the nearer to the true result it will be, meaning the spread about the overall result will be reduced as the study size increases, hence the funnel shape. If there is publication bias, then the studies will not be equally distributed within the inverted V. The usual sign of publication bias is the absence of studies in the green box that represents where small negative studies lie.