It's not clear to me why this study would set those thresholds for the underlying data (4+ Victims & 2+ deaths), this seems to downplay the number of incidents. At the very least there should be some explanation of why this limit is in place, as it has a significant impact on the findings.
Closest thing I can find to an accurate data-set is here: https://www.kaggle.com/ecodan/us-school-shootings-dataset/no.... Data seems to be a blend of a northwestern study and the wikipedia list of shootings.