Fig. 4From: Identification and evaluation of risk of generalizability biases in pilot versus efficacy/effectiveness trials: a systematic review and meta-analysisAssociation of the three most prevalent risk of generalizability biases with pilot and efficacy/effectiveness sample size. Note: The x- and y-axis represent the log of the total sample size per study. The tick marks represent the actual total sample size across the range of sample sizes in the studies.Back to article page