How generalizable is the interactive relationship between outcome favorability and procedural fairness?