cesare
shady's dreams ♥
Any examples of unions doing this?
Psychometric Testing
- Thread starter Yelkcub
- Start date
Any examples of unions doing this?
Onket
Je suis [CONTENT REMOVED]
They are rarely used properly, and that's the point.
Onket
Je suis [CONTENT REMOVED]
We did it in various re-structure consultations here. They dropped their use for any positions below a certain pay grade.
cesare
shady's dreams ♥
That's simply not the case. The BPS has strict guidelines which I've given you a source for, and they investigate complaints.
Onket
Je suis [CONTENT REMOVED]
cesare
shady's dreams ♥
What were the objections based on? The type of test? the brand of test? The administration of test? The interpretation of the test? The confidentiality and use of test results? The structure of the test? The cost v benefits aspect? What?
cesare
shady's dreams ♥
Better to just resist their use in the first place & not have to go through all that.
Any union worth their salt would resist interview only based selection
kabbes
First intersubjective, then internalised
FTR, in my last job I was involved in graduate recruitment. We had a whole recruitment day. We brought in a team of employment psychologists, who administered psychological testing. This was a small but crucial part of the whole day. The psychologists would not reveal what was discussed in their testing, but instead would only give their interpreted analysis of it. This was taken within the context of all the other testing that had occurred.
We were taking on graduates to be actuarial students, who would absorb a huge amount of cost and time within their first two years, so it isn't something you want to get wrong. We found the psychologists' input very informative.
Tellingly, after starting the regime as a whole, we thereafter got a 100% acceptance rate for the roles offered and a 0% drop-out of students within the first three years. That compares pre-regime to a high drop-out rate and a number of refusals.
We were taking on graduates to be actuarial students, who would absorb a huge amount of cost and time within their first two years, so it isn't something you want to get wrong. We found the psychologists' input very informative.
Tellingly, after starting the regime as a whole, we thereafter got a 100% acceptance rate for the roles offered and a 0% drop-out of students within the first three years. That compares pre-regime to a high drop-out rate and a number of refusals.
Onket
Je suis [CONTENT REMOVED]
Any union worth their salt would resist interview only based selection
Where have I said it was interview only?!
And if you really want the info, I'll have to get back to you on the objections, they were from before I came to this dept.
kabbes
First intersubjective, then internalised
Definitely. Interviews are pretty much the worst possible option. They introduce all kinds of systemic biases. They're also the death knell for diversity in the workforce.Any union worth their salt would resist interview only based selection
This has been studied in depth by a number of management institutions. You look at rates of retention and the success of candidates that have come from different methods of recruitment.
Pure interviews are a terrible, terrible method of assessment.
cesare
shady's dreams ♥
Where have I said it was interview only?!
And if you really want the info, I'll have to get back to you on the objections, they were from before I came to this dept.
I'm more interested in how you've formed this opinion tbh. If you don't know what the union's objections were, then what are you basing your own on?
kabbes
First intersubjective, then internalised
I still haven't got a straightforward answer from you as to whether the personality of the candidate matters as well as their experience and knowledge.
cesare
shady's dreams ♥
*looks for image of grandmother sucking egg
*Yep, it's a field I'm not entirely unfamiliar with.
kabbes
First intersubjective, then internalised
Now now, you should know that I wasn't talking to you. I was addressing the wider audience. Playing to the gallery. It's what I do best.*looks for image of grandmother sucking egg
Yep, it's a field I'm not entirely unfamiliar with.
Onket
Je suis [CONTENT REMOVED]
What I was told at the time when the subject came up in the latest re-structure. Because the use of the tests this time round were only for certain pay grades, we didn't have to go into it in depth.
The fact that the employer agreed with the union last & dropped their use supported the unions views.
Onket
Je suis [CONTENT REMOVED]
As I said, a decent interviewer will pick this up in an interview situation, and to a certain extent from the application. It matters, and those situations will give enough of an insight.
Going on what you've written about your graduate programme, I think we're slightly at crossed purposes anyway. We were involved with resisting them being brought in for fairly low paid office based admin type jobs at the council.
kabbes
First intersubjective, then internalised
Gosh, but you're being slippery.
So you are agreeing that it matters then. You said "It matters", which I think means that it matters.
So that's something. And maybe we can drop the "only experience and knowledge matters" red herring.
Because it really just comes down to what is the best way of assessing the personality. You are claiming that interviews are the best way of picking it up. I disagree, because interviews are absolutely rife with all kinds of endemic bias. They place all the power of personality assessment in the hands of one or two individuals, who might just happen to have a personality clash with a perfectly reasonable candidate. Or a nervous candidate that presents themselves negatively simply because they are nervous.
I'm not saying that psychometric testing is always appropriate. Or even that it is often appropriate. It should only be used, for example, where you have pre-identified the PRECISE personality issues that need to be avoided or selected for. If you don't have a professional reason for avoiding or selecting such issues then there is no point in using the tests.
And I have no doubt that they can be and often are misused. People are, fundamentally, rubbish. And they can easily carry the same biases that would make them fuck up being an interviewer into the interpretation of results, in order to get the answer they first thought of.
None of that means that psychometric testing should be dismissed out of hand, though.
Onket
Je suis [CONTENT REMOVED]
cesare
shady's dreams ♥
Now now, you should know that I wasn't talking to you. I was addressing the wider audience. Playing to the gallery. It's what I do best.
cesare
shady's dreams ♥
Not necessarily. They may have conceded that in the interests of industrial relations, or even for a more significant concession in return.
Brainaddict
slight system overdrive
OUt of interest, what are the preferred methods of selection that come out of these studies?
cesare
shady's dreams ♥
Not necessarily. Who knows why they were conceded. Pure conjecture. But they were. In favour of what? Interview only based selection?
kabbes
First intersubjective, then internalised
A combination of approaches massively beats any single approach.
This combination should ideally include:
* Actual assessment of them doing the job (obviously good, but in practice unlikely to be achievable!)
The above is the best, but if unavailble then (and in approximate order of how good they are individually)...
* Formal testing of work-based skills
* Formal testing of experience-based knowledge
* Formal testing of logical deduction
* Formal testing of creative thinking
* Psychometric testing.
That's all from the top of my head, so I might be misremembering some things or forgetting to include others.
Well, well below the above came structured interviews. Then miles below that comes unstructured interviews, random selection and graphology, which are amusingly in that order. Yes -- testing somebody's handwriting turned out to be marginally worse than actually picking at random. (There is no statistical significance to it being worse though, it's just how it happened to come out in testing).
kabbes
First intersubjective, then internalised
A major reason why such formal testing isn't done in practice, of course, is that it is very difficult and time-consuming to construct the tests that actually measure what you want them to measure. It took many man-weeks of time for us to put together the combination of objective testing that would do the job. And we were people who, by our very nature, were very logical and methodical in our approach and who were used to tackling large, ill-defined projects. And we were guided on the path by the aforementioned team of employment psychologists. It's no wonder that most people, faced with little time and no real understanding of exactly what to do, simply give a candidate an interview.
Brainaddict
slight system overdrive
Interesting, thanks. I'm surprised that the 'formal testing', which I take to mean exams, are so highly thought of considering how bad a test they are considered to be in education thinking.
Onket
Je suis [CONTENT REMOVED]
Interview & competency testing. These are internal appointments following a re-structure remember, the people interviewing/selecting already knew the applicants and had already worked with them, in some cases for years.
cesare
shady's dreams ♥
How were the competences tested?
kabbes
First intersubjective, then internalised
Exams of a sort. But not necessarily written and not necessarily with a "right answer". Also, they have to be carefully structured to be testing for something in particular.
You have to go through the process of:
* Identifying the exact skills and traits desirable in the candidate..
* Working out how to measure those skills
* Working out a way to test for that measure
* Working out a series of questions that fulfil this
* Working out how to translate the answers into the measure.
So, for example, to test for signs of ability to present to clients, we had the candidates prepare and present a five minute talk. When grading that talk, we had REALLY precise measures for four specific criteria. Measures that followed a four point scale of, basically, "Didn't exhibit behaviour/skill at all" --> "Some evidence" --> "Much evidence" --> "Lots of evidence". The criteria were to do with structure of talk, confidence of talk, ability to handle questions and communication skill.
The candidate then had to have zero "no evidence" ratings to get more than a 1, at least three out of four "some evidence" ratings to get more than a 2, at least two out of four "much evidence" ratings to get a 3 and at least two out of four "lots of evidence" ratings to get a 4. That gave them their score for the "presentation" element.
So exams of a type, but live exams testing something specific.
Similar threads
