Skip to main content
Fig. 2 | BMC Medical Research Methodology

Fig. 2

From: Sample size requirements are not being considered in studies developing prediction models for binary outcomes: a systematic review

Fig. 2

Scatterplot of the actual number of events used to develop the prediction model against the minimum required sample size as calculated by the Riley et al. formulae. Blue triangle = studies where the events per predictor parameter was ≥ 10; red circles = studies where the events per predictor parameter was < 10. The 45-degree reference line indicates where the used sample size was equal to the minimum required sample size

Back to article page