5 Tests Everyone Trusts (That Are Shockingly Inaccurate)

5 The SAT

High school has two final bosses: prom night and the SAT. Of the two, the SAT is undoubtedly worse -- all it can offer you is sleepless nights before, moments of confusion and despair during, and restless hours of existential terror afterward (prom night is essentially the same, but there's also dancing).

This one test plays a huge role in determining which colleges will accept you, and therefore it significantly impacts the rest of your life afterward as well. It's basically the Hogwarts Sorting Hat for boring reality. It is also about the worst test short of the Pepsi Challenge that we could choose for determining the life paths of every teenager in the country.

The SAT as we know it today has been around since 1926 and is the progeny of an old Army IQ test. It's not just a product of a time when the world teetered between two world wars and prejudices ran rampant -- it's an offshoot of a military product of said time. Why is the Prohibition-era military dictating whether you're eligible to spin the sign outside of a Little Caesar's (sadly, about the only job a college degree guarantees these days)? There's no quick answer for that.

At the very least, the SAT is not too good at its task -- one psychologist estimates that it tests less than 20 percent of the skills necessary for scholastic success. In addition, different studies have found that the test shows a clear racial bias, with several questions subtly phrased in a way that gives white students "an edge based not on education or study skills or aptitude, but because they are most likely growing up around white people."

We knew there was something suspect about that story problem asking us how many coffees Chandler could drink in an hour if it took Gunther 20 seconds to make them and he delivered them while walking a distance of 32 feet at an average pace of 1.7 steps per second. We mean, they didn't even give us stride length, much less cool-down time. Fuck the SAT.

4 Body Mass Index

Go hit up the gym. While you're standing there in the doorway, trying to catch your breath after heroically struggling the front door open, you will be approached by an eager, lean-framed young man named Thad. He will tell you he is a personal trainer and that he is only as effective as you are motivated, and then he will bust out the body mass index (BMI for short), to point out that most of you is sharp cheddar and fried bread. The BMI is a handy measure of relative weight that can be presented in an easy-to-read chart form, and it's considered a fair and accurate way of determining whether you are a human being or nothing more than the pilot of a wobbling fat-ship.

It's also wildly outdated and hilariously inaccurate.

A 5-foot-7-inch dude is overweight if he weighs, like, 155 pounds? OK, sure. But what if that person happens to be naturally stocky, or a well-honed bodybuilder the width of a barn door with barely any fat on his body? What if he's undergone a surgery to graft unbreakable metal to his bones? Too bad, Wolverine; our handy chart here says you're technically a walking cheesesteak.

When taken as what it was intended to be -- a general set of data describing the relative fitness of an entire population -- the BMI is mostly fine. It's certainly due for a revamp, considering that it was devised by a Belgian mathematician (read: not physician) nearly 200 goddamn years ago, but it may still be helpful in certain contexts. The problem is that we do not use the BMI in that capacity. We use it like it's an individual scale, even though it allows no adjustment for individuality.

And that goes way beyond making you feel chubby in gym class. Your BMI results can cost you a hefty chunk of change. Insurance companies are constantly probing your medical history for any sign of weakness and frailty like a bunch of ambitious Klingons. Since overweight people are more of a financial risk, they charge more money to cover them ... and they use BMI as their sole source of information.

A BMI rating of 30 or more will increase your monthly premiums by over 20 percent, despite the fact that a perfectly healthy individual could be classified as obese should they pack an appreciable amount of muscle. One report found that individuals with an unfavorable BMI were charged higher premiums than smokers, who are looking at a mere 14 percent raise. Yes, according to your insurance company's logic, the Rock (wrestling-era BMI: 31.24) is in poorer health than your two-packs-a-day uncle.

3 Mammograms

All forms of cancer are a big deal, but breast cancer definitely has the best PR people, and the mammogram is widely considered the sole procedure for catching it. Which is why it's a bit terrifying to consider that there's a pretty good chance mammograms aren't very effective. Now, before you start spittin' hate and boycotting Cracked because we hate breasts (we don't; we love them more than most people love their pets), let us clarify a few things. If your options for breast cancer prevention are between "possibly ineffective mammograms" and "an optimistic attitude," go with the former. Of course. Obviously. But the test is pitched like it's infallible, and it is way more fallible than anybody is letting on.

According to a study published in January 2014, even annual mammography doesn't do much to reduce breast cancer mortality. Physical examination (the slightly less embarrassing and considerably less painful "let's fondle for lumps" process) does the job fine. This study was not performed by some quack looking for excuses to touch boobs, either -- it followed almost 90,000 women aged 40 to 59 over the course of 25 goddamn years, solely to find out whether having your boobs crushed by robot foreplay is at all beneficial. The results suggest not only that the test is pretty useless, but also that it may have some pretty nasty downsides:

"Overall, 22 percent (106/484) of screen detected invasive breast cancers were over-diagnosed, representing one over-diagnosed breast cancer for every 424 women who received mammography screening in the trial."

Considering what breast cancer means for the patient -- chemo at the least, or even full-on Amazon-style breast amputation -- that misdiagnosis percentage is kind of a big deal. See, mammograms catch cancer early by finding abnormalities in the breast. The problem is that there can be a whole lot of abnormal goings-on in a human body, and many of our little lumps and bumps are less "cancerous" and more "just kind of a weird thing your tits are doing right now." But often, these benign tumors can get treated as if it they were cancerous, because it's really hard to tell the difference. You could be exposed to more drugs and radiation than a character in a Warren Ellis comic, then discover it was all for no particular benefit.

There are up to 70,000 misdiagnosed women per year, so again, while we're not saying "do nothing and never trust science," we are absolutely saying "don't panic at a positive result," as well as "maybe get a second and third opinion before potentially losing your wonderful chest-miracles."

2 Fingerprinting

Fingerprints are far from the slam-dunk form of identification that pop culture represents them as. There are tons of worrying cases of mistaken identity caused by faulty fingerprinting, in which even the justice system has to admit that the suspects are probably not Keyser Soze-ing their way to freedom. While prints are unique and pretty easy to match when you have two perfect samples to compare, that is almost never the case.

In the real world, cops are usually trying to match a partial print, which may be smudged, distorted, and/or fossilized in a mess of blood and clown makeup. We all know a crime scene is not exactly an optimal workplace condition, and some errors are to be expected. But that makes it all the more concerning that even in accuracy tests with complete prints and ideal conditions, the error rate is still anything between 3 and 20 percent.

However, because fingerprints are ideally reliable and have been used with success for years, they continue to carry a lot of weight in courtrooms. Even if the experts know the system has flaws, that's not what's ingrained in the minds of the jury. A fingerprint match has to equate to guilt; the alternative is that Law & Order lied to us, and we simply can't take that kind of hurt from you, Detective Stabler. Not from you. So when an expert goes in front of the jury and says they have a match, it doesn't really matter how many "howevers" and "partials" and "inconclusives" they bandy about. The jury heard that the prints match, saw the stern but paternal glare of Christopher Meloni in their mind's eye, and came back with a guilty verdict.

Some have tried to fix the system, or implement new measures, but there's one problem with that: If a judge decides that fingerprints cannot be used to convict, it will leave every single conviction based on fingerprint evidence open to reinterpretation. Thousands of convicts -- many of whom are likely guilty, hardened criminals, and the rest of whom are probably justifiably pissed off about their false convictions and have had a lot of time to work out lately -- would flood the streets, and society at large would devolve into Purge-style mayhem.

And that is not a movie we'd like to live in. Maybe if it was Road Warrior, we could be convinced, but gas prices are way too high for that to be an economically viable apocalypse these days.

1 The PCL-R

The Psychopathy Checklist-Revised determines what kind of punishment a person gets after they're convicted of a crime, whether that person is granted parole, and even whether they should be executed. But at its core, it's something simpler: a test for determining if you're a psychopath.

The idea is to measure the levels of 20 different "psychopathic" personality traits (such as glibness, promiscuity, superficial charm, callousness, and need for stimulation) in a subject by asking them questions and rating the answers. If that sounds a bit more terrifyingly vague than you'd like the "should you die" test to be, well, you're not alone.

The PCL-R can be skewed by a whole host of factors, all revolving around the score keeper: how they feel about the subject, how good they are at administering the test, how they are feeling that day, pretty much anything. If you show up for the test wearing your Team Edward shirt, and the tester is rooting for Team Jacob, there's a decent chance your results are going to be somewhere in the range of "irredeemable maniac that we must expel from the face of the Earth."

But hell, this doesn't apply to you; you're not in jail. Mostly because nobody's actually caught you committing any of your many, many terrible crimes. You get a pass on the psycho test, right? Nope! The PCL-R can be used to detain people who score high enough, even if they haven't committed any crimes yet. Use of the test isn't confined to those "let's stone people to death for eating birds on 'don't eat birds' day" legal systems, either -- right now it's employed in places like Canada, Australia, and (of course) the United States.

Maybe this still doesn't concern you. Maybe you're confident that you can pass this test, no matter how unintentionally skewed the results may be. What about the times the results are intentionally skewed? One study found that testers acting on the behalf of legal prosecution routinely give higher psychopathy scores to subjects than the ones working for the defense.

Do try to calm down: Raving fearfully about the conspiratorial psycho-test is an automatic failure of the psycho-test.

Sara Ohlms spends her time playing with her dog and tweeting jokes that sounded a lot funnier in her head.

Related Reading: There are a few tests our species could really use. The Cracked forums put a few suggestions together. If you're curious about the wide and exciting world of life as a human test subject, Cracked can help too. For a look at more of the bullshit people believe about psychology, click here.