First thing I'll say is that I'm not an advocate for how industry often interviews today, whiteboard and puzzle problems and all, and their efficacy is worthy o

First thing I'll say is that I'm not an advocate for how industry often interviews today, whiteboard and puzzle problems and all, and their efficacy is worthy of discussion. It's a hard, hard problem and there are major deficiencies. But this academic "study" seems highly suspect. 48 subjects and we can make a generalization? People's stress reactions seem intimately tied to enormous number of contextual factors. Looking at the abstract it seems the "goal" of this study is "to make problem-solving assessment more equitable and inclusive". A noble and worthy goal but this does not sound like academics or science. It sounds like its conclusion was sought for.