We looked at 232 school shooters and 232 matched controls using logistic regression to boil down a dozen "discriminating characteristics" or predictors. These authors have no comparison group and merely present case studies without a large sample, robust statistics, quasi-experimental design and threats to the construct, internal, external and statistical conclusion validity of discovering school or workplace shooters before they cause deaths, injured victims and cost (increase taxes, more liability insurance premiums, lost health and mental health). Watch for my upcoming publication to see what are concrete changes the US and world needs to do to predict and prevent these horrific events.
