The sample sizes of studies using secondary data were significantly larger than those using
primary and mixed data sources. An analysis of variance analysis conducted across the three
independent groups of primary, secondary, and mixed data found significant differences in sample
size between the groups (F = 6.02, p < 0.005). Applying post hoc Tukey tests discovered significant differences between secondary and primary data sources. On average, primary data users utilized 175 cases per study whereas secondary data averaged 1250 cases. No significant difference was found between mixed and secondary data even though mixed data users averaged only 182 cases per study. The result was attributed to the low number of mixed data cases resulting in low statistical power. Suspecting that the result may have been affected by skewness created by a few very large samples in the data set, we repeated the analysis using a nonparametric Wilcoxon rank sum test but also found significant differences in sample size between primary and secondary data sources(chi-square = 47.06, p < 0.001).