A reasonable question sits behind a lot of searching on this subject: if I had to bet on a child, would I rather know their IQ or know what they are like?
The research gives a clearer answer than you might expect, and it is not the flattering one that personality beats intelligence. It is that they measure different things, both matter, and only one of them is much use as a prediction on its own.
| Intelligence | Personality | |
|---|---|---|
| What it describes | How well you solve novel problems | How you habitually behave |
| Measured by | Performance. There are right answers | Self-report. There are no right answers |
| Can you fake it | No, you cannot fake solving a puzzle | Easily, which is a real problem for the research |
| Stability through life | High from adolescence onwards | Moderate. Conscientiousness rises with age |
| Predicts school results | Strongly | Moderately, and largely independently |
| Responds to effort | Slowly, mainly through education | More than people assume, especially habits |
What intelligence predicts
For academic outcomes, a great deal. The largest UK study of the question followed more than 70,000 children and found a correlation of 0.81 between intelligence at eleven and educational achievement at sixteen, measured as latent traits (Deary et al., 2007).
That is a strong relationship by the standards of anything in psychology. There is an important caveat about what a latent-trait correlation means in practice, and it is worth understanding before quoting the number, but the headline stands: ability predicts attainment well.
What personality predicts
Poropat’s meta-analysis pooled the research on the five-factor personality model and academic performance. One trait did most of the work.
Conscientiousness correlated with academic performance at around 0.22 (Poropat, 2009). Openness to experience contributed something smaller. Agreeableness, emotional stability and extraversion contributed little or nothing.
Taken alone, 0.22 is modest. The reason it matters is that it is largely independent of intelligence. Conscientiousness is not a proxy for ability. It adds predictive power on top, which is why selection processes that use both do better than either alone.
There is a second reason to care about conscientiousness specifically. It is the trait that moves most through life. It rises naturally with age, and the behaviours underneath it, turning up, finishing things, planning, are closer to habits than to fixed dispositions. You cannot teach a child to be cleverer in a term. You can plausibly help them become someone who finishes what they start.
Where the comparison breaks down
The two are not measured in the same way, and this matters more than most comparisons admit.
An IQ test is a performance measure. There are correct answers and you cannot fake your way to them. A personality questionnaire is self-report, with no right answers, and it can be faked trivially by anyone who wants to look organised. In a low-stakes research setting that is tolerable. In any situation where the answer matters, it is a serious weakness in the instrument.
So when you see personality and intelligence compared as predictors, remember you are comparing a measurement that cannot be gamed against one that can.
In my experience
I taught for several years before I had children of my own, and the pupils I was most often wrong about were the quick ones.
There was a boy who could see the answer to almost anything before I had finished writing it on the board, and who produced, across an entire year, perhaps a third of the work I asked for. And there was a girl in the same class who took visibly longer over everything and handed in every single piece, on time, redone when it came back marked. By the end of the year she was further ahead, and the gap kept widening afterwards.
If someone had shown me their two test scores in September I would have bet on him without hesitating, and I would have been wrong in a way the scores could not have warned me about. That is roughly what an r of 0.22, sitting independently alongside a much larger one, looks like when it walks into a classroom.
Where emotional intelligence fits
This is the question that usually arrives next, and it deserves a straight answer rather than either the enthusiastic or the dismissive one.
Emotional intelligence is measured two quite different ways, and the two barely agree with each other. Ability-based measures present scenarios and score the responses against a key, which makes them closer in kind to an IQ test. Self-report measures ask people to rate how well they read emotions, which makes them closer in kind to a personality questionnaire, and in practice they overlap heavily with agreeableness and emotional stability.
That matters because the two kinds of measure produce different conclusions. The strong claims you will have seen, that emotional intelligence matters more than IQ for success, generally rest on the self-report versions, which is to say on personality traits under a more appealing name.
The sober position is that ability-based emotional intelligence looks like a modest, real thing that correlates somewhat with general intelligence, and that self-report emotional intelligence is largely a repackaging of traits the five-factor model already measures. Neither displaces reasoning ability as a predictor of academic outcomes.
Can either one be changed?
Asymmetrically, which is the practical point.
Reasoning ability moves slowly and mainly through sustained education, in the region of a few IQ points per additional year, and not through anything you can buy as an app.
Conscientiousness moves more readily, partly because it rises with age anyway and partly because the behaviours underneath it are habits rather than dispositions. Turning up, starting before the deadline, finishing a thing badly rather than not at all. A child can be helped towards those in a way they cannot be helped towards faster pattern recognition.
If you are deciding where to put your effort, that asymmetry is the answer. The trait with the smaller correlation is the one that is more open to influence.
What to do with both
If you have a score, treat it as one input. It tells you about the kind of problem your child will find easy, and it predicts attainment reasonably well at a population level. It does not tell you whether they will do the work.
If the score is high and the results are not, you are probably looking at a conscientiousness or interest problem rather than an ability one, and that is the more tractable of the two. If the score is low and the results are fine, the score is probably understating them, and there are plenty of ordinary reasons for that.
And be wary of treating either number as a description of a person. A child who learns they are the clever one can become a child who avoids anything they might not be immediately good at, which is a conscientiousness problem created by an ability label.
For a current reading, the free IQ test for adults and the children’s version both report a percentile rather than a bare score. What they measure is reasoning, which is one of the two things on this page, and a nonverbal test measures it with less cultural interference than a verbal one does.
References
- Poropat, A. E. (2009). A meta-analysis of the five-factor model of personality and academic performance. Psychological Bulletin, 135(2), 322-338. doi:10.1037/a0014996
- Deary, I. J., Strand, S., Smith, P., & Fernandes, C. (2007). Intelligence and educational achievement. Intelligence, 35(1), 13-21. doi:10.1016/j.intell.2006.02.001

