Hard News by Russell Brown

Read Post

Hard News: The shaky ground of psychometric testing

154 Responses

First ←Older Page 1 2 3 4 5 7 Newer→ Last

  • Danielle, in reply to wasabicube,

    As an aside, it is interesting to observe the level of duplicity exhibited by some commenters on this subject. Clearly misrepresenting yourself to your prospective or current employer whether through a psychometric assessment, interview or otherwise is something that should be of concern.

    All of this stuff is just hoop-jumping. You know it, I know it, we all know it. The point is not how honestly you represent yourself, but how well you tailor your own representation to fit whatever the HR department is looking for. What would be the point of me saying "I think this is all time-wasting bullshit" to someone who has bought into this enough to sit there and present it with a straight face, playing their own little part? I'd never get hired.

    Charo World. Cuchi-cuchi!… • Since Nov 2006 • 3828 posts Report

  • Lyndon Hood,

    Of course, even if the tests worked, you'd need a reliable way to tell if the result you were looking for actually *would* be a good fit in your workplace.

    Wellington • Since Nov 2006 • 1115 posts Report

  • BenWilson, in reply to Danielle,

    What would be the point of me saying “I think this is all time-wasting bullshit” to someone who has bought into this enough to sit there and present it with a straight face, playing their own little part?

    It's important that we weed out the malcontents from our public libraries. Whenever I'm trying to find a book on 18th Century card tricks of the Ukraine and the librarian eventually gives up, I wonder what tests they cheated on to get where they are.

    Auckland • Since Nov 2006 • 10657 posts Report

  • Rich of Observationz,

    those with solid scientific studies to back their construction and use

    Got any references for that?

    Back in Wellington • Since Nov 2006 • 5550 posts Report

  • Bart Janssen, in reply to wasabicube,

    It’s also worth noting that most people’s ability to judge character is much worse than they credit themselves, especially in interview situations.

    This is very true. It took us several years and several disasters before we learnt which questions were most useful to determine if someone would be a disaster in the lab or not. You might think that measuring how bright or how much knowledge candidates had was most important in science but in reality building a team that are willing to work together is much more important. My partner and I are now pretty good at picking good ones, good enough that we have frequently face-palmed some of our colleagues choices and then been proven right months later.

    A big part of our interview process is the debrief afterwards where we compare notes and opinions and double and triple check ourselves. So we never rely on one persons judgement any more.

    Auckland • Since Nov 2006 • 4461 posts Report

  • Andre Alessi, in reply to Danielle,

    The point is not how honestly you represent yourself, but how well you tailor your own representation to fit whatever the HR department is looking for.

    And often HR is just the conduit. The bottom line here is that management see these tests as a way to validate things they think they already know about their employees and the business. If they don't get the answers they want, they see the "wrong" answers as a problem to be fixed. There's nothing perfidious about employees trying to avoid the entire painful process by responding in a way that keeps the higher ups happy.

    We can't all be Jean Luc Picard, insisting, "There are FOUR lights!"

    Devonport, New Zealand • Since Nov 2006 • 864 posts Report

  • Bart Janssen,

    Just a side comment. These tests typically present as binary questions ... they aren't. Each individual question is binary but the output depends on the accumulation of multiple questions. They can actually be informative.

    However, never (IMO) more informative than the opinion of a good manager who has genuine personal contact.

    Auckland • Since Nov 2006 • 4461 posts Report

  • wasabicube, in reply to Danielle,

    I’m not entirely sure it's just hoop-jumping though I admit we all like to put our best side forward at interview. My point is more towards deliberately gaming the personality assessment by attempting to project the personality you think is the one being sought. This involves second-guessing the mapping between questions and traits and the weights typically put on those responses. Competent assessment report interpreters can see through this behaviour in the results of the assessment and will check with referees for confirmation.

    Our focus has been on fit within organisations and teams and we feel we bring good success by seeking to understand the way people operate and tailoring their roles and management accordingly. It’s very rare to find the perfect candidate but understanding their motivations and behaviours can help produce a good outcome for both candidate and employer. This is starting to sound like marketing patter; I guess I’m playing my own little part with a straightish face.

    Whangaparaoa • Since Nov 2006 • 15 posts Report

  • Bart Janssen, in reply to Rich of Observationz,

    those with solid scientific studies to back their construction and use

    Got any references for that?

    Heh. When they first started using HR consultants with us to try and teach us leadership the consultant got very very tired of alway being asked for the primary references. She was very good humoured about it and did change her approach to one more grounded in "data first-conclusions second".

    Auckland • Since Nov 2006 • 4461 posts Report

  • BenWilson, in reply to wasabicube,

    Competent assessment report interpreters can see through this behaviour in the results of the assessment and will check with referees for confirmation.

    Heh, I've read that they always ask you about porn during a lie detector test, because you'll probably lie, and then they can get the baseline. It is actually possible to ask people questions that they perfectly legitimately do not want to answer honestly.

    Auckland • Since Nov 2006 • 10657 posts Report

  • BenWilson, in reply to Lyndon Hood,

    Of course, even if the tests worked, you’d need a reliable way to tell if the result you were looking for actually *would* be a good fit in your workplace.

    Indeed, there's quite a lot of work where having a duplicitous nature makes you the perfect fit. Management, for example, which is probably why they want to weed out the competitors :-)

    Auckland • Since Nov 2006 • 10657 posts Report

  • Stephen Judd, in reply to wasabicube,

    As an aside, it is interesting to observe the level of duplicity exhibited by some commenters on this subject. Clearly misrepresenting yourself to your prospective or current employer whether through a psychometric assessment, interview or otherwise is something that should be of concern.

    Not if you think that they’re actually useless bullshit. Especially in contexts where it’s the employer’s obviously clueless agent.

    Logically, on this basis, no one should ever follow advice on how to do well in employment interviews, as such advice invariably involves stifling your natural reactions to convey a false impression, sorry, I meant the best impression.

    I also don’t think we can overlook that in an economy with structural unemployment, in many cases potential employees need the employer much more than the employer needs them. Few among us are perfectly upright in the face of such strong incentives.

    Personally I’ve done one of these tests once. At the time, the employer swore that its purposes was so they could get a good balance of personalities on the team, rather that to select for a particular type. These days, I would regard reliance on testing as a red flag about a workplace and a sign of somewhere I don't want to be.

    Wellington • Since Nov 2006 • 3122 posts Report

  • tim oliver,

    The bit that leaves me boggled is, how the hell do you go from 'the candidate has personality type X according to a test' to 'will fit in in a department?'

    I mean let's say i'm outgoing and bubbly - what rigorous mechanism are you going to use to decide that this is suitable for the job? Is a psychologist actually testing the department and manager as well and using evidence to decide that type A and type C personalities get on well? Are there double-blind studies that test this?

    I mean all of this 'fit' talk seems to assume that the way to produce an effective team is to bring together a bunch of people with similar personalities. Why is that necessarily true?

    We all SUSPECT that what's actually happening is the hiring manager is being handed a test result in exchange for a pile of money, and they're just picking based on their own biases.

    In short, not getting a job because a dumb test thinks you have the wrong personality, and then being told this is a 'good outcome' for you, is the most infuriating thing I can imagine.

    Since May 2011 • 8 posts Report

  • Ross Mason,

    It’s good to know there are people out there concerned enough to ensure you are well prepared to....well...cheat these tests.

    Let me count the ways.

    1. Dress smartly. You are competing with other people; every little part of you that suggests you care and are taking this seriously will count in your favor.
    2. Have the right attitude. …expect to give the tests your absolute best shot by taking the right mental approach and being determined to work hard. [ Dude! ]
    3. Practice in advance…..Practice is the best chance that you have for succeeding, especially when you work with materials that are similar to those you'll be tested on.
    4. Broaden your test material. As well as using psychometric tests, find all sorts of relevant test material to broaden your knowledge and fast answering ability.
    5. Seek feedback. …While there are no wrong or right answers in personality tests, there can be indicators of areas in which you would benefit from self-improvement, such as training in ethics or assertiveness. [ ETHICS???? FFS ]

    Good to know our academic orgs (Auckland Uni) are right behind them!

    Why do employers use psychometric tests? Psychometric tests are objective and fair. The results are not influenced by tester bias……

    Yeah right.

    And gosh. That AKL link has some really close matching text of the one above. Plagiarism is rife…..somewhere around here. Who copied who over?

    Upper Hutt • Since Jun 2007 • 1590 posts Report

  • linger,

    Each year, I present my Applied Linguistics class with the Jung-Myer-Briggs test, and the “language learning styles” predicted for each of the 16 personality types. (One module of the course is factors influencing language learning success, including a range of – proposed, but with few exceptions not conclusively supported – psychological factor effects.)
    So far, I have observed that:
    (1) Most students score close to the midpoint on most scales. Which is to be expected: all four scales are normal distributions, rather than binary categories. This implies that most classifications using the test are arbitrary on at least one dimension. Basically, for any dimension(s) where you score 25% or less (= 2 questions away from the midpoint), you should also consider the neighbouring type description(s).
    (2) If I simply present each student with the learning style description corresponding to their personality type as scored, students tend to think their assigned learning style description accurately matches their own learning experience.
    But…
    (3) If I give each student three unlabelled learning style descriptions (one matching their type score, one for a type differing on 2 dimensions, and one for the complementary type, differing on all 4 dimensions) then their choices of the “best” learning style description are not significantly above chance. Which suggests that the learner descriptions (and possibly also at least some parts of the JMB test), may have little more use than a horoscope.

    That being said: the extrovert/introvert scale does at least have a clear genetic basis, and does seem to correlate to some extent with preferred learning styles.

    Tokyo • Since Apr 2007 • 1944 posts Report

  • Rich of Observationz, in reply to linger,

    I suppose you could teach a class in a manner heavily biased to one learning style and see if their results correlated with the predicted learning styles - e.g. the ones with a compatible learning style did better.

    Unethical, I guess.

    Back in Wellington • Since Nov 2006 • 5550 posts Report

  • linger, in reply to Rich of Observationz,

    Yep ... and impractical, since the reliability of the "learning style" assignment is itself very much in doubt.

    Tokyo • Since Apr 2007 • 1944 posts Report

  • David Hood, in reply to linger,

    Not only is the reliability of the "learning style" assignment in doubt, I have seen some pretty convincing evidence that "What person A wants to spend their time doing" (learning style) correlates very poorly with "How to actually improve the understanding of the topic" (learning outcome).

    Dunedin • Since May 2007 • 1445 posts Report

  • Lilith __, in reply to linger,

    (3) If I give each student three unlabelled learning style descriptions (one matching their type score, one for a type differing on 2 dimensions, and one for the complementary type, differing on all 4 dimensions) then their choices of the “best” learning style description are not significantly above chance.

    That's fascinating!

    Dunedin • Since Jul 2010 • 3895 posts Report

  • Russell Brown, in reply to bronwyn,

    The test pushed me from irritation to near-rage. Here were some of the questions: “Would you prefer to go to a party or read a book?” “When going to the movies, would you prefer to see a romantic comedy or an action film?” It was preposterous. When the agent came back and said “So, how did you find that?” I replied “Not much room for shades of grey in there, is there?” Her reply? “Yes, that’s the way we like it”.

    Sweet Jesus. I'd have been getting mighty uncivil by that point. Not just having to endure a plainly idiotic testing process, but then hearing that that's the way they intend it to be ...

    Clearly, I should not be employed by anyone, ever. Which does kinda suit me.

    Auckland • Since Nov 2006 • 22850 posts Report

  • BenWilson, in reply to tim oliver,

    In short, not getting a job because a dumb test thinks you have the wrong personality, and then being told this is a ‘good outcome’ for you, is the most infuriating thing I can imagine.

    It might be a good outcome if the boss was a dick.

    Auckland • Since Nov 2006 • 10657 posts Report

  • BenWilson, in reply to Lilith __,

    That’s fascinating!

    Yes, it's sort of goes to my point about people near the middle roving between opposites, but I'd never really thought through that most people are near the middle. So most people can switch reasonably easily between any of the 16 types! But perhaps it does have some merit when people are more extreme in one or other dimension.

    Auckland • Since Nov 2006 • 10657 posts Report

  • Martin Brown,

    Ah, the Briggs-Myers test. I've seen that gem of HR passive-agression used to start the covert constructive dismissal process a few times.

    Auckland • Since Mar 2013 • 137 posts Report

  • SteveH, in reply to Geoff Lealand,

    Those questions on the US immigration application? Draft as a coot! I always wonder if anyone ever answers ‘yes’ to any of them

    I'm not sure what the original intention for them was, but I think the primary use is to detect people who were not paying attention when they filled in the rest of the form.

    Since Sep 2009 • 444 posts Report

  • Martin Brown, in reply to Andre Alessi,

    Worked for a large multinational ad agency which used to do that "survey" across the entire group, each year. "Don't worry it's completely anonymous." They would also select "representative samples" of staff who would be asked to complete "confidential" reviews of specific execs or staff. It was all rather creepy and loaded. And compulsory.

    Auckland • Since Mar 2013 • 137 posts Report

First ←Older Page 1 2 3 4 5 7 Newer→ Last

Post your response…

This topic is closed.