Psychometrics and the obsession with "g"
Over the years I have completed more psychometric tests than I care to remember. If you have never taken one of these they are the kind of test you may be required to do if you are seeking a job. If you are old enough to have done the 11 plus test for secondary school you will know the form. That one was an ability test. The British Army uses similar IQ-type tests for officer selection.
Psychometrics covers two main fields of testing: ability tests and personality tests. These latter tests which have grown popular in recent years are not really tests at all but questionnaires. The answers you give to the propositions on the questionnaires reflect your personal preferences. A classic here is a series of questions designed to find out the extent to which you are gregarious or introverted; so you might be asked whether you prefer to be socialising at a party or to be tucked up with a book.
Like me, you may be thinking "it depends". But the questions are trying to force a choice and they will come at the subject more than once. I'm not a great fan of personality tests but I can see how someone wanting to put together a traditional sales team might want to find outgoing people rather than those who prefer their own company.
I think that says as much about the nature of selling as the people engaged in it. There are different ways to sell a product that might well benefit from approaches other than those used by sales staff who want to be your biggest chum within the space of two minutes.
I took a test the other day that I'm featuring in my FT column this week. It was a Gallup test designed to find your five greatest strengths. One of mine, apparently, is seeking to put things in context. When confronting a concept or some development I like to get back to the beginning, asking "Where did it come from?" I used this method in my book, Blood Sweat and Tears, The History of Work. In fact I used it in a chapter on the history of psychometrics.
Whenever I'm trying to work something out in psychometrics there are a handful of occupational psychologists I return to again and again. One of these is Steve Blinkhorn. I first came across him when I found that he and Charles Johnson, another respected psychologist, had written an article saying that personality tests could not predict future job performance.
This is a contentious issue, one that has been rejected by SHL, the big test publisher. But Blinkhorn and Johnson have never stepped back from their assertion. The point they make is that a personality test will tell people about your working style but it won't say anything about the quality of the finished product.
You may, for example, demonstrate that you are, to use the jargon, a "completer finisher". But what you have dashed off with the utmost efficiency in a few minutes might be a load of rubbish. On the other hand you might be a perfectionist and take forever to get something done. There are issues here on both sides and that's why people still need to be managed or to find ways to manage themselves.
Blinkhorn is a historian too. A week or two ago he told me he had found the original research used by Charles Spearman, a pioneer of factor analysis in the study of human intelligence. Spearman built on Francis Galton's studies of correlation - a way of drawing apparent relations between random variables such as age and height, for example. Sometimes it is used to imply causation. With age and height that seems reasonable, at least in our earlier years. We grow up as we grow older. But with other variables a causal relationship is less apparent and may not exist at all.
Spearman came up with "g", his definition of general intelligence that led, inevitably, to the development of the IQ test. His work was based on a group of boys in a school where he used to teach. What Blinkhorn has done is to track down the school records of this study group, a group of 13-year-old boys. It seems that if one of the boys, missing from school at the time of the study, had been around at the time, then the research outcome might have been quite different.
I sometimes wonder whether we would have been better off if "g" had never been discovered at all. It is "g" that makes some members of MENSA, the IQ-obsessed organisation, wonder why their membership is the best thing they have to show from their high IQs. The problem with "g" is that it doesn't mark you out for riches, for great discoveries or for a position of power and influence. On it's own it might be helpful for doing Sudoku puzzles but not much else. For those other things you will need other qualities and you may not need to be overloaded with "g".
In selection one of the main question facing recruiters is "g" and what else?
Psychometrics covers two main fields of testing: ability tests and personality tests. These latter tests which have grown popular in recent years are not really tests at all but questionnaires. The answers you give to the propositions on the questionnaires reflect your personal preferences. A classic here is a series of questions designed to find out the extent to which you are gregarious or introverted; so you might be asked whether you prefer to be socialising at a party or to be tucked up with a book.
Like me, you may be thinking "it depends". But the questions are trying to force a choice and they will come at the subject more than once. I'm not a great fan of personality tests but I can see how someone wanting to put together a traditional sales team might want to find outgoing people rather than those who prefer their own company.
I think that says as much about the nature of selling as the people engaged in it. There are different ways to sell a product that might well benefit from approaches other than those used by sales staff who want to be your biggest chum within the space of two minutes.
I took a test the other day that I'm featuring in my FT column this week. It was a Gallup test designed to find your five greatest strengths. One of mine, apparently, is seeking to put things in context. When confronting a concept or some development I like to get back to the beginning, asking "Where did it come from?" I used this method in my book, Blood Sweat and Tears, The History of Work. In fact I used it in a chapter on the history of psychometrics.
Whenever I'm trying to work something out in psychometrics there are a handful of occupational psychologists I return to again and again. One of these is Steve Blinkhorn. I first came across him when I found that he and Charles Johnson, another respected psychologist, had written an article saying that personality tests could not predict future job performance.
This is a contentious issue, one that has been rejected by SHL, the big test publisher. But Blinkhorn and Johnson have never stepped back from their assertion. The point they make is that a personality test will tell people about your working style but it won't say anything about the quality of the finished product.
You may, for example, demonstrate that you are, to use the jargon, a "completer finisher". But what you have dashed off with the utmost efficiency in a few minutes might be a load of rubbish. On the other hand you might be a perfectionist and take forever to get something done. There are issues here on both sides and that's why people still need to be managed or to find ways to manage themselves.
Blinkhorn is a historian too. A week or two ago he told me he had found the original research used by Charles Spearman, a pioneer of factor analysis in the study of human intelligence. Spearman built on Francis Galton's studies of correlation - a way of drawing apparent relations between random variables such as age and height, for example. Sometimes it is used to imply causation. With age and height that seems reasonable, at least in our earlier years. We grow up as we grow older. But with other variables a causal relationship is less apparent and may not exist at all.
Spearman came up with "g", his definition of general intelligence that led, inevitably, to the development of the IQ test. His work was based on a group of boys in a school where he used to teach. What Blinkhorn has done is to track down the school records of this study group, a group of 13-year-old boys. It seems that if one of the boys, missing from school at the time of the study, had been around at the time, then the research outcome might have been quite different.
I sometimes wonder whether we would have been better off if "g" had never been discovered at all. It is "g" that makes some members of MENSA, the IQ-obsessed organisation, wonder why their membership is the best thing they have to show from their high IQs. The problem with "g" is that it doesn't mark you out for riches, for great discoveries or for a position of power and influence. On it's own it might be helpful for doing Sudoku puzzles but not much else. For those other things you will need other qualities and you may not need to be overloaded with "g".
In selection one of the main question facing recruiters is "g" and what else?
Labels: Charles Johnson, Charles Spearman, factor analysis, Francis Galton, Gallup, IQ, MENSA, personality tests, psychometric tests, Soduko, Steve Blinkhorn
posted by Richard Donkin at
10:11 AM
0 Comments:
Post a Comment
Links to this post:
Create a Link
<< Home