I'm seeing personality tests and cognitive tests in tech now

[Ponderosa]

I'm new to applying to tech jobs (potential career change) and now I'm finding personality tests popping all over the place!

One such employer had a cognitive test as well. 50 challenging questions within 15 minutes -> which works out to ~18s a question..

WTF? Seriously? Not only do you have to do phone screens, now you have these tests on top of homework, and a battery of interviews. I had an interview with an employer and got to the final step. The night before they said a personality test and asked me take it. I told them no, since it was last minute and I had plans. Still showed up at the interview the next day, and it went well. I took the test afterwards and got a call next week denying me the job. I asked why, and they said I seem a bit driven and they think I'd leave the company soon.

No wonder why we all want to be FI.

[MilesTeg]

I'm new to applying to tech jobs (potential career change) and now I'm finding personality tests popping all over the place!

One such employer had a cognitive test as well. 50 challenging questions within 15 minutes -> which works out to ~18s a question..

WTF? Seriously? Not only do you have to do phone screens, now you have these tests on top of homework, and a battery of interviews. I had an interview with an employer and got to the final step. The night before they said a personality test and asked me take it. I told them no, since it was last minute and I had plans. Still showed up at the interview the next day, and it went well. I took the test afterwards and got a call next week denying me the job. I asked why, and they said I seem a bit driven and they think I'd leave the company soon.

No wonder why we all want to be FI.

There's a lot of garbage to sift through in the tech market when it comes to prospective employees. Bootcamps, googlers, etc.

[Bloop Bloop]

I'm not a fan of personality tests because I think they're very very pseudo-sciencey.

Cognitive tests, on the other hand, should be encouraged. IQ correlates highly with most measures of workplace performance. Americans have this weird "don't talk about the war" thing going on with IQ, probably because of your country's screwed up racial issues, so you don't commonly do testing in an employment context. Here in Australia cognitive testing is a common and accepted part of job hiring, and it's probably one of the easier ways of weeding out applicants. I'm not saying that cognitive performance is the only, or even primary, measure of an applicant, but for certain jobs it's a starting point.

[Buffaloski Boris]

This sort of silliness is why FI folks should do “side hustles” and run businesses as often as they can. 

[SunnyDays]

I'm not a fan of personality tests because I think they're very very pseudo-sciencey.

Cognitive tests, on the other hand, should be encouraged. IQ correlates highly with most measures of workplace performance. Americans have this weird "don't talk about the war" thing going on with IQ, probably because of your country's screwed up racial issues, so you don't commonly do testing in an employment context. Here in Australia cognitive testing is a common and accepted part of job hiring, and it's probably one of the easier ways of weeding out applicants. I'm not saying that cognitive performance is the only, or even primary, measure of an applicant, but for certain jobs it's a starting point.

I'm curious as to what cognitive tests are used.  I have an MA in psychology and am qualified to do IQ testing, and I wonder if this is a requirement of your testers as well.  If they aren't using standardized tests, then there's likely no validity to them.  In addition, simply arriving at a number isn't very useful - a full analysis of strengths and weaknesses and a proper profile of the subtests really is needed to get any meaningful information.  It can generally be assumed that anyone applying for a professional job has a substantial IQ, so testing just to confirm this is a waste of everyone's time.

[Ponderosa]

I'm not a fan of personality tests because I think they're very very pseudo-sciencey.

Cognitive tests, on the other hand, should be encouraged. IQ correlates highly with most measures of workplace performance. Americans have this weird "don't talk about the war" thing going on with IQ, probably because of your country's screwed up racial issues, so you don't commonly do testing in an employment context. Here in Australia cognitive testing is a common and accepted part of job hiring, and it's probably one of the easier ways of weeding out applicants. I'm not saying that cognitive performance is the only, or even primary, measure of an applicant, but for certain jobs it's a starting point.

I'd be interested to have you share more what you mean by the comment 'Americans have this weird..', not sure I understand.

In any case, I highly doubt IQ tests have anything to do with the job past a certain level. Once you have a baseline of IQ then any job is more expressed mastery of a particular skill than having extra points of IQ.

[Bloop Bloop]

I'm not a fan of personality tests because I think they're very very pseudo-sciencey.

Cognitive tests, on the other hand, should be encouraged. IQ correlates highly with most measures of workplace performance. Americans have this weird "don't talk about the war" thing going on with IQ, probably because of your country's screwed up racial issues, so you don't commonly do testing in an employment context. Here in Australia cognitive testing is a common and accepted part of job hiring, and it's probably one of the easier ways of weeding out applicants. I'm not saying that cognitive performance is the only, or even primary, measure of an applicant, but for certain jobs it's a starting point.

I'm curious as to what cognitive tests are used.  I have an MA in psychology and am qualified to do IQ testing, and I wonder if this is a requirement of your testers as well.  If they aren't using standardized tests, then there's likely no validity to them.  In addition, simply arriving at a number isn't very useful - a full analysis of strengths and weaknesses and a proper profile of the subtests really is needed to get any meaningful information.  It can generally be assumed that anyone applying for a professional job has a substantial IQ, so testing just to confirm this is a waste of everyone's time.

One I took had pictograms similar to Raven's Progressive Matrices, as well as basic problem solving word questions/maths questions (if you had two bundles of hay and one bundle had three more than the other....) as well as analogies as well as some spatial questions, under time pressure.

Saying that simply arriving at a number isn't useful is like saying that simply looking at a candidate's GPA isn't useful because it doesn't flesh out strengths and weaknesses. True, but as a culling strategy, test performance/GPA are the first two employed.

As for your last line - are you saying that it should be assumed that everyone who applies to be, say, a graduate accountant or lawyer or policy officer is fully capable of the cognitive demands of the job? Because that's the only scenario where "testing to confirm would be a waste of time". And even if everyone were fully capable, it might be that an employer wants those that are intellectually more capable than others, again, not as a final metric, but as a culling strategy.

[SunnyDays]

I'm not a fan of personality tests because I think they're very very pseudo-sciencey.

Cognitive tests, on the other hand, should be encouraged. IQ correlates highly with most measures of workplace performance. Americans have this weird "don't talk about the war" thing going on with IQ, probably because of your country's screwed up racial issues, so you don't commonly do testing in an employment context. Here in Australia cognitive testing is a common and accepted part of job hiring, and it's probably one of the easier ways of weeding out applicants. I'm not saying that cognitive performance is the only, or even primary, measure of an applicant, but for certain jobs it's a starting point.

I'm curious as to what cognitive tests are used.  I have an MA in psychology and am qualified to do IQ testing, and I wonder if this is a requirement of your testers as well.  If they aren't using standardized tests, then there's likely no validity to them.  In addition, simply arriving at a number isn't very useful - a full analysis of strengths and weaknesses and a proper profile of the subtests really is needed to get any meaningful information.  It can generally be assumed that anyone applying for a professional job has a substantial IQ, so testing just to confirm this is a waste of everyone's time.

One I took had pictograms similar to Raven's Progressive Matrices, as well as basic problem solving word questions/maths questions (if you had two bundles of hay and one bundle had three more than the other....) as well as analogies as well as some spatial questions, under time pressure.

Saying that simply arriving at a number isn't useful is like saying that simply looking at a candidate's GPA isn't useful because it doesn't flesh out strengths and weaknesses. True, but as a culling strategy, test performance/GPA are the first two employed.

As for your last line - are you saying that it should be assumed that everyone who applies to be, say, a graduate accountant or lawyer or policy officer is fully capable of the cognitive demands of the job? Because that's the only scenario where "testing to confirm would be a waste of time". And even if everyone were fully capable, it might be that an employer wants those that are intellectually more capable than others, again, not as a final metric, but as a culling strategy.

The tests you are describing are not formally recognized IQ tests, as far as I can tell.  "Pictograms similar to RPM" is NOT RPM, and the others sound like a homemade test, to be honest.  They aren't standardized, as in there are no norms to compare results to, so likely the tester is just comparing one applicant to another.  I would not give a lot of credence to the results of these kinds of questions.  Even with well-recognized IQ tests, such as the Weschlers, 2 people can have the same number, but a vastly different profile of abilities. 

As for capability, I would suggest that anyone who has the education to apply for the types of jobs you mention, likely would be capable of the demands of the job, intellectually.  There may be other aspects of their abilities (such as personality or emotional factors) that would make them less suitable, but these would not be picked up strictly in the "IQ" testing process.  People applying for professional level jobs most likely have higher IQs than the general population, and a difference of say 5 - 10 points is not going to affect job performance greatly.

I think that these types of tests simply give the illusion of an objective measure for decision-making.  It's easier to rely on a number when comparing candidates than more subjective measures, such as social skills, likeability, past job performance, etc. 

[Kyle Schuant]

In any case, I highly doubt IQ tests have anything to do with the job past a certain level. Once you have a baseline of IQ then any job is more expressed mastery of a particular skill than having extra points of IQ.

See, I've always thought that your performance on a job should be measured by... your performance on a job. Not a bunch of other stuff. But I'm a crazy radical in that way.

"What if they haven't done the job yet?" Well then of course they won't be competent, but they might have potential. How about you teach them?

"What if they don't have potential?" So... you're a manager, and you're not able to discern competence and potential? Um, why then are you the manager?

The real reason we have personality and cognition tests is because we have so many applicants for a job, and realistically at least 50 of the 100 would be as good as any other, but we need some arbitrary way to choose between them, and personality and cognition tests seem more legit than astrology.

https://www.nytimes.com/2017/04/08/opinion/sunday/the-utter-uselessness-of-job-interviews.html

[Bloop Bloop]

SunnyDays - I don't know if the tests are "formally recognised". They were made by some big company that contracts to large employers and government agencies to conduct applicant screening (written application, test, group interviews, in that order, before the final interviews are done by the individual employer).

"so likely the tester is just comparing one applicant to another."

Correct. The test raw results can't be used to calculate for example an SD or a bell curve.

However, the only point of the test is to compare one applicant to another.

And even if I were to accept that "As for capability, I would suggest that anyone who has the education to apply for the types of jobs you mention, likely would be capable of the demands of the job" (this is a vast and sweeping generalisation; someone might have a law degree with barely passing grades from Podunk College versus a first class honours degree from a top 10 university), there is still a utility in ranking candidates. Even if 20 people are smart enough for a job I might wish to interview the 10 smartest ones of the group.

Finally, no doubt other aspects like personality, drive, etc, are more important. However, they are impossible to measure quantitatively. They are in fact difficult to even measure qualitatively, and are subject to interview bias in a way that testing is not.

"a difference of say 5 - 10 points is not going to affect job performance greatly."
If it affects job performance at all as a correlate, it would be rational to use it as a screening tool.

Grades are also subject to variation, are rarely of the utmost importance, and say nothing about a candidate's soft skills. Nonetheless grades are used by all professional employers when hiring. Test performance is along the same lines.

[mozar]

My advice is to see a personality test as a red flag. HR clearly has no idea what they're doing. You're better off staying away from these companies. My guess is that they were testing for compliance and doing whatever you are told. That's why you failed, because you took the test when you felt like it.
I can understand testing math skills for a job where it's important but 90% of jobs could be done by a monkey, and not a particularly bright one.
I have a masters degree in accounting and I was a federal auditor for ten years. If I wasn't saving my money I would have dressed a monkey in a suit and sit it in my chair and nobody would have noticed. Ok, ok, I'm just making myself laugh now.

Anyways for my first job out of grad school I had to take a personality test. After I got the job they told me I had failed the test but I got the job because of where I went to grad school.

[Kyle Schuant]

Think of the surgeon who left their instruments in a patient, the cop who shot an unarmed person in their home while doing a welfare check, or the software engineer who caused the spacecraft into Mars because they didn't think to check whether they were using NASA miles or ESA kilometres.

They all passed stringent tests.

[scottish]

I'm new to applying to tech jobs (potential career change) and now I'm finding personality tests popping all over the place!

One such employer had a cognitive test as well. 50 challenging questions within 15 minutes -> which works out to ~18s a question..

WTF? Seriously? Not only do you have to do phone screens, now you have these tests on top of homework, and a battery of interviews. I had an interview with an employer and got to the final step. The night before they said a personality test and asked me take it. I told them no, since it was last minute and I had plans. Still showed up at the interview the next day, and it went well. I took the test afterwards and got a call next week denying me the job. I asked why, and they said I seem a bit driven and they think I'd leave the company soon.

No wonder why we all want to be FI.

There's a lot of garbage to sift through in the tech market when it comes to prospective employees. Bootcamps, googlers, etc.

There's also a lot of garbage to sift through in the tech market when it comes to prospective employers.   :-)

[mizzourah2006]

I'm not a fan of personality tests because I think they're very very pseudo-sciencey.

Cognitive tests, on the other hand, should be encouraged. IQ correlates highly with most measures of workplace performance. Americans have this weird "don't talk about the war" thing going on with IQ, probably because of your country's screwed up racial issues, so you don't commonly do testing in an employment context. Here in Australia cognitive testing is a common and accepted part of job hiring, and it's probably one of the easier ways of weeding out applicants. I'm not saying that cognitive performance is the only, or even primary, measure of an applicant, but for certain jobs it's a starting point.

I'm curious as to what cognitive tests are used.  I have an MA in psychology and am qualified to do IQ testing, and I wonder if this is a requirement of your testers as well.  If they aren't using standardized tests, then there's likely no validity to them.  In addition, simply arriving at a number isn't very useful - a full analysis of strengths and weaknesses and a proper profile of the subtests really is needed to get any meaningful information.  It can generally be assumed that anyone applying for a professional job has a substantial IQ, so testing just to confirm this is a waste of everyone's time.

I have a PhD in I/O psychology and most cognitive tests being used are variants of validated cognitive assessments, like Raven's matrices,  or proxies for cog given a job-specific task, like asking a candidate for a supervisor position to look at a couple of tables about employees requested days off, combined work performance, etc. and develop the most ideal schedule, etc. There are of course companies out there that are "gamifying" cognitive ability assessments or say they are, but in most cases, the variants are validated.

I'd also disagree with you on the thought that everyone applying for a professional job has a relatively high IQ. I've seen executives at fortune 50 companies that only get a 10 on the Raven's. Plus when you throw in the fact that so many people today inflate their resumes and all of the information out there about the coding challenges that each company does, etc. It's possible that someone with a relatively low IQ (for a tech position) could get through without encountering a novel question or scenario. IQ is extremely important to solving novel tasks/problems, which I'm certain any tech person will tell you is basically their job.
Now if you are talking about testing doctors then I'd agree with you, it's a waste of time because they all had to get high MCAT scores and graduate from medical school.

Also, with regards to norms...those are extremely easy to create within house. It's basic math. If I get 5k people applying across 10 companies in professional roles to take my "cog test" I can create a very stable norm group and say this person scored in the 75th% for professionals in similar roles. Let's not pretend that Raven's or Weschler's or the Wonderlic has some special rights to normative samples. You don't have to call them formal IQ tests, but they are indeed proxies for IQ just as the SAT or GRE is a proxy for IQ. Most studies show that the correlations between GRE scores and actual IQ tests are in the 0.8 range...in psychological measurement, that threshold typically meets the requirements to be considered a reliability estimate.

It's also extremely easy to collect performance data on 4-500 of those individuals and provide criterion validation evidence that scores on the cog test are correlated with job performance. I've routinely seen assessment batteries (that include cog measures, personality measures, work history measures, situational questions, etc.) that have uncorrected correlations with job performance approaching 0.5. In most cases these assessments aren't ideal for tech companies when it's fairly easy to weed out candidates based off of simple qualifications, but what about if you are hiring for a call center rep job and you have 150 apply all with similar qualifications...i.e. none. Your options are give them an assessment that's correlated with job performance to narrow it down to maybe 25-30 to review more closely...randomly choose some to look at, or interview them all.

[Bloop Bloop]

Wow, thank you Mizzourah for such an informative post.

I would be interested to see whether the opposition to workplace IQ tests is based on there being insufficient correlation between the test and job performance, or some other factor.

[Focus_on_the_fire]

Yep, I've seen them. I think places that use them want a shortcut (a human "decoder ring" if you will), so they don't have to talk/listen to someone about their strengths, weaknesses, and preferences, etc.

 

[SAR]

I'm not familiar with what the various companies are using to assess IQ, but Raven's progressive matrices are the most highly g-loaded of the IQ tests.

"g" is the common factor that subsumes all IQ tests, and it is g that is most predictive of job performance, salary earned, and prestige of occupation (correlations in the .5 range). G also predicts longevity and (very mildly) sperm quality, humor ability (and, in turn, reproductive success) consistent with the hypothesis that g is a biological fitness indicator.

I would think a reputable company is doing something equally as valid as Ravens, so if I were an employer, I would very much want to know what the scores of a prospective employee is.

As for personality, measurement of the Big 5 has being going on for sometime and it has been validated, but the correlations with job performance are weak (from memory, conscientiousness is the most important there).

But beyond that, you are going to spend many thousands of hours of your life with people at work. Would you, for example, want to work with someone who is neurotic, disagreeable, and who lacks conscientiousness? Not me. Again, not a strong predictive validity, more of a screen for red flags.

[Bloop Bloop]

Yep, I've seen them. I think places that use them want a shortcut (a human "decoder ring" if you will), so they don't have to talk/listen to someone about their strengths, weaknesses, and preferences, etc.

I think the point is that if you have 20 applicants for a position or a series of position, it might be helpful to screen out, say, 15 of them via grades/cognitive performance, so that you only have to interview five of them about strengths, weaknesses, preferences etc.

[thesis]

Finding good people can be hard. Finding good people that fit your company's culture can be even harder. So I understand that companies are fearful of hiring, especially in tech where a bad hire can cost a great deal. However, I do think there's a lot of BS in the process, and this probably has something to do with trends. It was extremely popular not too long ago to have programmers sit down and do coding tests (this is of course still around), but there's power in being in demand - many of the best programmers simply refuse to play the game.

I'm not marking myself as one of those, but when I was last looking for a job, I was super excited to finally have the skills that put me in a position for several companies to be interested in hiring me. One job got back to me but then wanted me to do this long programming test online. I had experience with those and knew they could easily take several hours to complete. After putting it off and worrying about missing a key opportunity, I realized that other companies that didn't require a test were also interested. So I simply decided to pass on that company. And I'm glad I did, because I like where I'm at now.

I think it may be more stringent the less experience you have because the quality of junior devs is so variable (okay, to be fair, the quality of devs in general is also variable...), and they may really need to assure that you can program. There are lots of posers out there. But if you have 10 years of experience and know what you're doing, you can laugh at any test you want and walk out, and I think a lot of companies know that. That's where I hope to be in a few years :)

My current and previous company really just screened by having you talk about various frameworks/APIs and answer a few questions. These are great, I think, as you can tell whether somebody knows what they're talking about by hearing them discuss it, though I'm sure some people can BS that, too. But I actually learned a few things from this screen and had fun in the conversation. So I was pretty excited when I got the job offer a day later.

[Focus_on_the_fire]

Yep, I've seen them. I think places that use them want a shortcut (a human "decoder ring" if you will), so they don't have to talk/listen to someone about their strengths, weaknesses, and preferences, etc.

I think the point is that if you have 20 applicants for a position or a series of position, it might be helpful to screen out, say, 15 of them via grades/cognitive performance, so that you only have to interview five of them about strengths, weaknesses, preferences etc.

Based on my experience, I have two issues with the tests. First, candidates completed the tests outside of a testing environment, so there was no guarantee that they were an accurate representation. This is why I think transcripts are better.

Second, the test attempts to identify personality characteristics/preferences. That's beyond knowledge and a much more difficult area to assess under the best of circumstances.

[Bloop Bloop]

I agree with your first point. There's nothing stopping an applicant from getting a smart friend to sit alongside and help with the test.

As for the second point, only some companies use psychometric testing (and I think that has little if any validity). Most I've seen focus on the cognitive part.

[partdopy]

I like these tests, it shows me what companies to avoid and what companies allow their HR department too much leeway in hiring when the decision to hire/not hire technical slots should be made by the team hiring.

[ender]

I had one like this for my current job.

I don't think it was used at all in the actual evaluation process which was kind of silly to me

[John Galt incarnate!]

Cognitive tests, on the other hand, should be encouraged. IQ correlates highly with most measures of workplace performance. Americans have this weird "don't talk about the war" thing going on with IQ, probably because of your country's screwed up racial issues, so you don't commonly do testing in an employment context. Here in Australia cognitive testing is a common and accepted part of job hiring, and it's probably one of the easier ways of weeding out applicants. I'm not saying that cognitive performance is the only, or even primary, measure of an applicant, but for certain jobs it's a starting point.

"IQ is strongly related, probably more so than any other single measurable human trait, to many important educational, occupational, economic, and social outcomes ... Whatever IQ tests measure, it is of great practical and social importance." Linda Gottfredson

 I agree with you and Gottfredson's overarching conclusion  in "Main Stream Science on Intelligence," published in the Wall Street Journal on December 13, 1994.

The rationality and usefulness of cognitive evaluation to screen  prospective employees is so  beneficial  that it ought to be standard hiring practice  in the United States.

For this reason I disagree with the Supreme Court's unanimous opinion in Griggs v. Duke Power (1971).

[MNBEN87]

I had to take a cognitive test for a job offer I just got.  There was a lot of comparing shapes, and stuff like that.  I do well on this type of stuff, but have a hard time focusing on more routine work (I am an engineer). 

I think I performed better in college, compared to as a professional.  College tests I could really focus on as the pressure was on, but work is just so boring.  I think someone less intelligent, but that was challenged by the work (and engaged) would do a better job than I do. 

New job sounds much more engaging, so let's hope I focus.

[MNBEN87]

@Kyle Schuant

How much you deadlifting in your profile pic?  Respect to anyone who deadlifts!

[Bloop Bloop]

Cognitive tests, on the other hand, should be encouraged. IQ correlates highly with most measures of workplace performance. Americans have this weird "don't talk about the war" thing going on with IQ, probably because of your country's screwed up racial issues, so you don't commonly do testing in an employment context. Here in Australia cognitive testing is a common and accepted part of job hiring, and it's probably one of the easier ways of weeding out applicants. I'm not saying that cognitive performance is the only, or even primary, measure of an applicant, but for certain jobs it's a starting point.

"IQ is strongly related, probably more so than any other single measurable human trait, to many important educational, occupational, economic, and social outcomes ... Whatever IQ tests measure, it is of great practical and social importance." Linda Gottfredson

 I agree with you and Gottfredson's overarching conclusion  in "Main Stream Science on Intelligence," published in the Wall Street Journal on December 13, 1994.

The rationality and usefulness of cognitive evaluation to screen  prospective employees is so  beneficial  that it ought to be standard hiring practice  in the United States.

For this reason I disagree with the Supreme Court's unanimous opinion in Griggs v. Duke Power (1971).

That's what happens when your country has screwed up race relations. It infiltrates every other aspect of your society and it prevents you from having constructive discussions on anything that might touch on (or even not touch on) race.

[Kyle Schuant]

How much you deadlifting in your profile pic?  Respect to anyone who deadlifts!

That's actually from a few years ago, I think it was 180kg. I do more cardio now, I'm 48 and have to take care of that too.

Deadlifts should be part of job interviews. It'd be no more arbitrary and pointless than most of what goes on. Someone I know is up to her fourth interview with a company, they still haven't decided. You'd think they were hiring for someone to become God. HR's main job, apart from amplifying minor disagreements and offences into full-blow lawsuits, appears to be coming up with new and different ways to fuck around potential employees.

[BradminOxt19]

How much you deadlifting in your profile pic?  Respect to anyone who deadlifts!

That's actually from a few years ago, I think it was 180kg. I do more cardio now, I'm 48 and have to take care of that too.

Deadlifts should be part of job interviews. It'd be no more arbitrary and pointless than most of what goes on. Someone I know is up to her fourth interview with a company, they still haven't decided. You'd think they were hiring for someone to become God. HR's main job, apart from amplifying minor disagreements and offences into full-blow lawsuits, appears to be coming up with new and different ways to fuck around potential employees.

Hear hear!!

I agree, personality tests and cognitive tests end up producing robots.  A company's greatest strength may be the diversity of its work force, instead of a bunch of similar robots with the same values / beliefs / education producing monotone work for a narrow slice of the population.

Look at the most successful people in the world, they all have a variety of backgrounds, experiences, values, and philosophies and yet they all managed to find niches to enable them to make a difference in the world and make lots of money.

[Bloop Bloop]

Unless you think higher IQ correlates with higher robot traits, it doesn't follow that IQ testing would produce robots. It just produces high-IQ people. You can select for diversity among the high-IQ people. When you are a hot start-up or a large company to which dozens or hundreds of applicants apply for a handful of spots, you need to make culling decisions fast, and IQ is probably one of the less arbitrary ones, in that it at least serves a function.

[bluebelle]

Unless you think higher IQ correlates with higher robot traits, it doesn't follow that IQ testing would produce robots. It just produces high-IQ people. You can select for diversity among the high-IQ people. When you are a hot start-up or a large company to which dozens or hundreds of applicants apply for a handful of spots, you need to make culling decisions fast, and IQ is probably one of the less arbitrary ones, in that it at least serves a function.

Rather than pure IQ tests, I'd like to see some basic reading comprehension tests done, which I'm sure would be considered discriminatory against any English as second language candidates.   And I don't want that, but I need to work with fellow IT geeks that read and understand emails, requirements etc.  I'd be happy if folks could read and write simple bullet form text, I don't want literary prose.   I spend too much of my day trying to figure out what the hell someone meant in their emails or responding that the answer is in the text below (and no it's not me, 90% of my co-workers understand me the first time)

I like the idea of cognitive tests, but they'd have to be appropriate for the job.   I like questions about how you'd solve a problem, regardless of whether you come up with the 'right' answer, it's seeing a thought process - but that's an interview question, not a multiple guess quiz.   Although, the cynic in my thinks that some companies use the 'how would you solve this' type question to get answers they don't have (personal experience and rumour mill).

For most technical aspects of the job, you can teach someone syntax and functionality, you can't teach personality, and cognitive ability.   I am referring IT type jobs, obviously I want my brain surgeon or car mechanic to already know the job.

[trollwithamustache]

meh, don't get to worked up over it. big companies have always had silly HR requirements. The Tech companies seem a bit more fad driven than others.   

I'm going to disagree with the poster who said they wouldn't work for this company... I suspect HR is screening out a lot of people this way, probably too many. So if via networking one can get to a hiring manager at one of these companies for a role that makes sense for them, there is likely very little competition!
 

[Psychstache]

Unless you think higher IQ correlates with higher robot traits, it doesn't follow that IQ testing would produce robots. It just produces high-IQ people. You can select for diversity among the high-IQ people. When you are a hot start-up or a large company to which dozens or hundreds of applicants apply for a handful of spots, you need to make culling decisions fast, and IQ is probably one of the less arbitrary ones, in that it at least serves a function.

If you are giving actual IQ tests (norm-referenced assessments given by a trained examiner in a controlled environment with standardized administration), you are not making a fast (or cost-effective) decision.

-Someone who gave IQ tests for a living for about a decade

[EscapedApe]

Deadlifts should be part of job interviews. It'd be no more arbitrary and pointless than most of what goes on.

Hell, I think that's a fantastic metric. Deadlifts test your discipline, your determination, and your tolerance for discomfort - all important qualities in a challenging workplace environment.

I say add them to the interview process.

[Kyle Schuant]

Unless you think higher IQ correlates with higher robot traits, it doesn't follow that IQ testing would produce robots.

No; learning to give people the answers they want to hear produces robots. This applies even when the answer they want to hear is technically difficult to figure out. If your recruitment process demands grovelling conformity, then you will get grovelling conformists as employees. Which is of course the goal of many managers. Assertive original people are a pain in the arse to deal with. Grovelling conformists ensure a minimum of interruptions to the smooth flow of emails, KPI reports and powerpoint presentations.

Hell, I think that's a fantastic metric. Deadlifts test your discipline, your determination, and your tolerance for discomfort - all important qualities in a challenging workplace environment.

I say add them to the interview process.

Many people say the purpose of selecting for degrees when in truth the day-to-day job doesn't require a tertiary-level education (for example, bank tellers) is to get people who have demonstrated they were able to focus and work hard in an organised and self-directed way for a few years. A deadlift would do the same thing.

As well, a decent deadlift would mean the person is less likely to end up on WorkCover from having to load the photocopier with paper, or something.

But in reality it would be arbitrary and silly. Just as IQ tests are (particularly when degrees are required, since a degree will filter out anyone who's particularly stupid). It's simply that with 100 applicants for 1 position, at least 20 of them will be able to do a decent job, and in truth you could just choose randomly and get similar outcomes.

This is particularly so at companies with hundreds or thousands of employees, where one person's fuckups are always going to be picked up by someone else, and any brilliance they have will be filtered through and damped down by a dozen committees and a hundred meetings, so that everyone's performance in the end trends to mediocre.

But choosing randomly means that we don't need as many people in HR, and middle managers have less to do. And so some silly arbitrary methods are employed to give the illusion that what is in reality a process of random selection actually has solid reasoning and assessment behind it, and that this process is important and useful and fair. The "fair" part is important in these days of diversity managers and the like.

[GuitarStv]

The last time I was searching for a job I ran across several personality tests, and a couple (completely unrelated to the job IQ tests).  It was weird.  I get testing people for the skills you need for the position . . . but the personality/"IQ" tests seemed kinda capricious.

[Bloop Bloop]

To suggest that a degree should tell you all you need to know about intelligence is a bit misleading. Even if a degree might assure you that the person has a decent level of academic ability, it doesn't measure fluid intelligence (though the two are correlated, they are not the same). Further, even if a degree were an assurance of meeting an intelligence threshold enabling an employee to do a task, you might still want to select for a higher level of intelligence. All things being equal, I'd rather have a more intelligent employee than a less intelligent one, even if both were smart enough to do the fundamentals of the job. Of course I'd have to make sure the more intelligent person had social skills, had motivation for the job, etc, which is what the group interview and final interview are for.

[John Galt incarnate!]

All things being equal, I'd rather have a more intelligent employee than a less intelligent one, even if both were smart enough to do the fundamentals of the job.

Suppose an employer seeks an all-around office manager who is an excellent typist.

Should they hire an applicant who is an excellent  typist  (IQ = 100) or one who is a mediocre typist  (IQ =120)?

The employer  should hire the higher-IQ applicant because given their higher IQ they can easily learn to become an excellent typist as well as be a more useful  office manager.

^

IIRC, Charles Murray, co-author of The Bell Curve, used this hypothetical to illustrate the inherent value of higher-IQ employees.

[Kyle Schuant]

To suggest that a degree should tell you all you need to know about intelligence is a bit misleading.

Which would be why I didn't say it.

One of the things a degree should teach you is to read carefully, and contend with what is actually said, rather than some other stuff you made up.

I said, "a degree will filter out anyone who's particularly stupid". Not the merely stupid, obviously, but the particularly stupid. I also noted that many jobs don't actually require a degree-level of education for their day-to-day work. So: filter out the particularly stupid, then look at work-relevant skills. When you start talking about "fluid intelligence" and the like, you're talking about general traits not specifically relevant to the job, traits which should make the person good at many different jobs, but not necessarily good at this particular job.

In other words, if you are assessing things like "intelligence" (however you choose to define it), what you're really saying is, "we don't know how to assess your ability to perform work-related tasks, so we're just going to assess this arbitrary general thing instead and hope for the best."

In other words, you may just as well test their deadlift.


But, it helps keep HR and middle managers busy, I suppose.

[Bloop Bloop]

I would agree that fluid intelligence measures "traits which should make the person good at many different jobs, but not necessarily good at this particular job." The key there is 'necessarily'. There is still a correlation, for jobs which require some level of reasoning and which aren't pure physical labour.

So, you are right to say that for, say, a lawyer a better measure (ie a more closely correlated measure) than fluid intelligence might be how well the candidate does in a written exercise or a mock trial.

However, that does not mean that fluid intelligence doesn't have a correlation. And to the extent it has a correlation, it has a use. Sort the 100 applicants by fluid intelligence first, and then take the top 20 and give them a written exercise or mock trial to do.

When you say: "In other words, if you are assessing things like "intelligence" (however you choose to define it), what you're really saying is, "we don't know how to assess your ability to perform work-related tasks, so we're just going to assess this arbitrary general thing instead and hope for the best." you are committing the fallacy of thinking that there can only be one round of testing. There can be several. First test intelligence and then test more job-specific measures.

Secondly, if there is a correlation between fluid intelligence and traits which should make the person good at many different jobs, then it's not arbitrary, is it? The height of your argument is that it might be less useful than a more specific evaluation.

[GuitarStv]

Can you explain the research that shows the intelligence test you're using actually gives a valid indication of intelligence?  I might be way off base here, but my understanding was that determining intelligence is pretty tricky to do, and it's easy to create a test that will misrepresent intelligence levels.

[EscapedApe]

Can you explain the research that shows the intelligence test you're using actually gives a valid indication of intelligence?  I might be way off base here, but my understanding was that determining intelligence is pretty tricky to do, and it's easy to create a test that will misrepresent intelligence levels.

Just so.

Intelligence is difficult thing to define, much less measure.

Is it the ability to learn quickly?

Is it the ability to learn difficult material? If so, how does one define what is "difficult"? By comparing against an average performance outcome in terms of time and resources used? (If so, then the definition of intelligence becomes fluid).

Is it the ability to take learned information and apply it "effectively"? If so, how does one measure that effect?

It costs resources (time, money, and effort) to know something. We use inaccurate, broad-stroke criteria (things like stereotyping) where those criteria produce noticeable desired results, and only invest in more expensive ways of knowing when the benefit-to-cost ratio is high enough.

[PeterParker]

I can't believe the IO Psych PhD is saying that many of these various "IQ test proxies" are scientifically validated --- uh, validated to what? As predictors of job performance?

The SAT and GRE -- well, perhaps that are correlated to IQ (as measured by...?) by a factor of 0.8 ---- but that definitely doesn't sound quite right.

I've achieved a perfect score in both the SAT and GRE --- and I'm fairly smart, but I'm not Stephen Hawking, either.
The GRE math is even "lower" than the SAT. It goes up to Algebra II maybe? And algebra ability/ speed is based on knowledge and practice --- not necessarily IQ.

Before I took the test, I (having not looked at algebra or geometry in years) --- did a ton of practice tests and booklets, etc as drills. Again, refreshing knowledge and priming common situations (practice). Did my IQ increase from ordering these $10 booklets off of Amazon? Of course not. Yet they definitely improved my score.

And just because something has a normal distribution doesn't mean it's more scientifically valid --- I would expect pretty much any test to have a normal distribution by default --- it's usually other uncommon features that cause a skew/ bimodal distribution or something else.

As always, there are companies making MILLIONS off of rehashing either IQ/ cognitive tests .... or worse, "Personality tests" --- which usually are just sloppy copies of Myers Briggs. "Personality Poker" or "Insights Training" or any 4-Quadrant test that couches itself in many scientific terms (like Scientology) but are ultimately psuedoscience.

I mean they open up good discussions, but so did Sigmund Freud, and most of his theories were debunked by modern social science.

The Big 5 Personality traits are really the most scientifically tested "personality traits" in psychology, but ironically, I rarely see any corporate personality test try to establish these (other than extroversion-introversion, because its easily understandable and lends credence to the other Mickey Mouse tests).

Yes, tests can be a good idea, but sometimes I wonder.

[mizzourah2006]

I can't believe the IO Psych PhD is saying that many of these various "IQ test proxies" are scientifically validated --- uh, validated to what? As predictors of job performance?

The SAT and GRE -- well, perhaps that are correlated to IQ (as measured by...?) by a factor of 0.8 ---- but that definitely doesn't sound quite right.

I've achieved a perfect score in both the SAT and GRE --- and I'm fairly smart, but I'm not Stephen Hawking, either.
The GRE math is even "lower" than the SAT. It goes up to Algebra II maybe? And algebra ability/ speed is based on knowledge and practice --- not necessarily IQ.

Before I took the test, I (having not looked at algebra or geometry in years) --- did a ton of practice tests and booklets, etc as drills. Again, refreshing knowledge and priming common situations (practice). Did my IQ increase from ordering these $10 booklets off of Amazon? Of course not. Yet they definitely improved my score.

And just because something has a normal distribution doesn't mean it's more scientifically valid --- I would expect pretty much any test to have a normal distribution by default --- it's usually other uncommon features that cause a skew/ bimodal distribution or something else.

As always, there are companies making MILLIONS off of rehashing either IQ/ cognitive tests .... or worse, "Personality tests" --- which usually are just sloppy copies of Myers Briggs. "Personality Poker" or "Insights Training" or any 4-Quadrant test that couches itself in many scientific terms (like Scientology) but are ultimately psuedoscience.

I mean they open up good discussions, but so did Sigmund Freud, and most of his theories were debunked by modern social science.

The Big 5 Personality traits are really the most scientifically tested "personality traits" in psychology, but ironically, I rarely see any corporate personality test try to establish these (other than extroversion-introversion, because its easily understandable and lends credence to the other Mickey Mouse tests).

Yes, tests can be a good idea, but sometimes I wonder.

1. Scientifically validated could mean any # of things. There are several forms of validation. Construct validation, content validation, criterion validation, etc. so if you want to argue with me be specific with your terminology. Most assessments used by employers that are concerned about being sued (anyone I’ve ever worked with) have studies demonstrating both construct and criterion related validity, meaning they are highly correlated with other measures that are demonstrated to measure the “construct” they purport to measure, I.e. a known IQ test like Raven’s or the like and the “assessment” they are giving and that they are predictive of work place outcomes like job performance. So...yes most assessments being given are “scientifically validated”. I’m a scientist and I perform 20+ of these validation studies per year for specific clients.

2. Just do a quick Google search and you’ll see the correlations between aptitude tests and g. But...you’re right...we should use your N=1 over all the research out there.

https://www.psychologicalscience.org/pdf/ps/Frey.pdf?origin=publication_detail

Taken straight from the abstract:

“These studies indicate that the SAT is mainly a test of g.”

Then there’s this...

“Research suggests that the SAT, widely used in college admissions, is primarily a measure of g. A correlation of .82 has been found between g scores computed from an IQ test battery and SAT scores.”

“According to research by Robert L. Thorndike, 80 to 90 percent of the predictable variance in scholastic performance is due to g, with the rest attributed to non-g factors measured by IQ and other tests.”

https://en.m.wikipedia.org/wiki/G_factor_(psychometrics)

3. Most personality tests are using variants of the big 5...but remember our job isn’t just to map someone’s general personality traits onto a common continuum...our job is to predict workplace performance and often times other personality traits are more predictive, so in addition to measuring conscientiousness we may measure self-efficacy or following authority in an hourly type role. If we can provide statistical evidence traits other than the big 5 differentiate people who are high performers from those that are low performers in these roles why wouldn’t we use them?

Regardless of whether or not you consider it scientific,the “validation” evidence you are requesting is there in every assessment I’ve ever built. I can’t speak for everyone in the industry, all I can speak to is who I’ve worked with, which is a large selection of the fortune 500s, I’ve also worked internally at a fortune 50 and we performed validation studies on anything we ever used.

[Kyle Schuant]

See? So much work involved in arbitrarily choosing from the 20 out of 100 applicants who could do the job. Keeps people busy for entire careers! "Fruitful, productive work," as Sir Humphrey would say.

[Bloop Bloop]

Maybe instead of focussing on the HR lackeys who create a career out of it, focus on the feel-good stories of the employees who are good at taking such tests, who use that fluid intelligence to get scholarships, promotions and eventually get paid handsomely and end up achieving FIRE like all of the rest of us on these forums.

If that's not a feel-good story I don't know what is.

[mizzourah2006]

See? So much work involved in arbitrarily choosing from the 20 out of 100 applicants who could do the job. Keeps people busy for entire careers! "Fruitful, productive work," as Sir Humphrey would say.

So your answer to the problem hiring managers have is to just....pick out of a hat. I’m a Data Scientist who just interviewed with a professional baseball team for a senior data scientist role. I’ll go ahead and tell them that’s how they should handle the draft....well if they played college ball they can all “do the job”...so just pick one. No need to use data and projections. They can all “do the job”.

As an aside...I hope you’re at least pulling 2x BW in that pic or you probably shouldn’t own a strength training gym. But...I guess anyone that can perform the main powerlifts is capable of training people...isn’t that right?

Maybe instead of focussing on the HR lackeys who create a career out of it, focus on the feel-good stories of the employees who are good at taking such tests, who use that fluid intelligence to get scholarships, promotions and eventually get paid handsomely and end up achieving FIRE like all of the rest of us on these forums.

If that's not a feel-good story I don't know what is.

I laugh when people consider me an "HR lackey"....I guess I've fooled all these companies...because I'm an in-demand data scientist that commands $200k+/yr and I work from home in a low cost of living area.  Sure sucks to me :)

[SunnyDays]

This whole discussion reminds me of something one of my psych profs said in 3rd year, when everyone was stressing out about getting into grad school.  Apparently, someone made an error one year and offered admission to the master's program to the wrong person, who then turned out to be one of the best students they'd ever had.  Just goes to show .....

[mizzourah2006]

This whole discussion reminds me of something one of my psych profs said in 3rd year, when everyone was stressing out about getting into grad school.  Apparently, someone made an error one year and offered admission to the master's program to the wrong person, who then turned out to be one of the best students they'd ever had.  Just goes to show .....

oh for sure. There's clearly measurement error. That's why the best use case is in large volume roles. The probability of an assessment identifying the best person out of 5 is pretty slim, even with a correlation coefficient of 0.5. A correlation of 0.5 really only demonstrates that 25% of the variance in job performance can be explained by the variation in the predictor.

The way I've always explained it to a laymen is like this; You have two bowls of red and green marbles. If you select a green marble you have a good hire, if you select a red marble you have a bad hire. Would you rather pick from a bowl that was 50/50 or one that was 70/30 in favor of green marbles?

[GuitarStv]

This whole discussion reminds me of something one of my psych profs said in 3rd year, when everyone was stressing out about getting into grad school.  Apparently, someone made an error one year and offered admission to the master's program to the wrong person, who then turned out to be one of the best students they'd ever had.  Just goes to show .....

oh for sure. There's clearly measurement error. That's why the best use case is in large volume roles. The probability of an assessment identifying the best person out of 5 is pretty slim, even with a correlation coefficient of 0.5. A correlation of 0.5 really only demonstrates that 25% of the variance in job performance can be explained by the variation in the predictor.

The way I've always explained it to a laymen is like this; You have two bowls of red and green marbles. If you select a green marble you have a good hire, if you select a red marble you have a bad hire. Would you rather pick from a bowl that was 50/50 or one that was 70/30 in favor of green marbles?

There's certainly the assumption that they do, but I'm not entirely convinced that personality and IQ tests give you a 70/30 bowl.