Too Smart for the New London Police Department?

http://isteve.blogspot.com/2011/12/wheels-of-justice-grind-slow.html

Apparently the guy who applied to the New London Police Department and who was rejected for being too smart (with @125 IQ per Wonderlic when they wanted only 100-115) has lost his appeal on his discrimination case against the city.

I'm somewhat amused by this honestly.  A lot of people I know are viscerally offended by this, especially those who inhabit the 'Second Sigma' that I've discussed here before.  Even a lot of those outside that range are mortally offended by the notion that you can score TOO well on a test.  I suspect a lot of that is the indignation at the lack of transparency---obviously they won't tell you openly that you need to somewhat tank the test in advance---that fact is likely insider knowledge and the existence of insider knowledge terribly offends a lot of fair-minded people.  I'll admit that the lack of transparency is the only part that somewhat offends me.  Obviously were I a taxpayer in said jurisdiction, I'd be annoyed, because this action fairly provably will reduce the effectiveness of the local police force (reference the Bell Curve and its cited research, police officer effectiveness was specifically discussed in that tome).

But society can choose whatever figure of merit it wants to hand out its goodies, and it isn't a moral issue.  God did not inscribe on his tablets that jobs must be handed out to the top N scorers on an IQ test, despite the protestations of the Second Sigma, and sometimes the third as well.

I know the legal perspective must confuse the hell out of a lot of readers here---I mean, why is it that it is ok to discriminate against people with high IQs but apparently, per Griggs v Duke Power, not against low IQs.
The short answer is who...whom.  The long answer is this---but keep in mind it is all, at the bottom, who...whom.

You're allowed, under US law, to discriminate against any member of a non-protected class as long as there is a 'rational basis' for said discrimination.  Basically a rational basis is anything you can plausibly argue isn't 'mere animus', and the scrutiny that your supposed basis gets depends a lot on the court and whether the class in question is angling successfully to get 'protected'.  White guys with high IQs are clear at the bottom of that list.  If you're talking about protected classes, you get something called 'strict scrutiny', which is pretty much what it sounds like.
Now, if you were to, say, discriminate against people with low IQs, what would be looked at is if this had a disparate impact on protected classes---i.e., on black people.  If it fails the 4/5 rule the scrutiny will be very very strict indeed.  What does this boil down to?  Legal sophistry surrounding the central pillar of the system...who...whom.

Some commentators have suggested that by artificially restricting the range of police officers, the IQ-related performance differences for those benefiting from Affirmative action will be far less noticeable, when the time comes for, say, a detective's examination.  Devilishly clever, I must admit, it stinks of something a Third Sigma adviser to the Second Sigma cooked up.  New London seems to be generating a lot of such schemes.

But, because we value the interests of our readers, if you happen to be taking a Wonderlic IQ test for whatever reason, and you need to emulate a lower IQ than you actually possess, here's the deal.

It's typically a 50 question test administered very fast, 10 minutes is the typical time if memory serves.  Scoring is 1 point per question answered right.  100 IQ (average) is 20 questions.  A 26 is NFL quarterback standard (@112).  The basic formula is roughly (RAW_SCORE-20)*2 + 100.  So if you need to present, say, a 120 IQ roughly, shoot for 30 questions correct.  The questions are generally really easy, you just have to work them very fast.  So if you're naturally, say, a 140 IQ, which would roughly equate to a 40 Wonderlic, just slow down a bit.  Don't finish the test.  Most people don't anyway, it's akin to the ASVAB's old computational section, where they tell you that you won't finish it.  Just make sure to allow a little cushion in case you make a stupid mistake or two, but with New London's range, the target region is pretty broad.  If you need to emulate a higher IQ, I can't help you, other than to say, you should take several practice tests so you aren't freaked out by the format of the test.  It isn't a sort of examination that most people have a lot of familiarity with, so you can probably get a bit of artificial support for your score by being less naive to the test format than the average person who is tested.  This is a good idea in general for those who need to take high-stakes examinations---be at least as familiar with the test type as the average person taking the test.  Yes, such tests frequently have reliability levels up past 0.8 and 0.9, but there's often room to gain a small advantage through specific preparation.