Read my blog at Huffington Post

I also blog at Huffington Post's new UK site; please click here to read my posts there.

Monday, 13 June 2011

No good reason for abstract reasoning

What do you think of first when asked to define 'abstract reasoning'? If you're at all similar to me, then the first definition that springs to mind is 'applying reason to the consideration of abstract concepts', or words to that effect. It is increasingly common, particularly in the public sector, to require job applicants to sit an abstract reasoning test as part of the selection process. However, these abstract reasoning tests have nothing to do with abstract reasoning; at least, nothing to do with 'abstract reasoning' as I have just defined it. Instead, these tests have to do with the purely visual aspect of abstract reasoning. It will surprise some of you to learn that abstract reasoning even has a purely visual aspect, so let me explain that abstract reasoning is, in the opinion of the burgeoning assessment-test industry, all about sequences of shapes. Along the lines of: "If these four shapes are the first four in a sequence, then which of these other shapes will come fifth?"

Because, yes, fellow citizens, your taxes and mine are being spent on purchasing tests that measure people's ability to put sequences of shapes in order - what a marvellous use of our increasingly finite public sector resources. Obviously, if you're recruiting a Head of Human Resources or a Director of Finance, their ability to put shapes in order will be central to the job. Should you recruit a Press Officer, and should the phone ring with an urgent query from a tabloid reporter, nothing will then be more important than the new recruit's ability to immediately engage in some abstract reasoning, not about the details of the reporter's query, but about the order in which some shapes should be positioned on a piece of paper.

It's all utter nonsense and has naff all to do with people's ability to do many of the jobs to which it is applied. Obviously, I'm biased, because one reason (abstract or otherwise) that I hate these tests is because I'm absolutely awful at them. I invite you to imagine that you were applying for a job or a promotion and that, as part of the assessment process, you were invited to sit this test. Some of you might be very good at it, in which case congratulations, as this tells me that you're good at ordering shapes on a screen, and a great career awaits you in visual design. What it doesn't tell me is anything about your ability to reason abstractly, or your ability to do most of the jobs for which this test is used. You could be dreadful at this test, but superb at abstract reasoning. These tests are a waste of time and money and should be scrapped. Anyway, as I am today feeling benevolent, here is a relatively easy version of the same test, but go on, don't do that one, do the first one, which is harder.

<SNIFF> If you'll excuse me, I'm off now to crush my sour grapes...


  1. I really don't want to be "that guy," but I need to point out that your definition of abstract reasoning is flawed, not the tests. While everyone is entitled to their opinion, this article came up on while I was on a google search for factual information, and I don't want future readers to be misled. Abstract reasoning is the ability to identify how objects or ideas relate to one another without prior factual knowledge of such a relationship. The "abstract" part means you must be able to conceptualize the relationship in your own mind. These tests use shapes and pictures because no one could possibly know the patterns set in the shapes going in to the test. Being able to see how each shape relates to one another uses the same cognitive processes as you normally would any time you imagine the "how and why" of any situation. Abstract reasoning is extremely important for being adaptable in novel situations and being able to come up with a valid "answer" without prior experience. You'd even be using abstract reasoning to see how abstract reasoning can relate to excellent job performance!

  2. I could not agree more. Tests strip reality from every situation and make it easier on those who should be able to assess things abstractly. These tests are the antibiotic of psychological assessment.

    Abstract reasoning should allow someone to use unrelated knowledge to address any problem. Lateral thinking is the start of this, finding the links where they appear not to be. And this will probably be case specific and relate to the context of the problem. So running someone through a bunch of graphic tests when that person views things in an organic way, may produce wrong results. Some see patterns where others do not - after all, it is abstract reasoning. The solution given is the logical one, but nobody ever explores why someone gave a "wrong" answer - that may be right by way of his or her excellent abstract reasoning. How can abstract reasoning have a "right" or "wrong" answer? To whom is it "right" or "wrong"?

    In my thinking, dreaming is abstract reasoning in the brain (at very least a high-speed, pedantic version) as it tries to build the indexing information that relates the day's memories to all other existing memories. As it goes about its night's work, it produces some strange images linking what we think in the mornings are unrelated things. Yet we should embrace these strange images to encourage our abstract reasoning abilities.