We discuss several limitations of existing
practices for probing interaction effects and offer solutions that are designed to better differentiate differential susceptibility from diathesis — stress models and quantify their
corresponding implications.