What IQ Actually Predicts, and What It Doesn't

It usually starts with a number. A school report, a gifted-programme cut-off, a psychologist's letter, and suddenly there's a figure attached to your child, and you're lying awake wondering what it means. Is 112 good? Is 94 a problem? Does this number describe who your child is, or who they'll become?

Here's the honest answer up front: IQ is one of the most studied measurements in all of psychology, it genuinely predicts some things, and it predicts far less than most people assume. The gap between what IQ actually measures and what parents fear it measures is enormous, and understanding that gap will change how you read any report that lands on your kitchen table.

This article walks through what the evidence really says: what IQ scores predict reasonably well, what they predict weakly, and what they don't predict at all. No hype in either direction.

What an IQ Score Actually Is

An IQ score is not a measurement of how much knowledge sits in your child's head, and it isn't a reading of some fixed quantity of "smartness" stamped into their brain. It's a comparison. Modern tests like the WISC-V (the Wechsler Intelligence Scale for Children, the most widely used children's test) compare your child's performance on a set of reasoning, language, memory and speed tasks against a large sample of children the same age.

The scores are placed on a scale where 100 is the exact average and most children, about two-thirds, fall between 85 and 115. Scores from 90 to 109 sit in the Average band. So an IQ of 100 doesn't mean "answered 100 things correctly"; it means "performed like a typical child of this age, on this day, on these particular tasks."

Three things follow from this that parents are rarely told:

• It's a snapshot, not a scan. The score reflects performance during a couple of hours on one day. Anxiety, poor sleep, a cold, or rapport with the examiner all leave fingerprints on it.
• It's an average of very different abilities. The WISC-V combines five distinct indexes, Verbal Comprehension, Visual Spatial, Fluid Reasoning, Working Memory and Processing Speed. A child can be exceptional in one and well below average in another, and the single "Full Scale" number quietly averages that story away.
• It's less stable in young children. Scores measured at five or six wobble more than scores measured at twelve. Early numbers deserve to be held loosely.

What IQ Predicts Reasonably Well

Let's be fair to the test. Decades of research show IQ scores have real predictive power in a few areas, and pretending otherwise doesn't help anyone.

School achievement, on average. IQ is one of the strongest single predictors of academic results that psychologists have. Children who score higher tend, as a group, to find schoolwork easier, learn new material faster and score higher on standardised tests. This makes sense: an IQ test and a maths exam both draw on reasoning, language and memory.

But notice the qualifier: on average, as a group. Even at its best, IQ typically explains somewhere around a quarter to a half of the variation in school grades. Flip that around and it's startling: half to three-quarters of what separates children's school results comes from things an IQ test never touches, effort, attendance, teaching quality, organisation, confidence, interest, home support, and specific skills like reading fluency.

Speed of learning genuinely new material. Children with stronger fluid reasoning (the ability to solve novel problems without relying on taught knowledge) tend to need fewer repetitions to grasp a new concept. That's useful for teachers to know, because it shapes how much practice and scaffolding a child needs, not whether they can learn at all.

Broad occupational complexity, weakly and distantly. Across whole populations, higher childhood scores are statistically associated with entering more cognitively complex occupations decades later. But the association is loose, it operates through education, and it says almost nothing about any individual child. Plenty of average scorers become engineers; plenty of high scorers don't want to.

What IQ Predicts Weakly or Not at All

Now the list that should help you sleep, because it's long, and it covers most of what you actually care about.

Happiness and life satisfaction. There is no meaningful relationship between IQ and how happy or satisfied a person is with their life. None worth building a worry on. Warm relationships, purpose, health and security do that work.

Character, kindness and integrity. IQ tests measure nothing about whether a child is generous, honest, brave or cruel. Some of history's most destructive people were intellectually brilliant; some of the finest people you'll ever meet sit squarely at 100.

Creativity. Beyond a fairly modest threshold, IQ and creative achievement part company. Original thinking depends on curiosity, tolerance for failure, deep engagement with a domain, and willingness to look foolish, none of which appear anywhere on a test form.

Effort, persistence and self-regulation. These behaviours, sometimes grouped under "executive function" or self-discipline, predict school grades impressively well, and in some studies of report-card grades they rival or beat IQ. A child who can plan, start, persist and resist distraction converts whatever ability they have into results. A child who can't, doesn't, regardless of the number.

Social and emotional skill. Reading a room, repairing a friendship, calming yourself down: invisible to an IQ test, enormously visible in a life.

Success in any individual case. This is the one to underline. All of IQ's predictive power is statistical, true of groups, hedged for individuals. An IQ score tells you about probabilities across thousands of children. It tells you very little, with confidence, about your child's particular future.

Why Bright Children Still Struggle

If IQ predicted school success cleanly, a high score would rule out academic trouble. It doesn't, and this is where the single number actively misleads.

A child can have strong verbal and reasoning scores alongside a real weakness in working memory (the mental sticky note that holds information while you use it) or processing speed (how quickly the brain handles simple tasks). That profile produces a child who sounds brilliant in conversation, then loses the thread of three-step instructions, or knows the answer but can't get it onto the page before the lesson moves on. The Full Scale IQ averages those peaks and troughs into a bland middle number that describes nobody.

Equally, specific learning difficulties such as dyslexia occur across the whole IQ range. A child with an IQ of 120 can have severe difficulty mapping sounds to letters, because decoding draws on phonological processing, a skill only partly related to general reasoning. This is exactly why modern diagnostic frameworks moved away from requiring a gap between IQ and achievement: the gap model kept bright strugglers waiting for help they needed years earlier.

So when a teacher says "she's clearly smart, she just needs to apply herself," and your gut says something else is going on, your gut deserves a hearing. The pattern underneath the IQ, not the IQ itself, is usually where the explanation lives.

The Profile Matters More Than the Number

Ask any experienced psychologist what they look at first in a cognitive assessment, and it's rarely the headline figure. It's the shape: which indexes are high, which are low, and how the gaps line up with what's happening in the classroom.

Consider three children who all have a Full Scale IQ of 100:

• Child A scores evenly across all five indexes. The number describes them fairly well.
• Child B scores 120 on Verbal Comprehension and 80 on Processing Speed. They're articulate and slow, often mislabelled lazy.
• Child C scores 115 on Fluid Reasoning and 82 on Working Memory. They grasp ideas instantly and forget instructions instantly, often mislabelled inattentive.

Same number, three completely different school experiences, three different sets of supports needed. Child B needs extra time and reduced copying; Child C needs chunked instructions and written checklists. The single number suggests all three are "fine."

This is why, if you're trying to understand a struggling child, the useful question is never "what's their IQ?" It's "what does their cognitive profile look like, and does it explain what we're seeing?" A structured screening that measures reasoning, memory, processing speed and academic skills side by side, the approach we take at GiraffeLens, which reports criterion-referenced ranges rather than IQ scores, exists precisely because the pattern is what points to action. You can see what's covered at [/what-we-measure].

Can IQ Change?

Less fixed than its reputation suggests. A few honest points:

Individual scores move. Test-retest studies show real fluctuation, particularly across childhood and adolescence. A score at six is a weaker forecast of a score at sixteen than most people assume.

Environment matters at the population level. Average measured IQ rose substantially across the twentieth century in many countries, far too fast for genetics to explain, driven by schooling, nutrition, health and the increasing abstractness of daily life. Whatever the tests measure, it responds to environment.

Schooling itself raises scores. Children who, by accident of birthday cut-offs, get an extra year of education score measurably higher than nearly identical peers who don't. Education doesn't just fill heads with facts; it trains the kind of abstract thinking the tests reward.

But targeted "IQ training" mostly disappoints. Brain-training apps reliably improve performance on the trained game and little else. The honest route to a child performing better on cognitive tasks is the unglamorous one: sleep, reading, rich conversation, good teaching, treating anything treatable (hearing, vision, anxiety, attention), and time.

What to Do With a Score, Whatever It Is

If a number is sitting in a report in front of you tonight, here is a sane way to hold it.

Treat it as one data point, dated. It describes performance on particular tasks on a particular day. It is evidence, not identity.

Look past it to the profile. Ask whoever produced the report to walk you through the index scores and what each means for the classroom. If a report gives only a single number with no breakdown, it has told you almost nothing actionable.

Check it against real life. A score that contradicts everything you and the teachers observe is a score to query, not to grieve over. Retesting contexts, anxiety, and measurement error are all real.

Never say the number to your child as a verdict. Children turn numbers into ceilings or pedestals with frightening speed. Talk instead about specific strengths ("you're brilliant at spotting patterns") and specific, changeable skills ("we're going to practise holding instructions in your head").

Act on function, not figure. Supports, adjustments and interventions should target what the child actually finds hard, reading fluently, holding instructions, finishing on time, not the abstraction that summarises it. And remember that only a registered psychologist can conduct a formal cognitive assessment or make a diagnosis; if the profile and the classroom story both point to something specific, that's the path worth funding.

Your child was the same person the day before the test as the day after. The number changed nothing about them. At its best, it's a torch beam, useful for seeing one part of the landscape clearly, and never to be mistaken for the landscape itself.