You are making an invalid assumption. Prejudicial discrimination is bad iff you have total knowledge of all variables. Admissions panels have very little information that in many cases is weakly correlated with the outcomes they are interested in.
What you are calling prejudicial in that case they call experience. They correlate past statistical data on how the variables they do have access to predict the outcomes they want.
If that a satisfying situation? No way. Is including your experience better than just using the data you have for a current set of candidates? Hell yes. On average, hell yes. Some students will get screwed but on average IF you are smart, not a bigot, rational, have good past data, etc, your results WILL be better. Of course, most people aren't all of those things.
But what can you do other than make your best try at an optimal outcome? If they went just on raw present data, quality of students admitted to Stanford/MIT would probably go down and 99% of students would be asian. Does that seem better?
s/sloppy thinking/nuanced yet practical problem solving/g
What you are calling prejudicial in that case they call experience. They correlate past statistical data on how the variables they do have access to predict the outcomes they want.
If that a satisfying situation? No way. Is including your experience better than just using the data you have for a current set of candidates? Hell yes. On average, hell yes. Some students will get screwed but on average IF you are smart, not a bigot, rational, have good past data, etc, your results WILL be better. Of course, most people aren't all of those things.
But what can you do other than make your best try at an optimal outcome? If they went just on raw present data, quality of students admitted to Stanford/MIT would probably go down and 99% of students would be asian. Does that seem better?
s/sloppy thinking/nuanced yet practical problem solving/g