Fig. 2

A funnel plot assessing publication bias for studies regarding the prevalence of sub-clinical mastitis in cattle in Africa (from 2010 through 2020). Number of studies included = 82. The x-axis is a measure prevalence estimates (double arcsine transformed). The y-axis is the precision of the study size (standard error) of corresponding study. The vertical line is situated at the transformed value of the summarised prevalence on the funnel plot. Circles that represent smaller studies are broadly spread towards the bottom (less precision; higher standard error), and further from the centre of the funnel plot (less similar to summarised prevalence), whereas circles from larger studies are narrowly distributed towards the upper part of the graph, and symmetrically clustered around the vertical line. The two limit lines symbolise the 95% CI around the summary prevalence value. More circles lie beyond the two limit lines, indicating high heterogeneity. Note: funnel plot as a measure of publication bias needs to be interpreted with caution because sometimes studies with undesirable results are not published due to other factors