Wednesday, 23 April 2014

How big does an IQ gap have to be before it becomes easily noticeable?


Just interested to hear people's personal impressions.

My first notion is that the intelligence gap needs to be at least one standard deviation (15 IQ points) to be easily detectable in normal human interactions.

And my second impression is that the gap needs to be larger looking upwards than downwards.

So, looking 'down' we can readily feel the difference of about one SD (15 IQ points) less than ourselves - but not above us.

Looking 'up' at people of higher intelligence, the IQ gap needs to be bigger than one SD (more than 15 points) for us to notice.



  1. Agreed. Trying to do something incredible simple (e.g. customer service) with what is probably 15-20 point deviation below is readily noticeable and often frustrating. Talking to someone 1 or 2 standard deviations above, I'm not immediately sure if they're just more knowledgeable/experience, or if their opinion is actually smarter or correct (how am I to evaluate?). I think the higher IQ individual has also learned to "talk down" or simplify their language or thoughts, as one might do to a child. Or to feign lower intelligence, but try to lead the other in the right direction so they can figure it out together. In general social situations, it's not particularly advantageous to make others aware of superior intelligence.

  2. 15 points? At that scale, ranking people would be an exercise in pretty large clusters. I would have thought that 5 points (1/3 SD) would be closer to the limit of detection.

    I suppose it depends on what you mean by easily noticeable. You can probably tell in a single conversation within 15 points.

  3. It is hard, if not impossible, to notice that someone is smarter than us, especially in ordinary daily interactions. You are certainly not going to see them behind the cash register. You would have to be having an intense intellectual discussion, and even then, it is possible that you think you're "getting it" when you really don't. If they point out that you don't get it, you think they are a churl not your intellectual superior.

  4. A few comments will follow, I hope you don't mind.
    I can demonstrate with tests my family members are all 3-4 SD's above average

    Our observations are,

    a) High IQ is more noticeable in the young as they have not yet learned to communicate with others; the older someone gets they more capable they are in 'blending in'

    b) Highly educated people overestimate their own IQ and underestimate the IQs of people without advanced degrees

    c) Most people with average intelligence confuse being well read with being very intelligent

    d) It seems to me anecdotally that the more toward the mean a person is the more trouble they have differentiating between a person 1 SD and a person 3 SD's, etc. In other words, the average person tends to lump everyone as 'smart' but have trouble telling the difference between 'gifted' and 'genius'. As you go right *or left* on the bell curve you seem to get better at this.

    e) My sons and I struggle in reverse; while I can recognize when a person is 1-2 SDs above or below me fairly well I often miss that someone needs more help than average - unless I really pay attention I tend to lump everyone with an IQ of between, oh, 75 and 115 together as 'average'. My sons are the same, but my wife is much more adept at noticing when someone needs more assistance.

    f) Since we educate our sons at home my wife and I comment that in some ways gifted children face hurdles similar to challenged children, although of a different sort and degree.

  5. Also, I used to think that I had average intelligence, and wondered why 99% of the people I met were so stupid. =D

  6. Auster’s anecdote about his sub-normal uncle may be of interest. I think what he’s saying is basically correct. Big differences aren’t always particularly noticeable in everyday (non-work) interactions.
    It’s here (search for “uncle” in the text):

  7. The greater difficulty of recognizing that someone is more intelligent, compared with recognizing that someone is less intelligent, seems consistent with the Dunning-Kruger effect.

  8. I don't think the perception is linear,the difference between 95 and 110 is much more noticeable than 115 and 130 for example.

  9. I've read other bloggers before complaining about the difficulty in communicating with people 1 sd or more below them in intelligence, and it's surprised me because this is not something I remember encountering a lot. In fact, I've sometimes wondered how much intelligence really varies among people; after all, IQ is normalized to the width of the distribution, so it doesn't tell you how wide this distribution is in any absolute sense. Maybe some people are being uncharitable and assuming that their listeners must be stupid when the speaker is just not explaining clearly. When my students can't follow my reasoning, I take this to be my fault. (Right now I'm haunted by the blank stares I was getting trying for the first time to teach special relativity to non-science majors. I must do better next time.) I certainly do notice differences in creativity and abstract reasoning ability in my graduate students, but they're only "easily noticeable" in technical discussions. Nor have I noticed some commenters at the Orthosphere being noticeably more intelligent than others.

    On the other hand, the illusion could be mine. I may well be in an environment where my own intellect is close to but somewhat below the average. Most people I interact with then would be close to the same IQ as me, and of those with markedly different IQ, most would be more intelligent, which would be harder for me to gauge. In fact, this seems not unlikely for a number of reasons.

    Anyway, I'm fascinated by your experience of clear differences in intelligence being easily noticeable. Does this mean that the difference between, say, IQ 85 and IQ 115 is large in some absolute sense, that an adult with the latter could do much more than one with the former? Can you give an example of something the latter could do or understand but not the former?

  10. In daily life, it's mixed up with so many other things - ability to access common sense, conscientiousness, how much someone has read, the richness of their upbringing and longer family tradition.

    If one is in the habit of interviewing people, and intelligence is one of the things one looks for, then it often becomes quite clear after a few minutes (which is not to say that one forms a closed judgement so early - it's just a phenomenological observation).

    It's a quality notable when it comes to foreseeing the distant consequences of actions today, and when trying to solve an unfamiliar problem. Of course instinct (there is a kind of complex knowhow that does run in families) and training may go a long way towards occluding the contribution of intelligence, but that does not mean one cannot distinguish it from other factors.

  11. @Bonald - It's a rather big topic for a blog comment - but I would say that IQ differences show up in two main areas: speed of learning and ability to abstract.

    Probably the best way to understand this would be to look at the way that the military uses IQ testing in order to allocate personnel to different kinds of job (and different ranks); and to exclude those who would be predicted (from past experience/ data) as highly likely to fail training programs, or need to need multiple repetitions of training, or who could learn only highly circumscribed tasks, or who would require close supervision.

    A less intelligent person might be able to learn a procedure, but not be able to compare procedures and decide which was appropriate. And only the most intelligent can critique a procedure - to know whether or to what extent a procedure was valid in terms of attaining objectives.

  12. Bonald I would say to you that you must yourself be a man of greater than average intelligence, perhaps significantly so, seeing as you do work in physics and the quality of your writing in general. By the Duning-Kruger effect:

    The Dunning–Kruger effect is a cognitive bias which can manifest in one of two ways:

    [1] Unskilled individuals suffer from illusory superiority, mistakenly rating their ability much higher than is accurate. This bias is attributed to a metacognitive inability of the unskilled to recognize their ineptitude.

    [2] Those persons to whom a skill or set of skills come easily may find themselves with weak self-confidence, as they may falsely assume that others have an equivalent understanding.

    You are therefore much more likely to overestimate the intelligence of your peers, being overly charitable. After all, it takes a great amount of intelligence to be able to explain something simply.

    As for your experience of teaching students, I would recommend reading this blog: perhaps his experiences will be illuminating

  13. I don't know; it's hard for me in practice to tell the difference between a genuine difference in IQ and a momentary lack of common interests. Most people can seem much stupider than they are when the real problem is that they simply don't know much or anything about what you are talking about, but for sake of politeness or intimidation cannot openly say so and ask for a change of subject. Conversely, most people have at least one topic on which they can converse with much greater knowledge and insight than others, and in which context they will appear much smarter than they may normally.

  14. @SJ - Yes indeed - but at a certain gap it does become obvious.

  15. "Yes indeed - but at a certain gap it does become obvious."

    I'd observe that it depends on what you're using as your indicator metric, as well. The vast majority of "normal human interactions" tend to rely by default and reflex on *verbal and social communicative* ability as a shorthand indicator of IQ; most people, I suspect, can think of at least one person they know for whom that doesn't correlate to actual intellectual ability, either up or down. In a highly heterogeneous society, lack of common language can also be a big obscuring factor.

  16. @SJ - What you are perhaps saying is perhaps that you would personally not notice any difference until somebody was, say, mentally handicapped. That might be about three of four standard deviations. Or are you just saying that there is always the possibility of making a mistake? That is true, but something that can always be said about anything.

  17. @SJ - What you are perhaps saying is perhaps that you would personally not notice any difference until somebody was, say, mentally handicapped. That might be about three of four standard deviations. Or are you just saying that there is always the possibility of making a mistake? That is true, but something that can always be said about anything.

  18. The truly gifted are better at using different types of intelligence and choosing when it is applicable or appropriate to do so. They can appear aloof and arrogant. Pseudo intellectuals are ego-driven, as are we all to some extent, however, what our ego is ultimately used for is of importance. High ego individuals are often satisfied to simply serve their ego. I would argue that the truly gifted - people like Einstein, Goethe, William Shakespeare - are venerated largely for their ability to innovate and successfully relate that knowledge to the great masses. It can also be observed in their ability to ascertain which degree of specificity or granularity is required in any given situation; when technical language is appropriate or when to break things down into layman's terms. One can be supposedly intelligent and yet fail due to an ego which serves only itself. Likewise some powerful figures in history may think selfishly only of "their" people and fail to address humanity on the individual level - the other side of this is the great many who do exactly the reverse of this - people who only think selfishly on an individual level at the expense of the masses. Many (although certainly not all) of the greatest catalysts in human history have been excellent at both relating to other individuals, adapting to the perspectives of others and then relating back to us their observations in a way beneficial to humanity as a whole.

    "Where wisdom reigns, there is no conflict between thinking and feeling" - Carl Jung.