You're right, this is a challenging study. I've probably noted elsewhere in this topic, I have first-hand experience of racist comments and attitudes from people who did not know to whom I was married. Racism is still very common - but, not universal. The presumption that it must be is a problem in and of itself.

I haven't read the original paper, and so don't know if it is methodologically sound. However, as an alternative perspective, it is worth thinking about the paper in the context of two (vaguely?) related ideas:

1. most women prefer a female obs/gynae specialist - are they "sexist" for doing so?
2. do we know whether "black" customers prefer "black" customer service staff?

Irrespective, the question remains as to whether the specific individuals in New Haven were screwed over when the city dumped the test results because they were "not the right kind." Most reasonable human beings would surely think that they were.

1. How many white males were compared to how many black males?

2. Were all white males rated above all black males?

3. Was other race groups (Asians) also present and how they were rated?

4. Were different ethnic groups among the White population compared and if any group outscored other groups? How large was this difference compared to white-black comparison?

5. Outside of the first set of experiments - was there done any comparison of service provided by a random group of white and black store clerks in the same store?

degsme:

So explain to me again how the time for Affirmative Action has passed? Particularly in light of the recent Academy of Management Study demonstrating continued existance of unearned preferences given to white males in any subjective evaluation setting.

First as I have demonstrated and as the study says it only applies to informal anonymous evaluations.

Second you have continually argued that the city was correct in discarding the test results in favor of another known testing method, namely "assessment centers" that has been shown to have less disparate impact.

However if it is true that there is a "continued existance of unearned preferences given to white males in any subjective evaluation setting", then how can an assessment center which is a more subjective evaluation have less disparate impact? The assumption based on your continued argument should have been that the advantage to white males would be more pronounced.

Either your asserted "unearned privillages" argument is grossly exagerated to completely false OR assessment centers must show more disparate impact. One of your arguments must fail.

1. How many white males were compared to how many black males?

Not clear. What threshold are you looking for?

2. Were all white males rated above all black males?

According to the report - yes.

3. Was other race groups (Asians) also present and how they were rated?

Yes. Rated lower than white males

4. Were different ethnic groups among the White population compared and if any group outscored other groups? How large was this difference compared to white-black comparison?

In the retail part of the study - irrelevant because it was scripted so ethnic alignment - which is not physiognomically determinable - could not be detected.

In the latter part of the study - the Hospital rating system - yes various ethnic groups were covered. No correlation.

5. Outside of the first set of experiments - was there done any comparison of service provided by a random group of white and black store clerks in the same store?

a) not relevant and not viable since you then lack an objective measure of performance

b) this is what the latter part of the study focusses on by looking at the data in a setting where objective performance CAN be evaluated.

Try again. Just go read the study.