There are modern studies that show the same thing - intelligence test score means still differ by race, and intelligence tests are the best predictor of job performance that we know of, substantially better than knowledge tests or interviews. It creates a bit of a legal problem, considering Title VII. The best predictor of job performance that we know of usually shows mean racial differences.

Personality test score means don't differ by race however (NEO included), which is part of what makes them so attractive to HR.

Well, many people suggest taking anyone that appears to be cheating and either 1) not hire them, or 2) put them through a more rigorous interview process. If you can hire the non-cheaters for $10 per applicant, and then do a more rigorous interview (at whatever hourly rate the hiring manager earns) for those that might be cheating, the company is ultimately saving money on their hiring system, which is the purpose of using these tests in the first place.

And on that note, I would never suggest using personality tests in small businesses - the correlation between personality test scores and job performance is just too low for it to be justifiable unless you are introducing it into an environment in which hundreds of potential employees will be taking it (in which case, it should still be part of a multi-stage system).

That sounds like a company correctly testing a personality measure that they are considering adapting in the future, even if the test itself wasn't very good. This is one of the ways to ensure a test will be legally defensible - have every new employee take it, but don't hire them based on it. Then see if the personality test scores correlate with their job performance (usually as measured by supervisors) at some time later.

There is no spiritual ancestry in modern personality testing with the MBTI - it lacks the psychometric properties required of tests these days (reliability and validity). It is still used because the creators still want to make money off of it. Few industrial psychologists with any decent statistical training would be caught dead using it.
 
An example of a modern personality test that is currently used (and has been successfully legally defended) is the NEO-PI-R. Scores on several scales in this measure have been demonstrated to correlate with job performance across a variety of jobs.

This is actually hotly debated amongst industrial psychologists.
 
Vendors of personality tests include items that "detect" patterns of responses that appear to be due to this kind of cheating. They then look at these cheaters (the ones who are purposefully answer how a "good employee" would answer instead of with their own tendencies) and check their level of job performance. Oddly enough, there is a correlation - people who pad their responses to look like a "good employee" also tend to have higher job performance ratings, at least as it appears to their supervisors.

There are actually several validation techniques that you can use to do just that. Personality actually has a moderate but fairly consistent relationship with job performance across most job types.

The problem is that the shinyness surrounding personality in the business world these days makes a lot of organizations think that they can just write some questions about the kind of people they like, throw that into an online test, and hope for the best. This does not work, and is not usually legally defensible.

Also - discrimination is legal. Hiring someone with more work experience is discriminatory in nature. Discriminating against a protected class is not. But generally speaking, as long as the results of the personality test do not correlate with membership in a protected group (race, color, religion, sex, or national origin - see Title VII), or predict job performance differently between members of the various classes, then that is not a concern.

Experience always trumps education.

Grades are important, but only while you are competing against other recent college graduates. If a company is hiring a new IT person and has 10 recent graduates to look between, the one with the highest grades will be an easy call for an interview.

But that isn't the situation now.

Right now, we have laid off IT workers who have already had a job, sometimes years of them, and that experience (and demonstrated success at holding a job for a while) is more valuable than your schooling, and a 0.5 difference in GPA.

Someone liked them long enough to let them keep an IT job for some number of years. You, however, are an unknown factor. Thus, they are the safer bet.

They have already proven they can stick to a college degree long enough to get it (as have you). They have also proven they can be successful in a real IT environment. Thus, they are 2 for 2. You are 1 for 2.

Just get any IT job you can find, at least for now. Trade up when options are better. Don't hold out for your dream job now, or you might not get anything at all.