Wow, that's a great question! I know the folks at the U. of Minnesota long used the MMPI to screen applicants to their clinical program (no surprise there, I suppose), but they were thought of as unusual in this regard.

Personally, I think that some assessment of resilience, adjustment, or emotional maturity is entirely appropriate when we're considering people for careers as mental health professionals. This is especially true given the large number of people who are attracted to these professions because of their own extended experience as patients. Some of them will make excellent care providers, but they are a "high risk" group. My one personal caveat would be that I would not want a rigid algorithm to trigger automatic rejection, except in extreme cases (e.g., where you had clear indications of paranoid ideation, a propensity toward violence, etc.).

The professional issues involved would center on the quality of the assessment tools used and their validity as predictors of success in graduate school and/or clinical settings. My guess is that aside from the MMPI-2, a few of the other major personality inventories will have some evidence of validity for related indications (e.g., hiring people for sensitive jobs). The MCMI-III and 16PF come to mind. You could contact the technical support people at the publishing houses to see what they know. Another approach would be to select a couple of promising tests and administer them for several years without using the results to make decisions - keep the files separate. Then follow up: who's had problems? What scales or combinations could have predicted those problems? (This would also make a fun research project for someone.)

I am sure there are also legal issues involved, but they would stem from the basic professional issues. Were you to deny admission to a litigious person based largely on their personality test results, you would need to be able to defend your use of said test in court. Probably no more onerous than trying to expel a paranoid student in their second year of the program, but no fun.

Personality assessment inventories such as the MMPI-2 can be used in clinical settings to reveal areas of psychopathology, of course. Using an inventory like this to make educational screening decisions can therefore constitute asking applicants about their disability status. If an application were rejected because of the MMPI findings, that could mean rejecting an applicant because he or she has a disability. There could well be important legal ramifications here!

I agree with the above. Interviews combined with symbolic convergence plus the MMPI2 could be very useful I believe, but the longitudinal study would be essential.

thanks for the responses. I understand your concerns about the MMPI. However, what if I were to use the NEO PI-3? This is a measure of normal personality and does provide indices relevant to counseling (e.g., empathy, compassion, emotional stability). Provided that we did a research study examining the relations between NEO scores and outcome ratings of students on clinically relevant dimensions (e.g., supervisor ratings of performance) and finding significant relations, would the use of such a normal personality inventory be useful and avoid the above noted clinical issues?

One of my grad school professors mentioned how Scale 4 on the MMPI-2 has been used as a positive selection criterion for doctoral students, because outside of clinical populations, it can indicate unconventional thinking. "Rule breaking" was viewed as positive in the context of scientific discovery. He didn't cite a publication, but he was a credible source. That said, as I recall the practice had been since abandoned because of legal and ethical reasons. That was in the '90s though, and I entered the private sector in 2008. A lot could have happened since then.