Efficiently merging fatigue datasets from diverse sources has proven to be a strategic approach for enhancing the reliability of fatigue assessment and design within industry, while concurrently streamlining costs and time. Statistical parametric analysis is an approach that can be applied to fatigue datasets to determine whether the datasets can be deemed statistically significant (different) or statistically insignificant (similar). This paper systematically employed statistical parametric test-statistic hypotheses to assess significance. To validate this approach the paper used as a case study, fatigue data sets generated from varied notched specimens with hole diameters ranging from 0 mm to 3 mm, in addition to data from the literature. In particular, gross stresses were utilized to ensure that the only means to identify differences in the fatigue datasets was through statistical analysis. This approach was observed to work well for geometries with differences in notch geometry as small as 1 mm and was able to identify notch insensitivity in cast iron. Thus, this method can be used to differentiate fatigue datasets based on statistical parameters rather than other physical parameters.