You don't care about the false negatives, you only want to avoid false positives. You might want to re-examine this assumption more critically. With a high en

You don't care about the false negatives, you only want to avoid false positives. You might want to re-examine this assumption more critically. With a high enough false negative rate, you end up burning out your candidate pool (and severely deterring those who "failed" from every applying again). Google themselves discovered this a few years ago when they did a fundamental re-evaluation of their hiring process -- having come to the conclusion that their previous filtering methods were woefully ineffective at predicting actual on-job performance. And specifically went back to their databases to try to re-recruit many of the people who had been "flushed" by the earlier incarnations of their filtering process.