Is class stereotype bias accurate? I don't know. But if so then the optimal amount of actual bias may be nonzero. So does the study show the right amount of bia

Is class stereotype bias accurate? I don't know. But if so then the optimal amount of actual bias may be nonzero. So does the study show the right amount of bias, too little, or too much? It wouldn't be easy to measure the correlation between classes and employee performance, but not impossible either. Lacking such data it's hard to know how to interpret such results. Of course any such bias should be deeply discounted by any actual personal data you can gather. Any group-based bias should be considered trivial in comparison. But this study is designed to only consider class signals.