What does it all mean?

Pre-employment testing has become one of the fastest-growing tools used to select employees from pools of hopeful applicants and a variety of testing tools are at the disposal of Human Resource managers and hiring managers.  There are some obvious problems with this particular strategy that jump to mind right away, such as "test faking".  Also, how can these tests account for factors such as gender differences?  If companies cannot control for these variables don't they run the risk of legal action due to discriminatory practices?  A lot of emphasis is placed on the Big Five measures of personality traits (conscientiousness, emotional stability, extroversion, agreeableness,  and openness).  While these are certainly important considerations, they don't account for proactive measures (the degree to which people take action) of an individual's personality, which makes their validity somewhat trivial depending on the position.  With regard to "test faking", few applicants are going to describe themselves as disagreeable, obstinate, self-preserving or self-centered.  The perverse outcome of "test faking" is that the applicants that enhanced (faked) their responses the most are the most likely to be hired.  The best part is that it's all based on some arbitrary test or set of guidelines in the first place!
Here's the question I have:  What is it that employers are really trying to measure?  Clearly, they're not interested in any quantifiable skills or they wouldn't be administering personality tests in the first place.  Here are two personal experiences:
1.  Out of desperation, I took a very time consuming assessment for a position that would require me to obtain a license to become a certified financial planner and I would be paid 100% commission on a draw (and self-employed, of course - no bennies!).  Not a good set of circumstances, I know.  Anyway, of course the integrity of the test was compromised because my mother and boyfriend were offering their assistance.  The three of us are intelligent people with strong analytical skills and, together, we could not pass the personality portion of the assessment because we reasoned it was better to indicate a preference for working with others.  Why would anyone suggest that they have a problem working with people?  Especially when applying for a position that involves constant human interaction?  The company reasoned that I was dependent on others for direction and that it could interfere with my ability to work independently.  The truth?  For the most part, I fucking hate working with other people.  Especially men (this would have been an office with two men), but why would I ever admit that on a test?  Ridiculous.
2.  I've applied for the same low-wage position with the same company for the third time in the last year and a half and, finally, I must be in the running because they e-mailed a personality index assessment for me to complete.  The only reason I'm willing to do this for a $10/hr job is because it is HR related and will provide me with some connections if I'm selected.  There are two parts to this assessment.  In the first section, I am to select from a list of adjectives how I feel I'm expected to act by others.  In the second section, I am to select from the same list of adjectives those that I feel really describe me.  From what I can reason, the only thing this test is going to tell them is if I act in accordance to how I feel people expect me to act.  Of course, I didn't select anything that I felt would shed me in a negative light (I'm not going to tell them that I'm frequently angry, depressed, worried, and cynical), so we're right back to test faking.  They might as well use the same methods as the company responsible for my major disappointment (see Major Disappointment Confirmed).  I mean, what is their formula?  My best guess is that they take the quadratic formula for brownies and divide it by the square root of gummy bears and that's how they decide who they're going to hire.  So, what does it all mean?