I'm not talking about gross prejudgment of individuals based on group observations, and in fact I would say that in my organization we have gone above and beyon

I'm not talking about gross prejudgment of individuals based on group observations, and in fact I would say that in my organization we have gone above and beyond to counter these traps. We've given opportunities to people in excess of what you might on average, hence I have actual observations about people who did and did not succeed from such less privileged backgrounds. All my observations are not talking about petty things like table manners, "in" jokes, or what -- they're about deep personality and motivational traits that affect the quality of work or reliability of the person. Tested after giving people fair chances over multiple exposures. You are correct that someone getting to a level of achievement from a disadvantaged starting point is quite an accomplishment and worthy of consideration. What I was talking about was countering the notion that someone's background has no predictive power or should be ignored. It has extremely strong predictive power, but should not be used to exclude other indicators of potential. The key is to not enable yourself to think you have examined all the important indicators with the few that are "easy" or "traditional". Finally, unfortunately I have to disagree that "diversity" in itself is a desirable goal. This is because I think that your "diversity" likely means appearance diversity or class diversity, for political goals, rather than actual valuing different opinions and approaches. I have not found that symbolic "diversity" leads to that. In fact, using symbolic "diversity" depletes your resources to find people of actual and valuable diversity. True diversity in people's thinking, areas of capability, and approaches professional work and interaction are more valuable and are the things I personally look for. And I have not found it more likely to find that in a group of people of random race, gender, class, etc. symbolic indicators, than a group from a more homogeneous source.