Advertise here: Contact FM.

"Our system allows you to clone your best, most reliable people"
August 24, 2009 11:39 PM Subscribe

Strongly Disagree: It bothers you a long time when someone is unfair to you. Other people's feelings are their own business. When people make mistakes, you correct them. You are a fairly private person. You don't believe a lot of what people say.
Strongly Agree: Any trouble you have is your own fault. It is maddening when the court lets guilty criminals go free. When someone treats you badly, you ignore it. You agree with people more often than you argue. You are careful not to offend people. You can wait patiently for a long time. You finish your work no matter what. You know when someone is in a bad mood, even if they don't show it. Your friends and family approve of the things you do. Your moods are steady from day to day.
....the expected answers... are always "Strongly Agree" or "Strongly Disagree." You actually have four options to choose from, when asked whether a statement applies to you or not... [but] "Disagree" or "Agree" are NEVER the right answer to any question, even though any sensible person will have mixed feelings about all these questions.

Benito was notoriously good-natured. People said of him that he could have got through life without ever touching soma. The malice and bad tempers from which other people had to take holidays never afflicted him. Reality for Benito was always sunny.

posted by orthogonality (147 comments total) 22 users marked this as a favorite

Clearly I don't know as much about US law as I thought I did. Can anyone explain to me why this isn't flagrantly illegal?
posted by Pseudology at 11:47 PM on August 24

Next up, can anybody provide me with the answer key that will have Scientologists showering me with money, and not the other way around?
posted by UbuRoivas at 11:51 PM on August 24 [3 favorites has favorites]

I seem to remember failing an early version of this.

I tested positive for personality, and was therefor found unfit to stock shelves in a grocery store.
posted by louche mustachio at 11:51 PM on August 24 [8 favorites has favorites]

fairness is an abstract concept invented by humans. Us automotons are better.

sscct
posted by MrTenacious at 11:57 PM on August 24

Pseudology, why should it be illegal?
posted by magic curl at 11:59 PM on August 24

Can anyone explain to me why this isn't flagrantly illegal?

Mostly because there's no law against it.
posted by mr_roboto at 12:05 AM on August 25 [19 favorites has favorites]

cf. Manna, by Marshall Brain, coming soon to an economy near you.
posted by @troy at 12:07 AM on August 25 [1 favorite has favorites]

It's been a decade or so, but I studied industrial psychology, and we looked at personality tests as one recruitment method. From memory, they were useful, but not highly useful (well-designed structured interviews, and work sample tests/IQ tests, were much more effective in predicting performance/retention).

That particular test (given in the second-last link) doesn't look very good, in my non-professional opinion. It's too much focused around specific traits ("always friendly", "always cheerful") that people are obviously going to answer one way. Better tests are more subtle, maybe more focused on behaviours. They'd also have 'lie detector' questions built in, to check if someone was giving what they thought were the expected answer, not the honest answer (e.g. "I am always on time", "I never argue with people"). This looks like a bit of a blunt instrument, that would be easy to game. I guess if you're going to test every applicant, it needs to be. [Normally I'd expect that only short-listed applicants would be tested, though maybe it's different in the US].

Personal experience: I've taken a couple of personality tests for jobs. I found honesty was the best option - I thought about gaming one test by giving the answer I thought they wanted (risk-taking, adventurous), and it turned out that they were looking for someone stable and dependable who didn't take risks).
posted by Infinite Jest at 12:10 AM on August 25 [1 favorite has favorites]

If I decide to game a personality test in order to obtain employment at a big box retailer, will someone please do me a favor and shoot me in the face with a bazooka?

Thanks in advance.
posted by dersins at 12:11 AM on August 25 [9 favorites has favorites]

From link 5: This is a totally unofficial key, compiled off the top of my head from my limited experience and study, using random materials which have been floating around the web for years. Use it at your own risk. I got the basic info from a blogger named "Yowling Cat" on an autism/Asperger's blog called WrongPlanet.

Keep that credibility in mind before you rage at the contents of that link.
posted by Mike1024 at 12:13 AM on August 25 [1 favorite has favorites]

Now that I think about what I've heard in the US it was a bad assumption.

In Canada an employer is not allowed to ask about what you do outside of work. I don't know how the law works exactly but in 2003 a case pressed by the human rights tribunal (they'll sue on your behalf for free!) and a guy who got fired for testing positive for marijuana basically made it illegal for an employer to drug test in Canada.

If a drug test can result in a successful human rights complaint (you're not exactly suing them but it's the same thing) I don't think this would fly either.
posted by Pseudology at 12:13 AM on August 25 [1 favorite has favorites]

Wait wait wait...so I should strongly agree with the statements that make me look like a bad emploGODDAMNIT!!
posted by obiwanwasabi at 12:14 AM on August 25 [2 favorites has favorites]

I am an automaton designed to do this job according to specification and not exist outside my scheduled working hours. Strongly Agree
posted by Pope Guilty at 12:16 AM on August 25 [8 favorites has favorites]

I work with screening tests that include personality components, though they make up only about five percent of the overall test. I've seen them tweaked and massaged over the years, and they're quite useful and accurate now.

But only to a point. They'll reliably give you Myers-Briggs style information, for example, and they will detect some serious mental/emotional disorders pretty effectively (think Leon and the tortoise, only less dramatically). The most useful aspect of them, to me, is that they provide a very nice tool to compare to what a candidate believes their strengths are, vs. what the (rest of the) tests actually show they're good at. In concert, it's a very handy self-awareness score, and a heck of a lot more reliable than "So, tell me about your weak points."

IJ, test-gamers are one of the easiest types to pick out algorithmically, too, so it's actually in your best interest not to try that. At best, even if you don't get pegged as a gamer, you end up coming across as wildly inconsistent, which is a red flag in itself.
posted by rokusan at 12:17 AM on August 25 [1 favorite has favorites]

From the "unofficial answer key":

Any trouble you have is your own fault ("strongly agree")
When you are done with your work, you look for more to do ("strongly agree")
It bothers you when you have to obey a lot of rules ("strongly disagree")
Other people's feelings are their own business ("strongly disagree")

Hmm, I see. The correct choice is always the maximum douchebag option.
posted by rkent at 12:18 AM on August 25 [11 favorites has favorites]

I've read books on how to crack this stuff for years. When I applied to my last position, they asked me to fill out one of these questionnaires. First, I showed initiative by tearing the piece of paper to bits. Then, I displayed leadership by ordering the proctors to pick up the mess. Finally, I showed off my team-oriented nature by cheering them on.

I didn't get the job.
posted by Christ, what an asshole at 12:20 AM on August 25 [40 favorites has favorites]

You do not fake being polite

This is listed as "strongly agree", but I would think faking politeness an essential skill for any workplace that would use this test.
posted by little e at 12:22 AM on August 25 [2 favorites has favorites]

Pseudology, these types of test are very popular and in use in Canada and Europe, not just in the USA. I have worked with many firms, not to mention government agencies, that use them as a routine part of hiring and screening.

(Not vouching for the particular tests mentioned in the article, no idea. But in general, they are quite commonplace. I believe any possible legal issue is sidestepped because every question is a theoretical construction: "Imagine you are standing at a busy intersection...")

It's also not true, in my experience, that only "Strongly Agree" and "Strongly Disagree" are valid responses. They're certainly interesting responses, and will often be the most heavily-scored, one way or the other... but in a well-designed test, the ability to detect and deal with nuanced issues of degree is often preferred.

I know one testing battery in particular in which "Strongly Agree/Disagree" responses are the worst possible answers, since that section of the test is measuring emotional maturity and/or creative thinking.

Absolutist thinking, as from those who toss "never" and "absolutely always" around in response to problems... these people are generally less desirable, especially in managerial or higher-level roles.
posted by rokusan at 12:24 AM on August 25 [1 favorite has favorites]

From the "unofficial answer key":

Many of those are just plain untrue, at least in tests I have worked with. This guy is either very, very wrong about how the tests are scored, or he's using some pretty bad tests.
posted by rokusan at 12:25 AM on August 25 [1 favorite has favorites]

In concert, it's a very handy self-awareness score, and a heck of a lot more reliable than "So, tell me about your weak points."

The last job interview I had to do asked me to pick three words that described myself and I told them that I was told that businesses that ask questions like that usually don't have their act together and refused to answer.

For some reason I got the job.
posted by Pseudology at 12:27 AM on August 25 [5 favorites has favorites]

Pseudology, these types of test are very popular and in use in Canada and Europe, not just in the USA. I have worked with many firms, not to mention government agencies, that use them as a routine part of hiring and screening.

Well the first job that gives me one of those tests and refuses to hire me because of it I'm filing a human rights complaint against.
posted by Pseudology at 12:29 AM on August 25

err crap I mean refuses to hire me because I refuse to take it
posted by Pseudology at 12:30 AM on August 25

Strongly agree.
posted by flabdablet at 12:54 AM on August 25

little e - No, see, that right there reveals the truly odious thing going on with these tests and the soulless companies that use them. Target, Lowe's, Sears, all these vile, big-box bastards that won't even call their employees "employees" but rather "team members" or some shit like that and won't call their customers "customers" but rather "guests," as if you're gonna pour'em a fucking cup of tea - it's not enough that someone is capable of professional detachment that allows them to set their feelings aside for eight hours a day, no ma'am - to HR, the ideal employeebot is some poor son of a bitch who doesn't suppress his pride and self-determination for the duration of their shift, because they have no pride and self-determination to suppress.

They don't want folks who see working a register as a necessary evil to make the rent - they want folks who want to be on that register, who need to be on that register and have so little self-respect that they won't even notice all the insults and indignities the shithead consumers they'll meet every day shall heap upon them. It's not enough to be competent and professional - what they want is someone who has really internalized this shit.

Take a look at some of these questions and the (alleged) preferred answer -

Other people's feelings are their own business, strongly disagree
You are a fairly private person , strongly disagree
You are not afraid to tell someone off, strongly disagree
You like to be alone, strongly disagree
You would rather not get involved in other people's problem, strongly disagree
It is hard to really care about work when the job is boring, strongly disagree
When you are annoyed with something, you say so, strongly disagree

These are the ones that stand out the most to me, but the whole damn test gives me the crawlies. To me, reading these questions and answers adds up to: "Fuck your wish for privacy, retail drone. And don't you dare attempt to assert your right to basic dignity or wish for a more interesting challenge. Get out there and see if those guests need anything - don't wait for them to come to you, either - get right up in their faces and help the shit out of them. Remember the Twelve Points of Superior Target Guest Service and never forget that a SECRET SHOPPER might be out there any place at all to grade your sorry ass against them, no matter what the practical outcome of your interaction may be. Now get moving! We ain't paying you part-time nickel-above-minimum wage with no benefits to have a fucking personality or creativity of your own!"

~long breath out~

Honestly, fuck Unicru and every company that uses it. I'm glad these keys leak out online, I'm glad crafty applicants game the tests to get a fraudulent "green rating" and I hope it keeps happening, over and over, to business after business, until these creepy, invasive bullshit screening exams are absolutely useless. On the other hand, if these tests drive folks away from humiliating retail work and back out to find a proper job with some dignity, where they need not submit to a personality screening or piss in a damn cup, then maybe Unicru is useful after all.
posted by EatTheWeak at 1:04 AM on August 25 [45 favorites has favorites]

"...we do know that cheating or having someone else take the assessment for you doesn't increase your probability of success."

It is OK to tell (obvious) lies in order to defend the company line.

[] strongly disagree
[] disagree
[] agree
[] strongly agree
posted by Bokononist at 1:18 AM on August 25 [3 favorites has favorites]

About four months ago I came across one of these personality tests when I submitted an application, online, to a Very Large Company for an engineering research position. So it's not just big box retailers that use them. I failed the test. Of course, the whole time I was filling it in (not gaming, just being honest), I was thinking "Wow, this is some serious bullshit!", because really, it is. We're talking a hard science job and they ask me this nonsense?? When I finished and the site informed me that I don't meet the criteria, I was really pissed off because it just seemed so... idiotic. And I was trying to figure out what was mentioned upthread: what answers are they looking for? Do they want an automoton or are they searching for honest answers? Couldn't someone just fill in the right "sounding" answer, and wouldn't that make them a terrible employee? Then realized that it's basically telling me I'm not cut out for corporate culture, and that made me feel much better.
posted by molecicco at 1:34 AM on August 25 [4 favorites has favorites]

This is an excellent recruitment strategy if you are hiring people to take multiple choice tests!

It is also a good strategy if you want to pay big bucks to unscrupulous industrial organizational psychology consultants who are passing off their tests as psychometricly validated peer reviewed science when they are not. APA where are you? Oh right you are run by unscrupulous clinical psychologists who have their own pseudoscience with little to no outcome evaluation to hide...

If I/O psych had cracked this particularly nut it would be field shaking news. Like the Hadron Collider finding what it is looking for. It would be THE ANSWER.
posted by srboisvert at 1:53 AM on August 25 [9 favorites has favorites]

I would never write a program to steal fractional pennies like in Superman III

[X] Strongly agree
posted by benzenedream at 2:05 AM on August 25 [5 favorites has favorites]

If I/O psych had cracked this particularly nut it would be field shaking news. Like the Hadron Collider finding what it is looking for. It would be THE ANSWER.

Yeah, that's pretty much it.

The awful thing about these tests is not that they help soul sucking corporations find ~~victims~~ employees who are just overflowing with happy happy joy joy while they work their shitty jobs - in fact, if that's what they did do, they would be win-win.

The awful thing about these tests is that their results have no predictive power regarding the value of a prospective employee. The implicit assumption that personnel selection can be reduced to a scripted checklist is an insult to both the interviewees and the interviewers. It's the ISO9001 "quality control" mindset applied to people, and it's stupid, pointless and wrong.

Using personality tests as a prospective employee filter is bullshit. Making people jump through these meaningless hoops wastes everybody's time, from the hapless corporate wage slave to the HR director to the shareholder. The tests are snake oil, and they're not even used for anything that snake oil is good for.
posted by flabdablet at 2:15 AM on August 25 [20 favorites has favorites]

Years ago, my History thesis was on the rise of personnel (spell it how you will)management in the US -- part of the "what was with the Progressives" concept. Before WWI, personnel managers tried all kinds of stuff, including phrenology, to match "the man with the job". Of course, phrenology (which looked at general physiognomy, rather than just bumps on the skull) thought long, straight noses and light complexions meant a better, more intelligent worker. All of this was supposed to match people who would be happy with a particular job to that occupation. And, the big thing, this would end labor strife and all those messy unions and so on. After WWI, the managers turned to sociologists for help. That's when this kind of test came in. Psychologists administer them now but they are the same old brujo. Some day the medics will run this. Piss in a jar and we'll tell you if you can be successful here. Or unworthy of anything but living in a cardboard box.
posted by CCBC at 2:19 AM on August 25 [7 favorites has favorites]

I would never write a program to steal fractional pennies like in Superman III

... but I would do it like in Office Space^*

Gimme back my stapler!
posted by twoleftfeet at 2:20 AM on August 25

This really makes me want to learn how to write mobile applications. It'd be a fine thing - and bloody easy to publicize, considering the way things seem to be moving - to write a natural-language-based mobile app in Python that interprets these kinds of questions and spits out the correct answers based on their similarity to known questions.

And a mobile app which people could, say, have installed on their phones, so that it'd be handy during the inevitable times when they're forced to fill these idiotic things out while shuffling through job applications... well, that'd be awesome. Increasingly awesome the more popular it became.
posted by koeselitz at 2:27 AM on August 25 [3 favorites has favorites]

eHarmony could replace all of this with a much simpler test involving only 29 dimensions of compatibility.
posted by twoleftfeet at 2:33 AM on August 25 [3 favorites has favorites]

When my ex-employers were offered some free personality/aptitude tests I slagged off the idea and pointed out that the results would obviously be designed to sell training courses.
Probably as a result of this, I wasn't asked to participate, which I see as passing with flying colours.
posted by malevolent at 2:41 AM on August 25 [5 favorites has favorites]

Mike1024: From link 5: This is a totally unofficial key, compiled off the top of my head from my limited experience and study...

Keep that credibility in mind before you rage at the contents of that link.

You're being a moron. The rage isn't inspired by the answer key. It's inspired by the fact that there are such tests at all, and any intelligent and civilized person who's had to do these tests before feels it.
posted by koeselitz at 2:44 AM on August 25 [1 favorite has favorites]

You know what I'd love to see tried? A Reamweaver clone of the Kronos website, backed by a Yes Men style sales force flogging a Unicru clone that, unbeknown to the client, works by completely ignoring all survey answers and green-lighting 25% of applicants at random.

I would be very surprised to find that Unicru, as currently implemented, worked better than this.
posted by flabdablet at 2:47 AM on August 25 [5 favorites has favorites]

I dunno, I find the whole thing both stupid and shrug-worthy. I mean, their test is obviously ludicrous, but they have the right to hire whoever they want to hire. If that includes a test that selects for people that lie, well, okay, that's their prerogative. And maybe lying on your application is indeed a sign of future success with Best Buy, since presumably if you're willing to do that for the mere hint of a chance at an interview, you'll lie enthusiastically to customers when real money's on the line. Or, if you're one who actually tells the truth and still passes, maybe you actually are a good fit for a very bad job.

At its core, retail sales of that sort has been turned into an awful job, and if you've got the sense to be outraged by this test, you can't even imagine how outrageous you'd find the day to day job itself.

Just be aware of who they want working for them, and make sure you let your friends know, and things will work out okay. It's not like they're the only choice for either employment OR for electronics. If, as I expect, this really will cripple their ability to sell things over the long term, then they'll eventually fail. If it doesn't, if people will still buy from these monkeys, well -- that's how the cookie crumbles.

They have the right to control their hiring to make their business successful by their definition of success. Like it or not, personality is a core component to job function, and they have the right to select for whatever sort of personality they want. Even if they're stupid or self-defeating, they have the right to do their best to present their customers with a particular experience. It's largely a fraudulent experience, mind, filled with deception and upsell from you walk up to the minute you walk out of the store. If you're sensible enough to not like that, then shop somewhere else.

Those who want to intervene are, largely, just being authoritarian assholes, very probably worse than this test. Best Buy has the right to be stupid, even appallingly stupid. Just make sure you tell your friends how evil they are, every once in awhile, and encourage them to spread the word further. That's the right way to stop this kind of nonsense... starve them of customers.
posted by Malor at 2:50 AM on August 25 [1 favorite has favorites]

Malor: Just make sure you tell your friends how evil they are, every once in awhile, and encourage them to spread the word further. That's the right way to stop this kind of nonsense... starve them of customers.

I disagree. This has nothing to do with how they treat their customers, although that's indirectly affected. This has to do with how they treat their employees. As such, the right way to stop this kind of nonsense is to starve these companies of employees.

Universal unionization should always be the goal of any civilized society, but this is something that can happen even before that does. Just refuse to take these tests, and urge everyone you know as strongly as possible to refuse to take these tests as well. Furthermore, it's helpful, I think, to remind would-be employers that the reason you're walking out of the middle of the interview is because no decent manager would ever swallow this snake-oil and believe that there's an automated way to judge human beings.
posted by koeselitz at 2:55 AM on August 25 [5 favorites has favorites]

It's not just the big-boxes: My girlfriend turned down a job offer from a non-profit because they wanted her to take a personality test. They were almost begging her to work for them, but wouldn't waive it. Buh-bye.

The answers are pretty obvious. They want someone who is willing to be shit on by management, but won't turn a blind eye to shoplifters. Everything else is gravy. The biggest tip is that you have to choose the extreme choices, no points for half measures. They're going to be hiring a combination of feeble-minded drones and clever sociopaths.

If things get so fucked up I have to work at Best Buy, I'll remember this.
posted by Jimmy Havok at 2:57 AM on August 25 [3 favorites has favorites]

If I/O psych had cracked this particularly nut it would be field shaking news. Like the Hadron Collider finding what it is looking for. It would be THE ANSWER.

Yeah, that's pretty much it. The awful thing about these tests is that their results have no predictive power regarding the value of a prospective employee.

But they do, as I said above. They don't have much predictive validity, but they have some. From memory (~10 years ago) they can account for about 5% of the variance in performance. That's not a lot (structured interviewing would get you 30% or more; work sample tests about the same). But it is something. It has some use. Using a personality test on its own would be ridiculous, but using it to help choose between 2-3 good candidates makes sense.

I/O psychologists certainly aren't claiming to have solved the problem, but there's a good body of research that shows some use for personality tests. (Didn't have time to dig up much via Google, but this book chapter looks relevant).

As a more general comment, I'm surprised that everyone is so hostile. Like I said, I've taken these tests 3 or 4 times when applying for jobs. I didn't feel like they were an invasion of my privacy or anything. In one case the test was helpful in confirming that I was actually well suited for the job (where I considered trying to 'fake good' on questions that were obviously looking at my willingess to take risks, but decided to answer truthfully (e.g. non-risk-taking) and it turned out that they didn't want a risk-taker).

I do agree with the criticisms of this particular test, though.
posted by Infinite Jest at 3:00 AM on August 25 [1 favorite has favorites]

I call bullshit on the idea that you can "strongly agree" or "strongly disagree" with anything.

My mother taught me to agree or disagree. That's just logic.
posted by twoleftfeet at 3:10 AM on August 25 [3 favorites has favorites]

10 years from now...

"Your poop is quite consistent, but it has some clench marks along the body that are the hallmark of an uptight person. You're no match for this job. Eat more fiber."
posted by qvantamon at 3:16 AM on August 25 [5 favorites has favorites]

koeselitz: As such, the right way to stop this kind of nonsense is to starve these companies of employees.

Hmm, that's not a bad angle, but it's gonna be hard to convince the typical 20-year-old in a lot of these jobs to unionize. Plus, typically you can't join a union until AFTER you're hired, and I don't know how much power modern unions have over the hiring process anymore. And they're not going to have a lot of money for dues, which will weaken their union structure.

I'm just thinking to attack from the customer angle because we're older and, presumably, better organized. But we do it the right way, through persuasion and talking, not by legally forcing them to comply with our view of How Things Should Be.

As Infinite Jest points out, it's not all black and white; these tests actually can be useful. The problem is that Best Buy and Unicru are evil, not the tests. :)
posted by Malor at 3:16 AM on August 25

You know who else...

For their own safety no less than in the economic interests of the firm, workers had to be helped to a new subtlety of self-adaptation to the machine.
Management became vitally interested in the length of apprentices' hair - lest it jam the machinery; and in the propensity of different character-types to boredom and carelessness - so that each man and woman could be placed in the job for which they were best suited.
Personality joined labour power as a factor of production, to the demonstrable mutual advantage of both employer and worker. There were those on the employers' side who saw in this new impersonal intimacy, a possible bridge over the social chasm created by the first, more primitive machines.

Labour in the Third Reich 1933-1939 - Tim Mason
posted by Abiezer at 3:28 AM on August 25 [4 favorites has favorites]

I wonder if they are culling out agnostics.

Their focus on forcing applicants to select answers from one of two polar opposites causes me to suspect they are searching for true believers.
posted by notreally at 3:32 AM on August 25 [1 favorite has favorites]

koeselitz: Universal unionization should always be the goal of any civilized society,

I know this is a bit of a derail, but that strikes me as a rather bizarre approach. Regardless of how one feels about unions, surely the need for a union is a symptom of a deeper problem? Seems something akin to "our goal is that everyone gives to cancer charities" vs "our goal is to cure cancer".
posted by Leon at 3:49 AM on August 25 [2 favorites has favorites]

The answers are pretty obvious. They want someone who is willing to be shit on by management, but won't turn a blind eye to shoplifters.

Again, maybe this particular test (referred to in the article) was poor, or the writer horribly misunderstood it, but the tests I have seen and worked with spend ~95% of their bulk attempting to identify such things as lateral thinking ability, attention to detail, logical thinking, and even "awareness of world events".

These are all very testable things, as every SAT, LSAT and MCAT test-taker probably knows, and based on matching them to later employee reviews, they can be very accurate in terms of predicting future performance. Sometimes eerily so.

Textbook personality type testing, whatever worth you believe it has, is also easy to do and about as accurate/reproducible as anything else in the psychology universe. There are also many known and reliable tests for (some, certain) cognitive disorders. The things employers wish to test for vary based on what is important to them, of course.

But the kneejerk hatred of evaluations seems off to me. Not all employers are The Man trying to exploit workers. For example, "Honesty" and "Integrity" are two of the most common things that employers wish they could screen for. That's a good sign, I think, in terms of an employer's interests and value system. (Sadly, those are also difficult to test for accurately. There are techniques, but they're among the least reliable indicators.)

I have no experience in using such testing for retail or low-level jobs as in the article examples, though, only for creative, technical, managerial and executive positions where a bad hire is a mistake that costs hundreds of thousands of dollars. So I've never seen one that attempts to identify the sort of "passive employee" Jimmy Havok posits. (I'd be pretty shocked if one exists that treats that as a desirable trait to screen in. I suspect this is a fear-based myth.)

Then again, I'd also be surprised if low-level jobs spend six to ten hours screening applicants: it's just not cost effective. At higher level or more "important" positions, though, the tools deployed to find and measure applicants get fancier, because the difference between an adequate and exceptional hire is much bigger. If Best Buy and Wendy's are really doing this for their teenaged employees, that's awful, and is probably an example of the mis-application of half-understood technology, as if your neighborhood grocer tried to institute ISO-9001 or Six Sigma for all part-time cashiers.

From a business process standpoint, a test is just a set of formalized interview questions and techniques. These are the same things a good interviewer attempts to glean in person anyway. But instead of an interviewer or panel of interviewers battering applicants with six hours of questions, and hopefully not forgetting any important ones, a test knocks off all of the "standard" questions and gathers the key information quickly and accurately, leaving the humans to do what humans do best: evaluate the human factors. Testing also ensures that the same information is gathered from each applicant, which can improve hiring fairness and success. It can make it possible to compare apples to apples.

As for ethical concerns: I believe that jobs that hire you for your thinking ability should be able to test and screen people based on how you think. I don't see a problem with this, and an applicant who refused to be measured in that way... well, as above I'm sure they'll find another job somewhere else that doesn't care. Their same abilities, strengths, weaknesses and cognitive issues (if any) will come out later anyway, in the workplace. It just takes much longer and is more expensive to test via crucible.

I know there are at least 700 firms in NYC alone that use such testing (because I have a client that does this kind of testing, and that's about the size of their client list)... and I'm guessing that means there are many, many more that I don't know about. So to refuse to ever work for a place that screened you first... well, I'm not sure why anyone would want to reduce their employment opportunities that much. The tests I have seen and touched are very definitely not anti-employee or anti-privacy; if anything, they benefit employees by better matching applicants to ideal positions. I've seen many cases in which an applicant applying for one position is offered a very different one as a result of their testing and interviews.

Refusing a test is like refusing an interview. I mean, you can certainly skip anything you think is nobody's business, if that even exists in these tests, just as you could always decline to answer certain questions in an interview... but would you really refuse the whole interview in advance?

And now, coffee cold, I have written a novel.
posted by rokusan at 3:51 AM on August 25 [6 favorites has favorites]

I like these tests. My current employer even showed me a movie after I finished it.
posted by FuManchu at 3:52 AM on August 25 [3 favorites has favorites]

I wonder if they are culling out agnostics.

On the tests I'm familiar with, and in the handful of anecdotal cases where I happen to know... agnostics tend to perform disproportionately well, while radical followers of anything tend to do poorly.

Critical thinking skills. High overlap.
posted by rokusan at 3:53 AM on August 25

(Selfishly: I love doing these tests too. I fell into working with them by editing/testing some for a client once many years ago. I find them to be a huge amount of fun, especially the puzzle-type questions. It's like finding a stack of old Games magazines on a rainy day.)
posted by rokusan at 3:55 AM on August 25 [1 favorite has favorites]

They don't have much predictive validity, but they have some. From memory (~10 years ago) they can account for about 5% of the variance in performance.

It's higher in the tests I know, BUT (see above) these are mainly cognitive with only small emotional/psych components, and it's quite possible to view a standardized test as a "structured interview" itself.

And yes, within narrow fields they can predict later performance results very closely. An applicant bad at basic math or unable to logic out a problem won't likely gain that skill magically during their first year on the job, after all, and if it's a position that requires a lot of that... they're in trouble. Better and cheaper for all parties to know that as soon as possible.

I agree that basing a hire on a personality test alone would be near-worthless, unless you're a Scientologist.
posted by rokusan at 4:01 AM on August 25 [1 favorite has favorites]

In Canada an employer is not allowed to ask about what you do outside of work. I don't know how the law works exactly but in 2003 a case pressed by the human rights tribunal (they'll sue on your behalf for free!) and a guy who got fired for testing positive for marijuana basically made it illegal for an employer to drug test in Canada.

posted by Pseudology at 12:13 AM on August 25

Wow. Are our countries really on the same continent, right next to each other?

Clearly I don't know as much about US law as I thought I did. Can anyone explain to me why this isn't flagrantly illegal?

posted by Pseudology at 11:47 PM on August 24

Workers in the US have jack shit for rights. Drug testing has been commonplace (at least on the lower end of the pay scale) since the mid to late '80, and these psychological tests have been around for at least that long, too. (I have a hard copy that I'm not supposed to have of one from about 1993 - the assistant manager didn't know she wasn't supposed to let people take them home because they might copy them, and in fact I only asked because that's exactly what I intended to do. It's not Unicru, btw, and some of the questions are bizarre, like "How often do you blush?".)

I don't know why these tests are making the news now, but I suspect it's related to the crumbling economy. People are taking jobs below their qualification level to make ends meet, and they're seeing a side of life they hadn't seen before. When a previously privileged group of well-paid and well-treated employees learn how the poorer half live by actually becoming the poorer half, then they get outraged and a few stories like this come along.

Anyway, employee rights are hard to come by in the US. Employers can't ask you about your activities outside of work in Canada? Really? What is this human rights tribunal of which you speak? Is it paid for with public funds or private donations? If we had such a law here (which we don't) then maybe the ACLU might take up a drug screened employee's case, but given how entrenched the War on Drugs is here, I would expect them to be laughed out of court.

USA = Bizarro Canada.
posted by Marla Singer at 4:22 AM on August 25 [2 favorites has favorites]

I used to find these kinds of tests to be kind of hilarious in their clumsiness. There's tons of tests out there that are better (I mean, even the Taylor-Johnson has a great predictive feel if you're just scanning for creeps) and it wouldn't have been too difficult to hire just a couple of psychologists and a decent statistician or two to devise something better. Hell, we could get together half the people in this thread and make a more predictive test over a long weekend.

This is why this test, which versions of which I've seen lumbering about for as long as I can remember, do not have their main value as tools for management. They're so you know you are taking a test. That's the purpose. The evaluation of you has begun, and if you have half a brain, you can see what a lousy test this is. You'll be taking it anyway.

This is the first of many small, meaningless, and ultimately irrational humiliations you will undergo. You will play along by giving the "right" answers. The test is that, if you grovel once through something which is obviously a joke, you will do so again.

I love management.
posted by adipocere at 4:25 AM on August 25 [10 favorites has favorites]

Next time you go to Target, take a close look at the cashier's monitor as you're paying. You'll see a score (red or green) displayed in big letters, which looks like it's based on how fast the cashier has gotten you out of the way, as well as the scores for the last ten people to go through that checkout line. This test does not surprise me at all in an industry that turns its employees' work centers into Skinner boxes.
posted by haltingproblemsolved at 4:42 AM on August 25 [2 favorites has favorites]

With no possible 'meh' response, are these tests an attempt to keep hipsters out of the workforce? If so, they should all like, get together or whatever.
posted by ob at 4:51 AM on August 25 [1 favorite has favorites]

On the tests I'm familiar with, and in the handful of anecdotal cases where I happen to know... agnostics tend to perform disproportionately well, while radical followers of anything tend to do poorly.

OK. I will rephrase. Maybe they are trying to cull out critical thinkers.
posted by notreally at 5:08 AM on August 25 [1 favorite has favorites]

Structured interviews and computer assisted personal interviews have a lot of validity in the mental health field. To the extent that "good employee" is a meaningful construct, I'd imagine that you could easily come up with an instrument with some predictive ability. Nuclear engineers for GE used to have to take the MMPI, though I think that they were looking more for pathology than subservient-drone prediction.

Regarding how the right answer in SA/SD, I bet that they pushed the question answers (scored 1-5) into a log linear predictive model of time-until-quitting, and they use your answers to get predicted time-until-quit, so picking an extreme always maximizes the predicted value.
posted by a robot made out of meat at 5:15 AM on August 25 [1 favorite has favorites]

You'll see a score (red or green) displayed in big letters, which looks like it's based on how fast the cashier has gotten you out of the way, as well as the scores for the last ten people to go through that checkout line. This test does not surprise me at all in an industry that turns its employees' work centers into Skinner boxes.
posted by haltingproblemsolved at 7:42 AM on August 25 [+] [!]

You are correct. The sadder thing is that regardless of how fast the cashier actually scans your items, if the customer takes forever to write a check or to find their credit card, the cashier's score takes a hit.
posted by Ziggy Zaga at 5:20 AM on August 25 [1 favorite has favorites]

Could I just point out how much this paragraph disgusts me:

Although the data showed that minorities “performed significantly worse on the test,” Autor, using phrases you would expect from an economics professor, said “statistically speaking the difference was a case of standard deviation.” The study showed that African Americans were 33 percent more likely to be flagged red than white applicants. Hispanics also had a greater chance of being flagged red.

Yes. Economics (& many other) professors examine the statistical significance of the data. No. They're not being condescending, or ivory towerist. That's just how data is supposed to be examined, you anti-intellectual jackass.

33% seems really high for it to be statistically insignificant, but I'm willing to give him the benefit of the doubt.
posted by Lemurrhea at 5:26 AM on August 25 [1 favorite has favorites]

You'll see a score (red or green) displayed in big letters, which looks like it's based on how fast the cashier has gotten you out of the way.

It does seem a bit inhumane, but this is not some "whip the slaves" maltreatment of the bottom-rung employees.

At places like Wal-Mart and Best Buy, it's not just low-level employees who are performance-metricked up the wazoo. It extends to store-by-store management, who also have their daily performance tracked, quantified, and on public display... if you know how to read it. Those numbers are really reflecting the combined numbers of each salesperson and cashier.

And at a higher level, regional managers definitely spend their time looking at dashboarded numbers indicating each store's performance hour by hour and day by day, combining those numbers further.

I imagine that at the top of the chain, high in a mountain fortress somewhere, there is a wood-paneled room full of cloaked and hooded board members, all staring at a single red and green light. The light blinks and blinks and blinks. Forever.
posted by rokusan at 5:39 AM on August 25 [6 favorites has favorites]

I'd like to find out which retailers use tests like this, so I can make a point of taking my business elsewhere. I've never worked in the service industry, so maybe my expectations are unrealistic, but I'm just astounded that companies depend on bullshit tests that might have well be some sort of How To Get Your Man test in Cosmo magazine, rather than screening for professional attributes like merit and experience.
posted by crapmatic at 5:44 AM on August 25 [1 favorite has favorites]

Deckard: You're reading a magazine. You come across a full-page nude photo of a girl.
Rachael: Is this testing whether I'm a replicant or a lesbian, Mr. Deckard?
posted by paddbear at 5:47 AM on August 25 [3 favorites has favorites]

A million years ago I worked for Phar-Mor, a discount pharmacy/medium-box retailer HQ in Ohio, but dotted around the country. I worked for one in Sunrise, FL.

I showed up for the interview in a jeans skirt and high-top sneakers, rightly assuming a business suit would be inappropriate (the hardest wardrobe choice I've ever made for a job interview.) After doing all the application and testing BS, the store manager explained their typical clientele and said, "I've seen old men beating each other with canes out here over who gets to pick up a penny on the floor. Think you could deal with that?"

The cash registers were nacent computers. You didn't scan a bar-code, but you pressed one of nine category keys as you manually rang in the price of each item. They could time you based on how many items you could scan per minute. As a cashier, working for minimum wage, you had to determine not just the price of the item, but its category. That is some pretty high-functioning shit there. You had to have a nimble brain and nimble fingers.

So, they task their lowest-level employees with some serious brain usage for minimum wage. Did I mention that in the early nineties that the company had a huge scandal wherein one of the principals tried to defraud the company of millions of dollars? Oh yeah, that happened.
posted by Ruthless Bunny at 6:05 AM on August 25 [2 favorites has favorites]

Mostly because there's no law against it.

I drove past an entrance to the State Hospital (for the mentally ill) here the other day, and I swear there is a sign by the back receiving gate that says:

Contraband Prohibited.
posted by Devils Rancher at 6:10 AM on August 25 [2 favorites has favorites]

Rokusan: It's higher in the tests I know, BUT (see above) these are mainly cognitive with only small emotional/psych components, and it's quite possible to view a standardized test as a "structured interview" itself.

Yeah, I was meaning strictly personality tests; I wasn't considering the whole range of tests: you'd get better results from cognitive tests, as you say. [And of course my knowledge is a few years out of date...]

Worth mentioning here, too, that standardised testing can be a good way of decreasing discrimination. Take it to one extreme: only hire junior staff whose parents are friends of the boss - highly discriminatory. Use unstructured interviews where you just chat about whatever - likely to be discriminatory - everyone's got cognitive biases, and the interviewer is likely to hire someone similar to themselves, or someone that they like, rather than the best person. But use a highly structured, standardised process, and (as long as it has been validated) you're helping to put people on a more level playing field.
posted by Infinite Jest at 6:15 AM on August 25

They want someone who is willing to be shit on by management, but won't turn a blind eye to shoplifters.

Dealing with shoplifters is something for security or loss prevention to do. Big box retailers want most of their employees to have nothing to do with shoplifters for fear of incurring liability. Occasionally we see news stories about an ~~employee~~ team member who chased a shoplifter and was fired for it.
posted by grouse at 6:17 AM on August 25 [1 favorite has favorites]

rokusan: I work with screening tests that include personality components, though they make up only about five percent of the overall test. I've seen them tweaked and massaged over the years, and they're quite useful and accurate now.

If the tests you're talking about look anything like Unicru, then I imagine they're worthless. But I'm willing to bet they don't.
posted by koeselitz at 6:36 AM on August 25

The first comment ever on MeFi where I would understand if someone labeled it "eponysterical."
posted by nax at 6:36 AM on August 25

The problem is not these tests, it's that they're often used by people who made the phrase "A people hire A people; B people hire C people." possible (or at least instantly believed by so many).

I have no trouble believing there are people out there who want to see the "correct answers" exactly as given. The world is full of idiots who think they're smarter than they are and just need people who'll do their bidding without raising a lot of questions or pointing out reality. (Not that they're going to describe things in just that way.) An automated test just makes it easier for them to "objectively" choose C people.
posted by Kid Charlemagne at 6:42 AM on August 25

Before WWI, personnel managers tried all kinds of stuff, including phrenology, to match "the man with the job". Of course, phrenology (which looked at general physiognomy, rather than just bumps on the skull) thought long, straight noses and light complexions meant a better, more intelligent worker. All of this was supposed to match people who would be happy with a particular job to that occupation.

When I took the PLAN (pre-ACT) test in high school, there was a fairly lengthy personality test component that was supposed to determine what kinds of jobs you would be best suited for. It was more or less a list of activities, and you were supposed to rank each one based on how much you would enjoy them. They were almost all boring or weird work activities, such as "Balancing a checkbook" or "Watching for forest fires."

I didn't take the test too seriously, because at that point I had been programming for a few years and had already decided to go to college for Computer Science, but I answered the questions honestly. I figured it was a bad sign that none of the questions even mentioned computers, but I was interested to see what kind of jobs it would end up suggesting. When the results came back, the top two jobs for me were Grocery Bagger and Bellhop. I laughed it off as being wildly inaccurate, but I felt sorry for any of the kids that actually made important decisions about their future based on that test.
posted by burnmp3s at 6:42 AM on August 25 [1 favorite has favorites]

Infinite Jest: As a more general comment, I'm surprised that everyone is so hostile.

People are hostile because these sorts of tests are really a study in miniature illustrating everything that's wrong with management today. It was once a careful and subtle art, interviewing people, and it took real skill; and while you were being interviewed, you were also assessing your interviewer. All of the personality-test switches which rokusan seems to think are so brilliant and innovative are actually child's play compared to what human beings have been expected to know over hundreds of years; and there's very little that can be gleaned from a paper test that can't be gotten from looking at someone face to face, whereas there are millions of tiny physical cues worth picking up on that you just can't get from a form.

It takes actual dedication and thoughtfulness, and relaxed contemplation, to give a good interview, both for the interviewer and the interviewee. People seem to think that human contact is a painful thing that will result in loss of productivity now, but in fact it's all that working ever is, really, even for engineers and lab rats.

To put it as simply as possible: companies that use this sort of form seem to be utterly daft as to the obvious fact that they are being tested, too; the intelligent person applying for a job is watching the company, thinking about whether they want to work there. Shoving a form across a desk that's supposed to test emotional outlook and personality and saying "here, fill this out" because you don't seem to have either the time or the inclination to ask the person questions yourself screams to the applicant: we are a company that doesn't give a fuck about you, and from the start we'd rather have you fill in forms than actually talk to you.

And the looming possibility that a whole slew of us will be forced to work for such companies tends to inspire some anxiety, I think.

By the way, it occurs to me: if I were a manager in a company that asked me to use these forms, I think I might try them, but only in a specific context: I think I'd read the questions out loud to applicants and see what they said. Yes, I know I'd get different answers, but those different answers would probably be a lot more interesting, and would be more helpful, I think, to my effort to try to determine how they'll get along with their coworkers.
posted by koeselitz at 6:48 AM on August 25 [23 favorites has favorites]

What is this human rights tribunal of which you speak?

The Canadian Human Rights Commission is an arms-length Federal entity empowered by two Acts of parlaiment: one for basic human rights and one for employment equity. The equity act mainly concerns itself with four groups: women, Aboriginal people, the disabled and visible minorities, but covers all employment cases, in general. They're essentially publicly-paid advocates and pro bono lawyers for those cases, legal aid for rights and equity cases, if you will. They do some government policy and research too.
posted by bonehead at 6:48 AM on August 25 [2 favorites has favorites]

What's weird about this is that if the tests are supposed to weed out people who aren't cheerful, calm, and helpful, they're not doing a very good job.

I don't shop at big box stores very often, but the last time I was at a Target, pretty much the only employees I saw were at the register at the fitting rooms. The Walgreens down the corner from my house? Staffed, apparently, by people who are really awesome at leaving aisles full of boxes and shelves half-stocked, and cashiers who shout into the PA system when another register needs to open (and it takes the person they're shouting for ten minutes to turn up).
posted by rtha at 7:00 AM on August 25 [1 favorite has favorites]

The study showed that African Americans were 33 percent ~~more~~ less likely to be ~~flagged red~~ willing to be treated as property than white applicants.

I have no hard data, but I'd also be surprised if Jews, Soviet POWs, Poles, Homosexuals, or the diabled did well on a test designed to screen aplicants for positions as guards at a forced labor camp.

It's almost like it bothers some people for a long time when someone is unfair to them.
posted by Kid Charlemagne at 7:03 AM on August 25 [4 favorites has favorites]

rokusan: I work with screening tests that include personality components, though they make up only about five percent of the overall test. I've seen them tweaked and massaged over the years, and they're quite useful and accurate now.

But only to a point. They'll reliably give you Myers-Briggs style information, for example...

"Myers-Briggs style information" is effectively meaningless, psychologically, and is generally only preferred by companies that sell this kind of snake-oil; actual psychologists in the field have known this for decades. [ 1, 2 ]

... and they will detect some serious mental/emotional disorders pretty effectively (think Leon and the tortoise, only less dramatically).

So - they'll detect psychological disorder, but less dramatically than spoken interviews. Hm.

The most useful aspect of them, to me, is that they provide a very nice tool to compare to what a candidate believes their strengths are, vs. what the (rest of the) tests actually show they're good at. In concert, it's a very handy self-awareness score, and a heck of a lot more reliable than "So, tell me about your weak points."

Anything is more reliable than "so, tell me about your weak points." Throwing a tennis ball at applicants and asking them to catch it is a more reliable benchmark than "so, tell me about your weak points." This is not a problem to be solved by programmatic pseudo-scientific tests; it's a problem to be solved by firing the crappy managers who can't conduct an interview and hiring some who can.

IJ, test-gamers are one of the easiest types to pick out algorithmically, too, so it's actually in your best interest not to try that. At best, even if you don't get pegged as a gamer, you end up coming across as wildly inconsistent, which is a red flag in itself.

Given that the people who take these tests are almost invariably more intelligent than the people who write them, I imagine they're getting gamed a lot more than anybody realizes already.
posted by koeselitz at 7:05 AM on August 25 [5 favorites has favorites]

I applied to work at Borders when I was in high school. The lady was all set to hire me, gave me the grand tour of the building, and then found out that I failed the test and seemed very embarrased about it because that had never happened before. After that I couldn't get a retail job anywhere because once you take the test you can't take it again. /but that was fine because i started getting office jobs.
posted by amethysts at 7:07 AM on August 25

You know what I'd love to see tried? A Reamweaver clone of the Kronos website, backed by a Yes Men style sales force flogging a Unicru clone that, unbeknown to the client, works by completely ignoring all survey answers and green-lighting 25% of applicants at random.

I would be very surprised to find that Unicru, as currently implemented, worked better than this.

This. Absolutely, this.
posted by gimonca at 7:14 AM on August 25

As a more general comment, I'm surprised that everyone is so hostile.

people get hostile when their innate personality traits are essentially commodified - and that's the real problem with these tests - instead of treating your labor as a commodity, which can be bad enough, they treat YOU as one
posted by pyramid termite at 7:26 AM on August 25 [8 favorites has favorites]

While these tests obviously dehumanize the applicants, they also speak volumes about how the company perceives its managers. These tests basically say, "we don't trust you to be able to hire competent workers, either because you're too stupid to be able to tell a good egg from a bad one, or because you're too busy to be able to spend the time interviewing potential candidates." Hiring quality workers is one of the most important things a business can spend its time doing; knowing how to do effective interviews and evaluations of people takes a lot of work. These tests are a way of side-stepping that.
posted by nushustu at 7:33 AM on August 25

there was a fairly lengthy personality test component that was supposed to determine what kinds of jobs you would be best suited for.

This test was still being used (separate from the PLAN, though) at my high school last year. I received an actually helpful result, but my best friend was given Tank Driver, and another boy ended up with Window Washer. I would only imagine that there is a special result -- Big-Box Employee -- for those who successfully agree with the test-makers at every opportunity.
posted by punchdrunkhistory at 7:33 AM on August 25 [2 favorites has favorites]

amethysts said: I applied to work at Borders when I was in high school. The lady was all set to hire me, gave me the grand tour of the building, and then found out that I failed the test and seemed very embarrased about it because that had never happened before. After that I couldn't get a retail job anywhere because once you take the test you can't take it again. /but that was fine because i started getting office jobs.

What? How is it possible to not take it again, or that a test precluded you from being able to get a job in retail ever? That makes no sense, as I understand the testing system, and I'm betting money that the ACLU and private litigators would be ALL OVER a case where someone was forbidden to work in an industry because of an interview test for one potential employer. Can you give some more specifics on this; I'd love to research it. This is the sort of thing that could be used to leverage these tests out of employment screening.
posted by dejah420 at 7:36 AM on August 25

rokusan: Again, maybe this particular test (referred to in the article) was poor, or the writer horribly misunderstood it, but the tests I have seen and worked with spend ~95% of their bulk attempting to identify such things as lateral thinking ability, attention to detail, logical thinking, and even "awareness of world events".

These are all very testable things, as every SAT, LSAT and MCAT test-taker probably knows, and based on matching them to later employee reviews, they can be very accurate in terms of predicting future performance. Sometimes eerily so.

This bit here, I'm sorry to say, sounds utterly and completely vague and unscientific; I'd like to know more about what exactly you mean. (The 'sometimes eerily so' bit - you know what that is, right? It's the Forer Effect - the observed phenomenon that people swear up and down that results are eerily specific if they're told that those results are about them, regardless of how general the 'results' actually sound.) Also, I've taken both the SAT and the LSAT, and, far from being a metric of 'thinking ability' or 'attention to detail,' both were woefully arbitrary; controversially so, in some cases. As I recall, there's a group of people up in arms every year about the LSAT being an exercise in test prep rather than a measure of ability or intelligence.

Even aside from those things, just put your statement there up to the light of some comparative reasoning for a moment: if we humans had uncovered tests which somehow measure thinking ability and rationality - really, that's no small feat - don't you think psychologists would know about them? Don't you think there'd be published research, data in the field, revolutions in science and other fields? Wouldn't this, for one thing, completely change neurology and neurobiology, given that it would mean having defined exactly how intelligence works? But as far as I can tell, none of those things have happened.

This is not to say that there hasn't been any work whatsoever in the still-murky field of psychometrics; but every person I've ever met in that field who also happens to care about being careful and scientific has had an overbearing abhorrence for the unfortunate fact that psychometrics has been so completely overrated and coopted by the mainstream. This has been going on since Stanford published Binet's Intelligence Scale - a scale which Binet himself had warned had a margin of error and didn't really measure any innate kind of intelligence anyhow.

Psychometrics as such is really still only a social science; I say that in the sense that it hasn't at all been able to demonstrate any predictable evidence on the personal level but only on the social and statistical levels. (And even on those levels these test often fail; Myers-Briggs has been pegged in some studies as failing to predict a stable and unchanging type around 76% of the time.) That is, psychometric tests can tell you roughly about a given population - say, for example, that "74% of people who answer a particular range of questions with 'Agree' will tend to be good at assessing problematic situations" - but since the range of error and the lack of any ability in the real world to test further in that population means that a psychometric test can hardly give you data about a person. And the fact is that companies don't generally hire populations; they hire people, and hiring is generally a process that lasts long enough that there is time to ask the questions face to face which have a demonstrably higher chance of predicting quality of applicants.

I'd like to hear what scientific basis you can give for the tests you're talking about. My small amount of knowledge in the social sciences and psychometrics - the Social Science department at Boston College is in the same building as Political Science, and sometimes we got to talking, that's all - leads me to believe that the worth of such tests has been greatly exaggerated by companies who stand to make a profit from them, and that scientists in the field have generally found this quite vexing. Maybe I'm wrong.
posted by koeselitz at 7:39 AM on August 25 [9 favorites has favorites]

Or probably a better, more focused question (ironic, that) would be:

Just how much more effective have psychometric tests been shown to be in hiring employees when compared to traditional interview-style hiring? And what independent scientific research can you point to that shows this to be the case?
posted by koeselitz at 7:42 AM on August 25

[The tests] will detect some serious mental/emotional disorders pretty effectively

There are a lot of things wrong with employers using these tests, but this one especially rankles. Some of these things were either designed to be or incidentally become diagnostic tools -- these fuckhead laypeople are gathering private medical information about you. In addition to being grossly unethical, that's got to be potentially actionable, doesn't it? Couldn't someone mount a massive ADA challenge if denied employment?

The fascist drug-testing policy of one joint I worked at (and quit with a big long ranty protest letter the moment the policy was enacted) included a provision that all employees had to give their immediate supervisors a list of any OTC or scrip medication they took regularly. Because nothing could possibly go wrong if the person doing your performance reviews and deciding on promotions is aware that you're taking fertility meds or Lithium or an HIV cocktail.
posted by FelliniBlank at 7:43 AM on August 25 [6 favorites has favorites]

You give them your social security number before the test starts. You're taking the test by putting your choices into a touch tone phone. So you put in your SS#, and they check to see if you're in the database before you can start. It wouldn't let me take the test again.
I don't know, maybe I'm just being paranoid. Even when they did let me take the test though I've never been able to get a retail job.

But as to why people are being hostile, it's frustrating to know that:
a. I really need a job
b. This is a job i can do well at. I've worked in retail before and i'm good.
c. The interviewer likes me and would hire me
And to never ever get the chance to do it because some bureacrat decided you were a bad employee without knowing anything about you, based on something you dont even know what you did wrong on. And then to go into these stores and wonder why all these people are allowed to have jobs and you aren't while you're waiting for someone to notice you.
posted by amethysts at 7:45 AM on August 25

All of the personality-test switches which rokusan seems to think are so brilliant and innovative are actually child's play compared to what human beings have been expected to know over hundreds of years

Koeselitz, as a consultant I actually do regular old human-based interviewing for and with clients, to help them find and screen better (and better-matched) talent for their projects, and generally improve their talent-level overall. I have conducted, good lord, perhaps five thousand interviews by now? Maybe more. Generally, I get my fees because I have grown good at matchmaking: employer and employee have to 'fit' together to work well.

Now, related: one of my clients happens to be a firm that develops automated screening testing (and training systems). This could have either dovetailed or collided horribly with my work, depending on whether I chose to fight it or work with it. As it turned out, I ended up doing some work for them in helping to test and improve those tests to increase their usefulness (from my perspective, and eventually theirs too). The tests were more clueful than the samples shown in the article here to begin with, though I don't know the history. They're definitely a lot more complicated.

The biggest piece of what I worked on was developing long-process "feedback loops" (testing the value of the individual parts of the tests over time, against later employment reviews) to evaluate and most of all adjust them based on how accurate (or not) they are at predicting suitability. If particular questions, types of question, or sections don't actually map to real on-the-job performance a year or two later, they're marked as "unreliable" and redeveloped until they ARE accurate predictors.

As such, the tests I know are in use now are both accurate at what they actually try to do, and constantly evolving as part of their nature, and vary a lot between employers who have different priorities. They also look very little like those Unicru examples shown above -- those look to me more like a sort of 1960's Readers Digest parody of modern testing. And, again, they're 95% cognitive ability, with about 5% personality and mental-disorder testing.

Good screening tests can be excellent tools, even if they only improve the quality of new hires by a tiny percent, it's a worthwhile result, especially in positions where a bad hire can cost six figures to repair. But like all tools they have to be used well. It's easy to take a great tool and use it horribly. (You should see me with a soldering gun.)

I think I'd read the questions out loud to applicants and see what they said.

Well, of course that can be useful, and one gets different and more nuanced information that way, too. But that is impractical for a certain kind of question, especially when there are hundreds of applicants and/or hundreds of questions or exercises, especially when in the interest of fairness one needs to collect the same responses from each. Again, my experience isn't in Best Buy salesperson hiring, it's for positions that take weeks or months of evaluation. Not all jobs can be awarded over a quick coffee with the boss.

If you say "all testing is useless", you are suggesting the equivalent of all college admissions boards throwing away the SAT/LSAT/MCAT and basing their selections only on the student interview. Nobody would argue that position. And choosing a job (or employee) is simiilarly important to choosing a school (or student).

But all that said, you'll never hear me arguing that any such testing (whether before or during employment) can ever replace interviews. Most of the positions I can think of that use testing as part of the screening process also do multiple in-person interviews, both one-on-one and panel style. Nobody would ever think the hiring firm is giving them too little attention, that is for sure.

But I also know that there are limits to what even a talented interview panel can evaluate. And the information they collect is also tucked into the new employee's HR file as their first pre-evaluation anyway, no matter who actually writes it down. The only difference between an interviewer's file note ("I'm worried she doesn't follow directions so well.") and a test score showing the same thing is that the latter is probably more accurate and reproducible.

Shoving a form across a desk that's supposed to test emotional outlook and personality and saying "here, fill this out" because you don't seem to have either the time or the inclination to ask the person questions yourself screams to the applicant: we are a company that doesn't give a fuck about you, and from the start we'd rather have you fill in forms than actually talk to you.

Well, note above I've already said I don't see much value in testing "personality" alone, especially (ridiculously) as the only hiring test... but even with that in mind, you're adding a lot of negative spin and loading there. What if instead of your presentation above, it's more like:

"As you've noticed by now, we screen and evaluate people very thoroughly to make sure we're the right fit, and know their strengths and weaknesses right out of the gate. And as you'll remember from last week, after the first two interviews and the security checks, we use a written test as part of that evaluation, so we'd like you to set up an appointment with you to do that. It's the same test you'll take again every year or two while working here, in fact, though of course the questions change each time. It will take about four hours, supervised, with a lunch provided in the middle, and of course you'll be compensated for the time. Now, when can you be available?"

(That's a rough version of what I am familiar with. As I said way above, I'm a bit boggled that people would try to slim down this kind of screening for, like, supermarket checkout staff.)

You are right, tough, that tests definitely also tell the applicant a lot about the company and what it values. And that's a great side-effect. For example, I know one firm once used a test that included the problem of optimizing Mario's route through a supplied level, and another that required translation of text from a "fictional" language (by inference) that just happened to be Sindarin.

Other tests sometimes include impossible, contradictory or erroneous instructions, because in those cases, employers wished to measure how applicants would deal with such a problem. (I'm guessing those firms must have a lot of managers giving contradictory instructions!) I've heard stories of tests deliberately interrupted by designed fire alarms, but those might be apocryphal.

So yes, an applicant can indeed learn a lot about a company's values and culture from a good screening test. Bonus value.

Like I said earlier, we take hard, grueling exams to get into schools, and the screening processes for better schools are harder and more complex than your local State U. And we accept that they're generally useful as a part of a bigger application and evaluation process.

Good employment evaluations can be valuable too.

(And I really need to stop talking about work stuff. I think I'll go back to one-liners about hipsters and boobies instead.)
posted by rokusan at 7:48 AM on August 25 [1 favorite has favorites]

And a mobile app which people could, say, have installed on their phones, so that it'd be handy during the inevitable times when they're forced to fill these idiotic things out while shuffling through job applications... well, that'd be awesome. Increasingly awesome the more popular it became.

I can't wait for the day when you're required to have a smartphone just to get a bullshit service-industry job. Between that and pay-for-internships, there are some exciting new developments in the world of poor people getting screwed these days.
posted by box at 7:50 AM on August 25 [4 favorites has favorites]

No, see, that right there reveals the truly odious thing going on with these tests and the soulless companies that use them. Target, Lowe's, Sears...

There's a reason why they lost their tower...

On top of that:

...and never forget that a SECRET SHOPPER might be out there any place at all to grade your sorry ass against them, no matter what the practical outcome of your interaction may be.

If there is any way of truly manipulating staff into fireable positions or major emotional holes, it is that right there. Or you could talk about a "compliance visit" or some upper-level manager/God coming. And in this economy, corporate has more power than they ever had before, as a loss of this job probably means long unemployment.

Unkind words for the very unkind souls making pretty decent salaries and spearheading this sort of junk: To hell with you and all your friends. -Taking Back Sunday

Good day!

PS: Sad thing was my company also had this third party employee survey recently that had some real BS questions with answers on a scale of 1-5. No open ended. Obviously, the company employed this third party for the best possible answers while giving the impression that they were listening to employee concerns.
posted by JoeXIII007 at 7:52 AM on August 25

People are hostile because these sorts of tests are really a study in miniature illustrating everything that's wrong with management today. It was once a careful and subtle art, interviewing people

Large chain retailers, especially, have been using this kind of screening tool since the 1930s. Unicru is just the Wal-Mart of this area, having bought up or run out of business most of the competition. Loompanics (I think) used to publish a book of cheat codes to the most popular screening tests administered in the 1950s-1980s, which included questions like "I know who is responsible for my problems" and "I sometimes hear voices."

Getting a job at Sears/K-Mart/Woolworth's, when it existed, and stores of that sort in the last 70 years has rarely involved the exercise of "a careful and subtle art" of interviewing. Multi-outlet retailers have used standardized testing as the central element of candidate assessment for that long, and any use of personal judgment by the interviewing supervisors had to be backed up by the results of the test.

If you want to read some real horror stories about American corporate looniness, check out Ken Alder's history of the lie detector.
posted by Sidhedevil at 7:52 AM on August 25

flabdablet: You know what I'd love to see tried? A Reamweaver clone of the Kronos website, backed by a Yes Men style sales force flogging a Unicru clone that, unbeknown to the client, works by completely ignoring all survey answers and green-lighting 25% of applicants at random.

I would be very surprised to find that Unicru, as currently implemented, worked better than this.

Hmm. What we need is a grad student in clinical psychology with an emphasis on psychometrics. This whole thing stinks of killer graduate thesis: 'A double-blind study demonstrating the inefficacy of common 'intelligence' or 'personality' tests frequently used in hiring employees..."
posted by koeselitz at 7:52 AM on August 25

Yeah, these tests are ridiculous. Even if it was theoretically possible to measure personality this way with a willing participant, who actually wanted to measure their personality accurately it would never work in an adversarial situation like this. Of course, people are going to answer the questions as if they were "nice pleasant people", even if they're not.
posted by delmoi at 7:55 AM on August 25

leads me to believe that the worth of such tests has been greatly exaggerated by companies who stand to make a profit from them, and that scientists in the field have generally found this quite vexing. Maybe I'm wrong.

I feel your doubt. It's a small but real value, in line with how actuaries evaluate populations for risk: a small percentage risk-reduction over many people or much time yields a measurable benefit.

I once thought it was all a joke too ("Bah, I can manipulate my answers to say anything!", I thought), but I have slowly gained respect for at least some testing from years and years of (sorry) anecdata. The real data collected by the particular feedback and learning systems I worked on lo those years ago? I wish I could sneak in and see the results in detail, but that's a contract that ended a decade ago. I know the process is still in use, feeding accuracy data back and flagging questions that don't correlate to later reality as "low value" for reworking, so the incremental improvement aspect is working... but I definitely don't get to see the numbers.

(I don't make money from testing, and I don't mean to sound like an apologist for the whole industry, which I agree includes a whole lot of snake oil. I just work "around" it as explained above, and there's really no way to not deal with it in modern hiring. In some ways (see above, again), formalized psychometric testing is actually my competition: I'm an old-school guy at heart.)

But I have learned to live with the Beast, and even see some value in it.
posted by rokusan at 7:58 AM on August 25

Some of the older standardized personnel tests include the Wonderlic Personnel Test, the Hogan Human Resources Assessment, and of course the always popular Minnesota Multiphasic Personality Test and California Personality Inventory.

Note: I think this stuff is dangerous bullshit. But it's not new dangerous bullshit.
posted by Sidhedevil at 7:59 AM on August 25

Of course what these tests will find are people with enough empathy to understand out what they're "looking for".
posted by delmoi at 7:59 AM on August 25

Delmoi, if you think you know what they're looking for, you're obviously too clever and you will get weeded out as a "gamer".
posted by amethysts at 8:01 AM on August 25

And for the tenth time maybe, because most times I buried it in a Wall of Text... my experience is with cognitive tests that measure (mainly) problem-solving ability and specific skill sets. Many of these tests include a small percentage of "personality" type questions.

I can't speak for these one-page personality thingies.
posted by rokusan at 8:02 AM on August 25

'A double-blind study demonstrating the inefficacy of common 'intelligence' or 'personality' tests frequently used in hiring employees...'

I also want this. Badly.
posted by rokusan at 8:03 AM on August 25

rokusan: If you say "all testing is useless", you are suggesting the equivalent of all college admissions boards throwing away the SAT/LSAT/MCAT and basing their selections only on the student interview. Nobody would argue that position. And choosing a job (or employee) is simiilarly important to choosing a school (or student).

Yes, I'm saying that college admissions boards should throw away the SAT/LSAT/MCAT. Good god, you're talking like somebody who hasn't really taken those tests or thought about their implications.

Look, I spent three years in graduate school as a master's student. I was seen as inferior in nearly every way. Master's students at the school I went to were supposed to be the 'professional people,' the folks who had enough money to spend a few years going back and getting their degree whilst working part-time; we weren't serious like the PhD candidates. I don't mind being seen as unserious, but that meant I paid my way without a drop of financial aid while people who often really couldn't care less got paid to go to the same classes. Why? Because I did badly on my GRE, a test which clearly measures test prep above all else, like almost every one of these sorts of tests. The SAT is even more of a joke, and I believe that fact would be a household commonplace if the College Board didn't have an army of lawyers preventing the publication of studies which undermine its profit margin.

The school I went to for my undergraduate work, a blessed place, pointedly did not require SAT scores; in fact, a colleague of mine and one of the best students that school had while I was there never finished elementary school. She would never have been able to go to most colleges; but they liked her in the interview (yes, they base most of their admissions on a several-day interview) and let her in.

It's my experience that people who think blithely that the SAT/LSAT/MCAT are somewhat akin to 'necessary evils' or 'not great, but functional anyway' are people who haven't really spent much time in education, caring about how it works; they certainly haven't had their life's goals hanging on the thread of a stream of statistically-derived questions before. Take it from me: those professors who watch students come and go every year and actually care about where it is they'll end up, those parents who want to see a smart kid do well, and those students who have something they actually want to achieve? Overwhelmingly those people can't stand the SAT, the LSAT, the MCAT, the GRE, etc.
posted by koeselitz at 8:07 AM on August 25 [6 favorites has favorites]

me: 'A double-blind study demonstrating the inefficacy of common 'intelligence' or 'personality' tests frequently used in hiring employees...'

rokusan: I also want this. Badly.

Then your belief that the psychometric tests you're talking about are useful is unscientific and has no basis in reality.
posted by koeselitz at 8:08 AM on August 25

Koeselitz: Just how much more effective have psychometric tests been shown to be in hiring employees when compared to traditional interview-style hiring? And what independent scientific research can you point to that shows this to be the case?

There's a lot of good, academic evidence that 'traditional interview style hiring' doesn't actually work that well. Traditional, unstructured interviews aren't terribly effective in predicting job performance. They just aren't. Possibly because people over-rate their own interviewing ability, possibly because there's too much variation between the questions asked of interviewees.

Structured interviews, as I've mentioned upthread, are effective - they are the most effective selection method. Work sample tests (where you actually demonstrate a specific skill that you would need to use on the job) are also very effective. Cognitive-style psychometric testing is next. Personality testing has a small but significant independent effect. Checking someone's references, graphology - basically worthless.

Afraid I can't give citations, I've long ago got rid of my I/O texts, but any basic text should give you similar information (I linked to one, upthread).

I want to stress, I'm arguing this from an academic perspective. I've never worked in a personnel selection role. Anything I'm saying comes from my memories of reading academic research into personnel selection.

Also want to stress: an interview should be the starting point. I fully agree with you there. Psychometric testing is an additional tool, to be used where appropriate, to provide extra information.
posted by Infinite Jest at 8:12 AM on August 25 [2 favorites has favorites]

Yes, I'm saying that college admissions boards should throw away the SAT/LSAT/MCAT.

Yes, and rokusan, many people in the field of post-secondary education are arguing the same thing: this is not a crazy fringe position within the field at all.

Some college administrators think it would lead to more diverse campuses. But even Charles Murray, no partisan of campus diversity, thinks colleges should abolish the SAT.
posted by Sidhedevil at 8:15 AM on August 25 [1 favorite has favorites]

You know, there are a lot of big box retail chains that use these tests, and back when I worked shitty retail I took a lot of them.

they don't matter. at all.

I've answered these tests mostly without using the SA and SD answers, and I disagreed with these right answers on a lot of them. I still got hired. the reason for this is that the test is a formality the corporate office insists on, but the local manager just hires who he can and picks the guy he approves of. then he files the test away, and doesn't even read it. Most retail chains have revolving door hiring policies because they don't need long time employees, they just need warm bodies. They'll always have kids back from college, or the unfortunate older people whose retirement funds aren't cutting it, or the young aimless slackers before they finally get whatever their ultimate career will be started. I've never seen a manager say "we can't hire you because you failed the Unicru test," but if they do it's probably just because then it's the test's fault and not theirs. it's easier than saying "I just don't like you." but honestly, most don't even bother with that. they just don't call you back.

any corporate efficiency expert who wants to revolutionize the industry would do well to immediately end any company's participation in this absurd test. It's like jaywalking: sure it's technically official policy, but nobody on the street is enforcing it unless they've had a bad day and they need to take it out on a stranger.
posted by shmegegge at 8:17 AM on August 25 [1 favorite has favorites]

"Myers-Briggs style information" is effectively meaningless, psychologically, and is generally only preferred by companies that sell this kind of snake-oil

Don't forget it's also preferred by LiveJournal and Facebook pollsters.
posted by Spatch at 8:18 AM on August 25 [2 favorites has favorites]

which included questions like..."I sometimes hear voices."

You know, I'd have to answer this one with a big yes.

Unfortunately, standing on your desk and screaming, "Will you people keep it down over there? I'm trying to get some work done!" over the cubicle wall is not viewed as good corporate ettiquette by some.
posted by Kid Charlemagne at 8:20 AM on August 25 [2 favorites has favorites]

people get hostile when their innate personality traits are essentially commodified - and that's the real problem with these tests - instead of treating your labor as a commodity, which can be bad enough, they treat YOU as one

Interesting point that I hadn't considered; but surely any selection method is going to do that? If the only method was an interview, they'd still be looking for specific traits. In my personal example above, they wouldn't ask me if I preferred sky-diving to reading poetry on a test, they'd ask me face-to-face questions that would hopefully demonstrate whether I was a risk-taker. They're still looking for the same traits, just in a different way.

amethysts: your situation sounds wrong and bizarre. The only reason I can see for a personality test is to add information to the decision-making process - helping choose between two or three candidates. Not as a reason to refuse to hire someone.

Interesting thread people. Thanks.
posted by Infinite Jest at 8:23 AM on August 25

Somewhere I read that you will know you've chosen the wrong career goal if your dream job includes having your name embroidered on your clothing. I guess this would be another indicator: if the job process STARTS with an automated questionnaire...

MDs frequently have their name on their coat, and their job process starts with electronic applications and standardized tests.

While these tests obviously dehumanize the applicants, they also speak volumes about how the company perceives its managers. These tests basically say, "we don't trust you to be able to hire competent workers, either because you're too stupid to be able to tell a good egg from a bad one, or because you're too busy to be able to spend the time interviewing potential candidates."
...
Just how much more effective have psychometric tests been shown to be in hiring employees when compared to traditional interview-style hiring? And what independent scientific research can you point to that shows this to be the case?
...
Psychometrics as such is really still only a social science; I say that in the sense that it hasn't at all been able to demonstrate any predictable evidence on the personal level but only on the social and statistical levels

In the mental health world, standardized instruments are more reliable and valid in establishing a diagnosis than unstructured interviews with a clinician. This pisses some professionals off to no end.

It is easy for one of these big box retailers to test the hypothesis suggested above. Randomize a sample of stores to the new test, and look at the change in labor expenses for those compared to the change in expenses for controls. Given that the seller pushes that they are a cost-saving measure and how metric-focused these chains are, I'd be really surprised if they haven't done this.

33% seems really high for it to be statistically insignificant, but I'm willing to give him the benefit of the doubt.

I don't know what he's even trying to say. A big problem with using any standardized test as an employment screen is the degree to which is functions to discriminate against minorities, who might on average actually be worse employees for whatever reason.
posted by a robot made out of meat at 8:23 AM on August 25

If we humans had uncovered tests which somehow measure thinking ability and rationality - really, that's no small feat - don't you think psychologists would know about them?

Again, my direct experience is not with the emotional/personality stuff (though I find it interesting) but with the adjacent testing for more technical/specific things that are applicable to the sorts of thinking/organizing/processing that employees would actually be doing in the jobs in question (see the Mario and Sindarin examples).

In those cases, testing has proven to be a much better indicator of ability and future performance than interviewing alone, as you'll agree if you choose to just accept my resume's claims that I am fluent in Swahili, have perfect pitch, and can throw a nickel through a soda can at 100 meters.

(The "eerily so" accuracy comes in comparing pre-hire screening test results with later, measured performance in the actual jobs that showed high and low performance in the "expected" areas. There's not much room for misperceiving that.)

The small sidecar percentage of psychometrics in tests I have touched seem to also track to later results, at least according to the managers and firms doing the hiring and evaluating. But I also agree that that is most definitely where it all gets pretty darn murky, at least for me (not my field)... and I can imagine a lot of confirmation biasing going on: everyone is likely to remember that the employee who was 'predicted' to be scatterbrained indeed lost the Important Case File ("aha, it worked!") but no, nobody will remember the other six people who lost files that year.

What I personally touched was all about adding feedback loops and processes in test-making so that question results that did not correlate with later performance (inaccurate predictors) were flagged, culled, changed, so the system would naturally produce more accurate measures over time. Generously: an evolving learning system. Unkindly: trial and error.

One partner at the firm called it "Preventing the PhD's from running amok."

Me, I consider it victory enough when I can get people to actually look back and measure results... like, ever, and this applies to a lot of processes in business, not just hiring. There's a heck of a lot of willful blindness + forging boldly ahead in Corporate 'Merica.
posted by rokusan at 8:24 AM on August 25

Don't know about corporate whatever, but the only job for which I've had to take a personality test was with the federal government.
posted by grobstein at 8:35 AM on August 25

Myers-Briggs isn't snake oil. There have been times I've figured out why I'm pissing someone off, just by thinking our differences in Myers-Briggs-like terms.

It's just that that's never how they sell Meyer-Briggs to you. The business model that has grown up is geared towards people who want to extrapolate in all kind of goofy ways while simplifying it into a brief presentation lest it cut into labor hours. In other words, the same kinds of people who swear by the unicru test want a Myers-Brigs consultant whose talk takes an hour at the most. As a result, you get something that sounds like, "Hi, I'm Bob and I'm a Taurus."

Similarly, if I had a dollar for every company who was all over the Total Quality thing, while all but taking a wizz on everything W. Edwards Deming actually suggested, I'd be a wealthy man.
posted by Kid Charlemagne at 8:46 AM on August 25 [1 favorite has favorites]

I just administer the Voight-Kampff. Sure it takes longer, but I know which employees to put on manual labor and which ones won't be around in five years.
posted by The Whelk at 8:48 AM on August 25 [7 favorites has favorites]

Then your belief that the psychometric tests you're talking about are useful is unscientific and has no basis in reality. -- posted by koeselitz

As I said, my experience is with tests that include a small (my guess is five percent) of the sorts of personality question this post was about. I began assuming they were nonsense filler, and I know for a fact that at least SOME of the "weird" questions are just inserted to clear the mind between sections, but I have slowly gained some respect for even the weird questions over the years. They do seem to identify some employee problems. I don't claim to know why: my direct involvement was only in adding QC and automation to test development, so that questions that proved worthless or low-value could be automatically weeded out. My indirect involvement is the need to work alongside these things (from firms I know and firms I don't), whether I like them or not. They are getting very common.

(I'm sorry that most of what I know of the psych stuff is anecdote. It's all I have on that part, and I'm trying to be clear about when I'm talking about what I know vs. when I'm reaching for anecdote or guesses, but in the thread here there's been a lot of jumping between the tests I know, the work I have seen, these tests in general, and the actual "personality" tests in the actual post itself, which are different and, it seems to me, the weakest type.)

But for me, good testing added to other methods has produced better hires than no testing at all. Plainly: when I have endorsed hires with who had supporting testing results (I don't conduct such tests but many employers require them now, so they're administered by whoever's running the thing), they have worked out well a higher percentage of the time (about ten percent more often) than those evaluated from interviews alone: yeah, I keep track of every person forever.

So, yes, all I am certain of is that certain testing has improved the value of the interviews and other screening that *I* have been involved in by about ten percent. And from that I speculate that others who claim they are getting SOME value from testing are probably not lying. Whether the percentage improvement is worth all the effort, cost and stress is a valid question, but I can't accept "completely worthless" as correct: that's too in conflict with my own experience.

And also again: the kinds of "good" tests I'm familiar with come a lot closer to "highly structured interview" (just step by step, and on paper) than to "personality test", and are composed mainly of problem-solving questions, general knowledge, and domain-specific knowledge. I don't see the value in "Are you a cat person or a dog person?" type questions, myself.

But in general, testing can act as a QA check on an interviewer, too, which is useful and can "save" a good applicant who just happens to have hit a personality conflict (or a bad day) with a certain interviewer. I imagine a scary-good SAT does the same thing to save some bad college interviews.

Must. Stop. Talking. About. Work. Now. Dammit.
posted by rokusan at 8:48 AM on August 25

Yes, I'm saying that college admissions boards should throw away the SAT/LSAT/MCAT.

Yes, and rokusan, many people in the field of post-secondary education are arguing the same thing

That's a good point. So there are equivalent opposed-to-all-testing voices everywhere, I suppose. And there are probably people who wish to just change/evolve those tests to make them more useful, or even (gasp!) try to measure their real value. I'd probably be with those folks: the boring practical middle-roaders.

My hunch is that they're neither worthless nor of paramount importance, but that they have some value somewhere in between, and should get some consideration combination with other factors.

The value of the "somes" is open for all sorts of evaluation, not just debate, and I'm all for that.
posted by rokusan at 8:55 AM on August 25

Myers-Briggs isn't snake oil... it's just that that's never how they sell Meyer-Briggs to you.

Ha. Yeah, exactly. It's the whole tool-vs-how-its-used problem.
posted by rokusan at 8:57 AM on August 25

Oh lord, how I've wished for a world in which all hiring was done by questionnaire only.

In my last interview, I was asked about my favourite magazines and the last movie I saw. It took every ounce of my strength not to remind the interviewers that I wasn't there for a haircut.
posted by Sys Rq at 8:57 AM on August 25

I have to wonder, too, if the HR people here on MeFi can see the writing on the wall with regards to the growth of tests like this and the continued used of automated systems. Your days as professionals are numbered. Your jobs are being reduced to simply compiling numbers and checking them against a pre-packaged "acceptable" range of results. Hiring and firing is now a simple pass/fail metric, with no regards to any actual human qualities. And, of course, anyone with a GED can do that...or it, too, can be automated.
posted by Thorzdad at 9:01 AM on August 25

If you say "all testing is useless", you are suggesting the equivalent of all college admissions boards throwing away the SAT/LSAT/MCAT and basing their selections only on the student interview. Nobody would argue that position. And choosing a job (or employee) is simiilarly important to choosing a school (or student).

Well, actually, I would argue that the LSAT/SAT/GRE etc. should be thrown out. With the advent of the test prep industry, all these tests test is your ability to prep for the test. Seriously, that's all. Kaplan et. al. are successful at gaming these tests. I myself was successful at "gaming" the GRE (99 percentile math, not to brag, but really, I'm only mediocre at real math).

You suggest that in the absence of these tests, all schools would be left with are the student interviews, but that's untrue. In fact, just considering the GRE and SAT, schools outside the US do not use these at all. The MCAT and LSAT are not widely used outside North America. Toss away the standardized tests, and you have: transcripts (useful outside of where grade inflation has taken hold), CV, letter of intent / application essay, interviews, correspondence with potential advisors, letters of recommendation from reasonably trusted sources... Seems like a reasonably basis for admissions decisions (which, in the case of a research Ph.D. is very much like a hiring decision: in my grad school, a grad student would be working for a prof. for 5 to 6 years and cost something like 70K a year just for tuition, stipend and overhead. Add research costs and the advisor and department is looking at something approaching real money...).

In my own industry (hires mostly Ph.D., some master's scientists and engineers), one of the largest companies were quite taken by standardized tests and war-games like simulations and team challenges scored by clipboard wielding psychologists / HR types. My understanding is that they have quietly moved away from that and returned to what everyone else is doing: hiring based on proven, demonstrated accomplishment and a series of interviews.
posted by bumpkin at 9:03 AM on August 25 [1 favorite has favorites]

Infinite Jest's order of effective filtering above is exactly right, by the way, from most reliable filter to least: (1) structured interviews; (2) work exercises; (3) cognitive testing; (4) personality testing; (5) references.

The tests I like (the ones I keep calling "good" because they work) are a planned ratio of (1+2+3) with a little (4) for... I am not sure what reason: see above for my guesses at that.

I am used to seeing the smattering of (4)-style questions, but I can't imagine using a (4) alone for any purpose other than a party game or Scientology recruitment.

References (5) are near worthless, indeed, a fact which shocks people the same way non-lawyers react when hearing that eyewitness evidence is the least reliable.

I've never seen anyone actually use Graphology, at least not consciously. I guess that could be (6).
posted by rokusan at 9:05 AM on August 25

Nobody would argue that position.
Well, actually, I would argue that the LSAT/SAT/GRE etc.

Forgive me. Per Sidhedevil above, I see now that such a no-testing position does indeed exist in academic admissions debate.

I'd try to focus on improving the testing, rather than throw out the whole baby, but that's me. I know what you mean about test-taking skills.

You suggest that in the absence of these tests, all schools would be left with are the student interviews.

That was in response to someone above advocating a Job Interview alone instead of any testing. It was an argument for interview-only, and I was saying that made as little sense in a business hiring situation as in a school application. Not a perfect parallel. I know all employers would love to see a "transcript" from previous employers, for example. But that isn't available.

An interview alone is very very weak, no matter how talented the interviewer.
posted by rokusan at 9:11 AM on August 25

1. These tests are fairly stupid.

2. Though some of the questions might be illegal in a better world, overall you won't ever be able to pass a law saying, "You can't ask job applicants questions."

3. Unfortunately, these tests mostly occur for un- or semi-skilled jobs, where there are many more applicants than positions (thanks to the Fed's policy of "tightening" money when unemployment falls below a certain level, but I digress) so that you have a lot of people, quite a few of whom are actually unemployable, competing for a job that isn't really worth very much to either the applicant or the employer.

4. In these positions, making the applicants demonstrate any sort of competence at all will weed out some percentage of the unemployables with some sort of accuracy. ("Applicant was unable to use pencil.")

5. If capitalism actually worked as advertised, tests that had no predictive value would fade away and eventually all the tests would be pretty good. That won't happen, unfortunately, as management only allows the "invisible hand" to apply to workers and not to themselves.

6. While a very few people will be wrongfully not hired because of these tests, even your below-average specimen should be able to fake their way through these without issue.

Conclusion: People should know about these tests, this is a valid concern, but it's under control, there are far greater injustices.
posted by lupus_yonderboy at 9:13 AM on August 25

Looking back over my comments here, I worry that I sound like I'm taking things far too personally - sorry for that. No insult or denigration intended or implied; all friends here.

One final point, though: I said earlier that this was "still only a social science." A small jibe from an erstwhile political scientist; what I meant by this and was trying to explain (badly) was that these sorts of tests are 'improving' constantly - but only along the statistical dimension. The unfortunate fact is that statistics is really only a very, very loose way of considering almost anything; especially something as complex and multifarious as human cognition. I don't doubt that people will be able to boost tests like this to the point where they're eighty-five or even ninety per cent effective, but unfortunately there's a ceiling that they will inevitably hit so long as the only data we're using to refine them is data about which portions of which populations answer particular questions a certain way.

To go further than that, tests like this will have to go the route of biology, of neurology, and various other studies. But until you can sit any human being from any population on earth down in front of me and have him answer standardized questions and be able to predict from his answers exactly what his thinking ability will be within, say, 99% - until research in this field focuses on that kind of testing, leaving behind counting heads and calculating percentages of this group and that group that answered the question a certain way - until psychometrics starts setting down a single human being and attempting to answer the question: what biological necessity led this person to give these answers, and what may we predict on the basis of that biological necessity? - until then, I don't really see how psychometrics can have the real worth that most other sciences currently in everyday use can have.

And, needless to say, until then, I don't think psychometrics will be refined enough to be used as a predictive theory for hiring; it's simply not specific or precise enough to tell us about individual humans. Yes, it can tell us general things about the statistical groups certain humans belong to, but things like race, social status, childhood, geographical origin, et cetera can often tell us as much if not more, so long as we read the data the way they ought to be read.

In particular, I strongly believe that psychometrics shouldn't be used in educational assessment. The whole point of accepting new students is to find people who will improve our ideas and refine them; that is, to choose people who can make the current range of psychometric tests obsolete by inventing a study of psychometrics that is a whole order and rank beyond the current one. The use of our own shoddy psychometrics seems to me to conflict unfortunately with that goal
posted by koeselitz at 9:26 AM on August 25 [1 favorite has favorites]

... and I guess one thing that I haven't realized is at the heart of my position is this: statistical testing, as opposed to actual direct scientific testing, is inherently flawed because it's vastly more difficult to judge completeness of a statistical test. In order to say something real about innate thinking ability, statistical testing of populations has to start from a point beyond social conditioning - which is, I think, an astronomical task, a task so incredibly difficult and monstrous that many people who conduct studies conveniently forget that it's even there. But consider: while psychometrics likes to claim some predictive powers, and the studies seem to back this judgement up in a certain way, nearly all of the studies and tests concerning psychometrics have been of the statistical variety, and nearly all of them have been conducted in modern western countries. What of our socialized habits of thinking? What of our own cultural and traditional views? I've seen the kinds of tests that linguists like to administer to remote tribes to determine to what degree thinking is affected by language; this is just a thought, but as I recall, those tests were oddly similar to the kinds of tests that are often claimed to be predictive of things like problematic thinking. Needless to say, the South American natives answered those kinds of questions in vastly different ways that we did, often. Does this mean they lack our ability to think critically? Or is it possible that tests like this are socially and culturally biased?

The only way to really start to eliminate that element would be to take the science in the material direction: to look for actual causes within the human body for particular answers and extrapolate from there.
posted by koeselitz at 9:37 AM on August 25

You know, there are a lot of big box retail chains that use these tests, and back when I worked shitty retail I took a lot of them.

They use them at their corporate offices too. I've had to slog through them during job searches and interviews at local behemoths.

The HR people who make the decision to administer the test are probably people who passed the test themselves. It's like a self-replicating virus.
posted by gimonca at 9:38 AM on August 25

eHarmony could replace all of this with a much simpler test involving only 29 dimensions of compatibility.

Yes, they could. Scroll down to "workflow"
posted by lysdexic at 9:45 AM on August 25

How do you fake politeness? That's like faking running. You're either behaving politely or you're not.

(I concede that giving a colleague the double finger behind their back may be a grey area.)
posted by chesty_a_arthur at 9:46 AM on August 25 [2 favorites has favorites]

I don't think anyone is trying to test for "innate thinking ability", koeselitz. At least, I hope they're not, for all the reasons you cite.

I don't even know what something as general as "How well does this person think?" even means, really.

But,"How skilled are you at solving this certain kind of problem?", on the other hand, is pretty testable, and it's also pretty useful if those problems are designed to closely mirror the kinds of real problems the person would be expected to handle on the job. The easiest example is a programming question ("Make the following sort faster.") but there are equivalents in most fields.
posted by rokusan at 9:48 AM on August 25

rokusan: But for me, good testing added to other methods has produced better hires than no testing at all.

I know this is at the core of it, and it's a fair point; you're doing what works, and that's the whole goal.

I guess I'm only trying to suggest gently (okay, I've utterly failed the 'gentle' test today, heh) one thing: I wholeheartedly believe that any gains made by refined testing, at least given the current state of the science, would be shown to be miniscule beside the gains that you would make if you instead spent time contemplating how to interview and choose people and how to gauge a person's intelligence and thoughtfulness.
posted by koeselitz at 9:50 AM on August 25

In the context of high-level jobs, it's what was said quite far above: a single bad hire can cost you hundreds of thousands of dollars in lost profitability. Any input that you can use with any hope of validity is going to be welcome. I strongly doubt that many high-level employers rely upon the personality dimension of psychometrics for more than a minor element of their overall recruitment process.

In the context of low-level jobs, I think the use of personality inventories is more a symptom of a several ills than it is a disease. Low wages and no benefits are the most important thing. You are not likely to get high quality workers when you are offering $9 an hour. You are virtually certain not to get high quality front-line management, of the sort who can choose and supervise $9-an-hour workers effectively, when you are offering $15 an hour for that job. Even a few percentage points of correlation to performance will seem helpful.
posted by MattD at 10:05 AM on August 25

The roots of the placebo problem can be traced to a lie told by an Army nurse during World War II as Allied forces stormed the beaches of southern Italy. The nurse was assisting an anesthetist named Henry Beecher, who was tending to US troops under heavy German bombardment. When the morphine supply ran low, the nurse assured a wounded soldier that he was getting a shot of potent painkiller, though her syringe contained only salt water. Amazingly, the bogus injection relieved the soldier's agony and prevented the onset of shock.

no, you only think i posted that to the wrong tab
posted by pyramid termite at 10:06 AM on August 25 [1 favorite has favorites]

fairness is an abstract concept invented by humans. Us automotons are better.

sscct
posted by MrTenacious

Acronym Finder: SSCCT stands for Seasonal System Capacity Combustion Turbine.
posted by StickyCarpet at 10:09 AM on August 25

People get hostile when their innate personality traits are essentially commodified - and that's the real problem with these tests...

Well, it's better than categorizing them for easy medication.
posted by rokusan at 10:34 AM on August 25 [1 favorite has favorites]

rokusan: But,"How skilled are you at solving this certain kind of problem?", on the other hand, is pretty testable, and it's also pretty useful if those problems are designed to closely mirror the kinds of real problems the person would be expected to handle on the job. The easiest example is a programming question ("Make the following sort faster.") but there are equivalents in most fields.

What kind of sort? What's the best answer? What language is the applicant required to use? These are all questions that relate to the utility of the main question: how would you make this sort? What I'm arguing isn't a sentimental hope that people have more face-to-face interviews; it's that the answers to these questions can't be parsed on the basis of a standardized mechanism, at least not meaningfully.

Programming is actually a very good example; I know some programmers and I've heard them grumble about bad interview questions, so it'd be good to have a better standard.

The trouble is that programming, like many employments, is a creative task; as such, it quite naturally tends to go 'beyond' questions like that. Great programmers - at least the ones I've met - are the ones who would've answered a question like "how would you do the following sort?" with something like "there's no point in sorting this data because x, y, and z." That's not an answer that's on the standardized answer sheet, but, as is often the case, the answer that's not on the sheet is often the best one.

There are examples of this all over, but I'll just give one: my dad taught calculus at a college in Nigeria in the 60s. He always says that one of the interesting bits was how much people who've never had a straight line in their lives besides the horizon think differently about geometry and applied geometry like architecture; the world to them was round, not a continuum of three dimensions jutting out in different directions, and as such when it got laid out flat in their minds it looked different than it does for us. Crudely, it's where Picasso got many of his spatial ideas; but it's hard to explain completely, beyond noting that "sides" turn out to be a somewhat artificial way of looking at the world.

And those kinds of people would approach a simple math problem designed to test this kind of thing in a wholly different way; the answer to such a question might be highly useful in determining their potential worth as an employee, but it's impossible to standardize a set of answers to select for that; or at least very difficult.
posted by koeselitz at 10:37 AM on August 25 [3 favorites has favorites]

I guess I'm only trying to suggest gently (okay, I've utterly failed the 'gentle' test today, heh) one thing: I wholeheartedly believe that any gains made by refined testing, at least given the current state of the science, would be shown to be miniscule beside the gains that you would make if you instead spent time contemplating how to interview and choose people and how to gauge a person's intelligence and thoughtfulness.

I believe you believe that! :)

You are speaking not only from faith yourself, but also from an academic or philosophical foundation, at least as I read you. I can respect that the world sure would be better if people all evaluated other people in the way you suggest, and clearly you have a passionate hatred for quantitative attempts to measure of human abilities. You, my friend, are a poet.

Me, I am deliberately speaking from a hands-on, practical basis. I am a better than average interviewer, I am told, and I have done this for a living (or half a living) for quite a while now, most of it before testing was commonplace in creative, management, or exec positions. My initial and gut reaction to testing was all negative.... other than a sick fascination with the puzzle questions, that is.

But to flat-out reject a potentially helpful tool like testing in favor of a "pure" interpersonal approach... that would be like a carpenter refusing power tools, or a musician flat-out rejecting amplifiers. It's a bit too Colbert: Forget all this newfangled crap... think with your gut! Hire who you feel!

Yes, it has some merit, and it even has a seductive appeal to the anarchist and the romantic in me. My soft, human ability (skill?) at 'measuring' people in interviews has been a very marketable and useful one, and it's why I started doing such work in the first place. (I am also a pretty good cold reader. Buy me drinks.)

But it won't move bottom lines as much as it can in concert with other methods, and in the churning business trenches, that's all that matters. If a test can provide a five percent reduction in the chances of a new hire burning out in a month, or identify those who can figure out some percentages without needing to run Excel? That five percent chance can be worth a lot. I suspect that's the reason these big chains linked in the story are trying to adapt some of these methods (the cheap, easy, ineffective ones) for their 100,000 high-churn employees. It's a misapplication, but I can understand the motivation.

Testing works (a little), and better testing works a little bit more. And on and on. This is the machinery of progress, and it won't stop grinding.

So... I figure I can either curse the tools (smash the bandsaw! burn the amplifiers!), or work on using them in a sane context. My prior influence in "making" them was all aimed at making them more useful, which means more accurate, which by extension means more humane. I still think that was (is) correct.
posted by rokusan at 11:00 AM on August 25 [1 favorite has favorites]

rokusan - "Well, it's better than categorizing them for easy medication."

No, it's not. It's the same kind of thinking that made the DSM-IV one of the more dangerous things in the hands of laypeople. Suddenly everyone thought they had a mental disorder and needed to be medicated. It's relying too much on simple numbers in a complex system.

The part that I think I find the most offensive is that it is entirely due to lazy thinking and ignorance (which I personally abhor). But the majority of the human population is lazy and ignorant. Thus, lazy and ignorant practices are prevalent and define the whole of society. This has the effect of creating systems that are ineffective, and are driven by bad ideas (deregulated capitalism, national socialism, prozac nation, rampant debt-laden consumerism). There are no easy answers to any of these problems. But they are the results of people thinking that the numbers are the only thing that matter, and by applying a metric to something, you can rely on the numbers more than you can rely on having well trained and competent people in place overseeing the whole thing. The reality is that you can't have well trained and competent people overseeing the whole thing, because there aren't enough people who can fill those roles available in society. So you end up with metric based testing, and a valuation of individuals that doesn't account for the outliers who do not follow this standard model.

In general, people want things to be black and white. Yes or no. Shades of grey unnerve them and shake their confidence. Business especially succumbs to this thinking because all business really wants is to make money doing something (or sometimes doing nothing). It's all geared towards the goal of finding people who will do the job without changing the system. Hence, mindless, boring, and overly metric-based "goal oriented" corporate culture.

See "The Trap" by Adam Curtis to see where there is clear evidence of how these limited systems can go horribly wrong when the things they are trying to shove into round holes are dodecahedrons.
posted by daq at 11:01 AM on August 25 [1 favorite has favorites]

StickyCarpet: "Acronym Finder: SSCCT stands for Seasonal System Capacity Combustion Turbine."

close but no cigar.

So Says Caustic Cthulu Today
posted by shmegegge at 11:08 AM on August 25 [1 favorite has favorites]

"how would you do the following sort?" with something like "there's no point in sorting this data because x, y, and z." That's not an answer that's on the standardized answer sheet, but, as is often the case, the answer that's not on the sheet is often the best one.

You really don't want to get me started on screening programmers. :)

Yes, that question would obviously come with more information about restrictions and requirements and such ('given x memory... without regard for..."). Specific language never matters much in such questions, at least not beyond the entry level: an answer in pseudocode is the same as an answer in C++ or Python, because any reasonable programmer can adapt to any language. And of course it's a long answer, not a multiple choice one.

As for your other example, well, you see, it's a romantic issue again... your Nigerian with the radically cool and different view of space-time and geometry might indeed be another Picasso, but... and I say this with all possible kindness: Nobody wants to hire Picasso.

It's not about measuring overall intelligence, whatever that is, or some kind of idealized notion of Total Human Worth. That is never the purpose, as far as I've seen, of any employment test.

It's a much simpler problem: how can I increase the likelihood that whatever person we hire will be able to perform this job well? All comes from there, and gut feelings are only one piece.
posted by rokusan at 11:10 AM on August 25

That was sarcasm, daq.

I meant that people object to one but embrace the other, despite the fact they're the same kind of awkward shoehorn: forcing many kinds of people into few categories of (personality/medication).

Which is, I think, what you meant too.
posted by rokusan at 11:15 AM on August 25

I had to take one of those stupid tests when I applied for a job at Target back in high school. (I was offered the job, but ended up not taking it) From the questions, I suspect most of them were filler, with a handful that, if answered incorrectly would cause you to fail. For example, there were questions along the lines of "is shoplifting okay if you don't get caught?" "are you habitually late?" and "do most people think you are difficult to get along with?" Being fed the questions rapid fire might make people answer those questions slightly more honestly than if asked by a human in an interview. If used properly, I think standardized tests can help weed out people who will not succeed.
posted by fermezporte at 11:24 AM on August 25

The trouble is that programming, like many employments, is a creative task; as such, it quite naturally tends to go 'beyond' questions like that.

Yeah, the nice thing about programming is that you can get a pretty decent idea of how skilled a candidate is in an interview, but it really depends on the skills of the interviewer and it would be difficult to put into any kind of test that could be automatically scored. There are a few easy questions that you can use to weed out the people who are overselling their experience in their resume ("What does const mean?" etc.) but once you get into actual problems it helps to be there and work through the problem with them. Programmers really need to be able to clarify requirements, handle changes on the fly, think about edge cases, explain their designs to other people, and various other things that are identifiable in an interview but can't be determined from a filled out programming evaluation.

any reasonable programmer can adapt to any language

Sure, in theory, but you try getting a hardcore relational database perl junkie to switch to object oriented Java web services coding, or a C coding unix systems guru to switch to Windows Forms in .NET. If you are expecting a candidate to stick to a certain sort of best practices in the language they will be working in, it's not always obvious if their skills and habits from previous languages will translate.
posted by burnmp3s at 11:30 AM on August 25

sscct

[x] STRONGLY DISAGREE
posted by katillathehun at 11:59 AM on August 25

Personal experience: I've taken a couple of personality tests for jobs. I found honesty was the best option

Not when they ask "Do you occasionally like to use cocaine or marijuana?"

I answered "yes" on that one on purpose (I'd also suggest answering "no" when asked "do you like to start fights?") and not surprisingly was not invited to interview, despite have much more experience than necessary at a virtually identical position.
posted by mrgrimm at 3:45 PM on August 25

At my old job, HR was thinking about adopting a certain personality test for hiring, and they administered it to selected employees, myself included, to see if it had any predictive power. I answered honestly, and I found out later that I "failed" the test and would not have been hired based on my results. Apparently some of the company's top employees and managers also scored poorly, and based on this information HR wisely decided not to use the test.

After that, I've never answered those tests honestly, and I feel absolutely no guilt about giving whatever answers I believe will help me get me hired.
posted by arithosa at 4:31 PM on August 25 [1 favorite has favorites]

Though some of the questions might be illegal in a better world, overall you won't ever be able to pass a law saying, "You can't ask job applicants questions."

Not true. There are laws that state some questions cannot be legally asked of job applicants.

A few years ago, I applied for a part-time seasonal (Xmas) position at a craft store (a company I was once a store manager for and left on good terms with a rehire status). It was the first job of any sort I'd applied for in ages, and I was amazed at how the job application process had changed since I had been the person doing the hiring.

First there was a six page application asking me, among other things, to disclose all manner of financial information and authorize them to do a credit check, as well as informing me I had to submit to a telephone "personality" test.

Then there was the 45 minute personality test on the telephone, the questions being much like the ones presented in this post. Some were even more ridiculous. I just answered honestly. Seemed easier than lying, and after filling out the application, I was already getting the feeling I didn't want the job anyway.

Next, the person who the boss had been interviewing before me left, I went into the room to be interviewed, and the next person behind me went in to take the personality test. The interview was REALLY outrageous. Not only did the somewhat creepy guy sitting across the table from me keep staring at my chest, he asked just about every one of the questions interviewers aren't supposed to ask. Seriously, we were only about three questions into the interview when he started asking me if I had kids and whether or not I planned to have any. If it wouldn't have been too rude to just get up and walk the hell out, I would have done so. I was certain at that point, there was no way I wanted to work there.

He called me back a couple of days later, but I let the voicemail get it. He left a message saying he asking me to come in the next afternoon to talk, and I did, just to see the game play out. When I got there at the appointed time, he said he was too busy to talk, and he'd call me the next day. He never did, and I didn't bother following up.

During the height of Xmas shopping season, I went shopping at that craft store (I used to shop there often), and the poor employees were being run mad. As I left, I saw one standing outside on a break, and I asked her why they didn't have more staff on duty during this huge sale they were having. Not a single person had been hired, as not a single person had passed all three portions of the hiring process. They'd either had bad credit, failed the personality test, or the boss didn't like them.

Over a hundred applicants had already been through those doors seeking employment, and not a single one of them had been good enough for a two month, part time, seasonal position at minimum wage, including, apparently, myself with years of retail experience starting at the bottom and going all the way to corporate and having a good employment record with that very company (and every company I have ever worked for).

I thought it was a fluke --just an insane manager-- until I started applying for part time retail jobs elsewhere, and it was the same insane system, and mostly people not getting hired. And this was before the economy tanked and there were more jobs than people seeking them. Just truly strange. I only wanted a little extra pocket change and to have something that forced me to get out of the house for something other than running errands. Good thing I didn't need a job in order to pay bills or eat, because I never did manage to get a part time retail position. It would seem I am not hireable in the current environment and would make an awful employee.

Gah ... sorry so long. Too much coffee. But I wrote it, so I am going to hit the post button.
posted by Orb at 5:42 PM on August 25 [4 favorites has favorites]

About eight years ago I needed a short-term job, a few months. Applied at a chain grocery store. Interview and all was going great, they seemed to be moving towards a hire, and then they pulled out a printout and a bubble sheet.

This, they explained, was more or less an honesty test. Answer it frankly, and give the bubble sheet back. The questions were a mix of filler, questions about minor ethical lapses like shoplifting a pack of gum, and questions about larger ones like punching the clock for someone who wasn't in, or selling information to competitors.

I suspected, and still suspect, the idea was to select the people who confessed to a few minor ones and no to any major ones. If they confess to major ones, they're unfit and they're out. If they confess to no minor ones, they're gaming the test, so they're both an unknown and a rogue, so they're out.

This is tremendously inconvenient if you happen to be one of those weirdos who isn't good at lying and doesn't care much for gum.

But what could I do? Reverse-gaming anything so as to appear not to be gaming it is tricky under any circumstances, and reverse-gaming an 'honesty test' was way too mindgamey and metawanky to be worth a grocery store job. So I decided to play it absolutely straight. No, I have not embezzled company funds to sustain my faltering pyramid scheme. No, I have not shoplifted a pack of gum. If you actually wanted to know about me, you could call the character references that you asked me for, instead of doing this scientistic mumbo-jumbo.

Gave back my bubble sheet, and they called a phone number and read off my letter selections, no joke, and listened for a moment, and ended the call. And then raised a concern that they hadn't had when I walked in and told them how long I was available for, namely that I wasn't available for very long, and worked around to not hiring me.

Later that year the store manager ran for city council—and was forced out of the race when it came out that he'd been falsifying accounting reports in the store.

Morals of the story: One, beware of the overly suspicious; there's a reason they see deceit at every turn. Two, test-believing pinheads, though irritating, are a self-limiting affliction if you give them the chance. Being frankly incompatible with annoying people is a skill that has served me well many times since.
posted by eritain at 9:46 PM on August 25 [3 favorites has favorites]

Not true. There are laws that state some questions cannot be legally asked of job applicants.

It isn't that someone sat down and said "asking these questions is illegal", it's that asking those questions increases the likelihood of being sued for discrimination. So if you ask someone their religion, then don't hire them, you open the door to a lawsuit.

Testing works (a little), and better testing works a little bit more. And on and on. This is the machinery of progress, and it won't stop grinding.

Yes, progress. Rather then treating people like disposible, replaceable cogs, we can treat them like statistically clustered cogs, which we replace after a quick automated test. Who could be against that? It improves profitability and shareholder value at the expensive of teh poor and lower middle class and that's progress.
posted by delmoi at 12:39 AM on August 26

(And of course there is a difference between skills testing and personality testing, but still)
posted by delmoi at 12:40 AM on August 26

« Older Matthew Weigman,... | Fedor Emelianenko is the best ... Newer »

This thread has been archived and is closed to new comments

"Our system allows you to clone your best, most reliable people" August 24, 2009 11:39 PM Subscribe

"Our system allows you to clone your best, most reliable people"
August 24, 2009 11:39 PM Subscribe