Apr 26 2011

IQ and Motivation

A recent study finds that motivation is a key component of performance on standardized IQ test. Researchers compared IQ scores with and without a material incentive to perform well. They found that everyone performs better with the incentive, but those with lower baseline IQ scored have a larger effect. They conclude from this that performing well on an IQ test requires both intelligence and motivation, but that those who score poorly may lack either one or both – and some simply lack motivation.

These results reflect a broader theme I often point out to students – when we examine the nervous system we are mostly asking people to perform specific tasks and then we infer how well the system is working based upon their performance. We cannot (with few exceptions) directly interrogate the nervous system to see how well certain components are functioning. So we design tasks that are meant to isolate, as much as possible, one function and to also maximally stress that function.

But it is impossible to completely isolate one subsystem within the brain, and so other functions can interfere with our testing, and we have to tease this out by using multiple tasks and then triangulating to the common element that seems to be causing trouble. This is most true for cognitive function.

For example, if I want to test Wernicke’s area (in the superior posterior temporal lobe) I will give the patient several verbal commands (without non-verbal cues) and see if they can interpret them. Wernicke’s area translates ideas into words and words into ideas, and so we test its function by testing verbal comprehension. However, in order to test this one piece of the brain, the patient also needs to be able to attend to the exam (they need to be alert and attentive), then need to be able to hear and get that information to Wernicke’s area, they need to have the general cognitive ability to understand what is happening and that they are expected to do something, and they need to be able to move to execute whatever command I gave them. And they may decide just to be uncooperative with the exam, for whatever reason. So many parts of the brain and nervous system need to work together to perform even a simple task.

We isolate Wernicke’s function, and other functions, by controlling for variables. So I will give the patient many commands, some easier for Wernicke’s area to interpret, and others more challenging – but with all other variables being equal. If commands that are difficult from a language comprehension point of view are what give the patient the most difficulty, then that is probably the part of the brain that is not working.

IQ testing is not different – it is simply a task that involves multiple neurological functions, only one of which is general intelligence (which itself is a tricky concept, as there are many components to what we would think of as intelligence, and many different kinds of intelligence).  We need to think of IQ as a measure of how people perform on IQ tests, and not necessarily of any one specific neurological function. Since IQ is measuring many variables at once (memory, fund of knowledge, judgment, perception, cognitive processing speed, attention, language, test-taking endurance, and yes – even motivation) it’s difficult to separate out these variables. It can be done to some extent, and the most elaborate battery of IQ tests (called neuropsychological testing) is designed to separate out at least several components of intelligence, like verbal-IQ from memory IQ.

This does not mean that IQ testing is worthless, we just need to have a nuanced and realistic view of what it is and what it is not. This new data on motivation is not surprising at all, and adds to our ability to properly interpret IQ testing.

Also, another way to look at IQ testing, or any type of testing, is not what it is actually measuring but what it predicts. If a test is predictive of some outcome, then it has utility, regardless of how messy a measure it is. The BBC quotes Dr James Thompson, senior honorary lecturer in psychology at University College London, who said:

“Life is an IQ test and a personality test and an IQ result contains elements of both (but mostly intelligence).

“If an IQ test doesn’t motivate someone then that is a good predictor in itself.”

I agree – if someone possesses the raw cognitive ability, but lacks the motivation to apply it, they will perform poorly on IQ tests, and perhaps in life also. For example, IQ testing does predict school performance (if imperfectly). Perhaps both are strongly influenced by motivation. Therefore, despite the fact that IQ testing is messy and complex, it can still tell us something useful. It just should not be viewed as a pristine measure of something innate and immutable about a person.

17 responses so far

17 Responses to “IQ and Motivation”

  1. HHCon 26 Apr 2011 at 11:44 am

    Alcoholics can test similarly to normals on IQ scores. They fail on reasoning, abstraction, and logical analysis.

  2. mvoetmannon 26 Apr 2011 at 12:46 pm

    Is this study online anywhere ? Or does anyone know what has actually been tested and on what kind of population ? Everybody seems to be quoting the BBC.

  3. yogzototon 26 Apr 2011 at 1:04 pm

    Here is a link to the abstract and full article (if you have PNAS access): http://www.pnas.org/content/early/2011/04/19/1018601108.abstract

  4. Jeremiahon 26 Apr 2011 at 1:10 pm

    I’ve often wondered if there are some people that are faster thinkers who excel at quick reaction tests, and on the other hand more abstract thinkers who inherently need to take more time with any problem while exploring for the abstract implications?
    And if there is that difference, say between short term and long term problem solvers, are the I Q tests that must be completed within a time limit a fair measure of the persons overall predictive abilities?
    In other words does motivation sometimes equate to necessity for a speedier response, skipping the harder problems for the ones with the more simple trickery involved?
    Clever beating contemplative perhaps?

  5. Steven Novellaon 26 Apr 2011 at 1:24 pm

    I could not find the study online – sometimes they send the press release out before the study is published. I will keep an eye out for it and link to it when I find it.

  6. James Foxon 26 Apr 2011 at 1:36 pm

    I recall that processing speed is considered a component of intelligence as measured in tests such as the WAIS. The quicker someone is able to use their raw intellectual abilities to process new and recalled data surely has an impact on overall functioning and intelligence. And I would suppose that capacity to quickly asses information and reach conclusions would add to someone’s motivation (or reduce a potential loss of motivation) when engaging in mentally taxing endeavors.

  7. Jeremiahon 26 Apr 2011 at 1:44 pm

    On reflection, that was probably a dumb question.

  8. micheleinmichiganon 26 Apr 2011 at 2:08 pm

    Is this the study you are looking for?

    Role of test motivation in intelligence testing

    Looks like you have to pay to download the full version.

  9. mvoetmannon 26 Apr 2011 at 2:57 pm

    That looks right. Thank you.

    My guess was that such an effect would be on the order of one half SD, so the 0.64 SD seems very plausible to me. That is not negligible, but it is not a very large effect either.

    The description of the second half of the study makes me a bit nervous, though. Some experts try to judge which children look more motivated? Really?
    And when they adjust for this perceived motivation, the IQ tests become less predictive. Just about _any_ adjustment to IQ measures will make them less predictive. If you adjust for bio-rhytms, IQ will also become less predictive. So that part of the description, I am not so impressed with.

    I guess I will have to pay $10 if I want to know the details. Maybe what they actually did makes more sense than what the abstract suggests.

  10. petrossaon 26 Apr 2011 at 3:54 pm

    I have HFA and had an IQ off the scale plus eidetic memory when i was a kid. School didn’t motivate me so i had very low grades if any. Eventually got expulsed. The only thing i ever used my intelligence for is to do as little as possible for the most gain. In that i succeeded very well. I never found a cure for cancer but i did manage to work very little and retire at leisure at an early age.

    Nobody will remember my name, but since i won’t be there to know it who cares?

    At a certain level everything becomes so abstract that nothing is important.

  11. Jeremiahon 26 Apr 2011 at 4:28 pm

    So does this mean someone for one or more such reasons can score highly on mathematical induction, i.e., mathematical analogy, yet at the same time not nearly so on the emotionally inductive recognition of abstract analogy?

  12. petrossaon 27 Apr 2011 at 2:32 am

    English is not my mothertongue so if i misunderstood your question let me know. What i understood from your question leads to this answer:

    In my view emotions aren’t part of your consciousness,of the flow of information that constitutes you. To me emotions are the result of a complex interpretative analysis by the neo-cortex of various corporal indicators, such as body stance, muscle tone, facial expression, hormonal levels, heart/respiration rate, etc. controlled by the limbic system.

    The ‘integration’ of emotions is in that view wholly dependent on the functioning of the ‘module’ doing to the monitoring.

    After the emotion has been analyzed it gets spliced into your consciousness giving the event an extra dimension.

    As such it has no direct role in higher order analytical processes, other then an obfuscating one.

  13. micheleinmichiganon 27 Apr 2011 at 10:15 am

    Thank you for the article, Steven Novella. It is quite thought provoking, for me.

    I have often been quite critical of IQ testing, due to an experience when I was 12. After some thought, I realize that my concerns should have been on how some IQ test are given and interpreted.

    I struggled with two issues through most of my k/12 years. Math and test/performance anxiety. In all topics, excluding math or math based subjects, I seemed to do well enough. But it was not unusual for a teacher to come to me with the results of some standardized test and say with confusion, “What happened to you?”

    In 7th grade science class (in 1976) we were given an IQ test. I remember being in a rush, because the test was timed, looking at the first problem, with blocks, not really getting what I was supposed to do, thinking I would get in trouble if I used the test paper for figuring and then thinking ‘oh, this is like math. I can’t do it.’ I don’t remember the rest of the test.

    When the teacher got the results back, I believe he told us that we weren’t supposed to see the results, but he would show us anyway, if we wanted. So I filed up there with the rest of the kids and found out that my scores indicated a low/average intelligence. To his credit, he did make an effort to keeps the results private, pointing to the scores one by one.

    Both of my parents were very intelligent, My father, who was a physicist, often talked about his high IQ and my mother, while far more modest, graduated college (while raising five kids) with a 3.9 GPA.

    When I saw my IQ test results, I remember being sort of ashamed and afraid that I would get in trouble. But, my parents never talked about the test results. Things pretty much went along as they had been, with me almost failing most of my math teachers, but being given passing grades, because I was generally a nice kid, who didn’t make trouble. But, I do believe that score did somewhat undermine my confidence. Before that, I felt like I was just bad at math. After that I felt that I just wasn’t generally as smart as most of my classmates.

    To some extent the IQ test was not predictive. I did graduate in the top 1/4 of my class. My ACT composite score was well above average (even with a shockingly low math score) and, with a good portfolio, I had no problem getting into the reputable Universities with Art programs to which I applied.

    On the other hand, the anxiety and math difficulties which I believe caused the lower than average IQ score did hampered my educational and occupational flexibility. If I had been able to suss out the individual issues and work through them at the time, rather than consider the score a guideline for my generalized cognitive abilities, that might have been helpful. But, as it was used, I believe that that particular IQ score was more of a hindrance than it was a help.

    Opps, this is long, sorry.

  14. SimonWon 27 Apr 2011 at 1:16 pm

    IQ is a good predictor, but it is apparent it is likely a combination of effects from the simple fact it is almost normally distributed (although that is in part an effect of test calibration as well).

    Whilst it might be useful as a predictor, in terms of assessing ourselves as Michele did from the test we should probably not look for one number, but look at a broad spectrum of tests that tell us our strengths and weaknesses. This is also probably more useful for educationalists, but it would require a change, or enrichment of culture.

    I actually have great concentration in tests, and enjoy maths puzzles, and so score fine on IQ tests but have pitiful little to show for the abundant IQ nature bestowed on me. I suspect motivation, or procrastination, is at the heart of that but my school never gave me a test for “ability to procrastinate”, I suspect another one I’d have come out top at 🙂

  15. leonardoon 28 Apr 2011 at 6:43 am

    In my experience, if one is ‘motivated’ to rehearse a couple of IQ tests prior to the real test, the outcome tends to be higher.

  16. daedalus2uon 28 Apr 2011 at 9:04 am

    Steve, The whole concept of IQ as a unitary characteristic is highly flawed, and even if there was such a thing as IQ or g, IQ tests would be unable to measure it.


    Look especially at note 2. The idea of intelligence tests is that each test has a fractional dependence on g, so if you do k tests, you have k+1 variables the k individual weighting factors of each test and g, but only k equations. The system is under specified. It is indeterminant. There is no unique solution. This has been well known for 80 years.

    The link to Peter Schonemann is important. He is passed away now, but there are links to his multiple papers on his homepage. In some of them he not only points out the problems with the mathematical approach being used, but outlines some of the political problems he had in getting things published.

    His experience is an excellent example of the problems with peer review. When the peers are wrong (as they are in the IQ debate), peer review blocks progress.

  17. samrockyon 07 May 2011 at 3:13 am

    Really helpful for me and i would like to get every article related to neurological blog,it change my thought of view.

Trackback URI | Comments RSS

Leave a Reply

You must be logged in to post a comment.