What an IQ Score Actually Means: Plain-English Explanation

A friend texts you a screenshot from some online IQ test: "I got 137! Is that good?" You don't quite know what to say. Good compared to what? More than 137 of what? Your own number from years ago doesn't help either — you remember being told you scored 122, but you never really understood whether that was supposed to impress you, comfort you, or quietly worry you.

The confusion isn't your fault. The way IQ scores are talked about in everyday life is roughly thirty years behind how psychometricians actually think about them. Once you understand what the number is measuring and how it's calibrated, the answer becomes much clearer — and a lot less mystical than the test marketers want it to feel.

The number is a percentile, not a quantity

The single most important thing to understand about IQ: it's not a count of anything. It's not how many "intelligence points" you have, the way grams measure weight. It's a position on a curve, expressed in a particular numerical convention.

Specifically: IQ scores are designed so that the average of a representative population sample equals 100, and the distribution forms a normal distribution with a standard deviation of 15 points (for most modern tests). Every score above or below 100 corresponds to a specific percentile of the population.

Here's what that looks like in practice:

So when your friend says they got 137, they're saying they're in roughly the top 1% of the population the test was normed against. The number 137 doesn't have any meaning on its own — it's a label for a position.

Why "the average is always 100"

This is where most people get confused. If the average is always 100, doesn't that mean people are getting smarter or dumber depending on what year they take the test?

Sort of. Test publishers periodically re-norm their instruments against a new representative sample — when they do, the raw scores that used to produce an IQ of 100 now correspond to something slightly different, because the population baseline has shifted. The most famous example is the Flynn effect: average raw scores on IQ tests rose by roughly three points per decade through most of the 20th century in developed countries. Re-norming kept the average at 100, but a 100 in 1950 corresponds to a different raw performance than a 100 in 2010.

The practical implication for you: an IQ score is meaningful only relative to the norming sample it was calibrated against. A score from a test administered in 1990 isn't directly comparable to one from a test administered in 2025, unless the test has been re-normed in between.

The difference between full-scale IQ and subscale scores

Modern intelligence tests don't produce just one number. The Wechsler Adult Intelligence Scale (WAIS-IV), which is the clinical gold standard, breaks down into four index scores plus a full-scale IQ:

Each of these is measured on the same 100/15 scale. The full-scale IQ is essentially an average across the four, weighted in specific ways. The reason this matters: two people with the same full-scale IQ can have radically different profiles. A 130 driven by high verbal and reasoning scores with average processing speed looks different in practice than a 130 driven by very high processing speed and average verbal performance.

For a quick visualization of where different scores sit on the distribution and what each band conventionally represents, an IQ score chart is the most efficient reference. The chart format shows the relationship between the standard deviation, the percentile rank, and the descriptive classification labels used in clinical reports — which is the framing you need to interpret any specific score against the population it was normed on.

What the classification labels actually mean

You'll see different terms used by different test publishers, but the conventional bands look roughly like this:

These labels are conventions, not biological truths. The cut-offs are smooth on the underlying distribution — there's no meaningful jump in capability between a 129 and a 130. The labels exist because clinical reports need shorthand, not because nature divides cleanly at those points.

What an IQ score is good for

The honest list, based on what the psychometric research actually supports:

What it isn't good for

The fact that IQ is the most-studied human trait in psychology doesn't mean it's the most important one. It just means it's measurable.

How to read your own number sanely

If you've taken a test and want to understand the result:

  1. Note the date and instrument. A WAIS-IV from 2018 means something more specific than "I once got 122 somewhere."
  2. Look at the subscale breakdown, not just the full-scale number. The shape of your profile is more informative than the average.
  3. Check the confidence interval. Test scores are estimates with statistical uncertainty — usually ±5 points for a single administration. A "131" might be a "127" or a "135" if you took the same test again under different conditions.
  4. Don't extrapolate too far. An IQ score predicts certain things on average, across populations. It doesn't predict your specific life with much resolution.

The takeaway

An IQ score is a percentile dressed up in a particular numerical convention. It's useful information when you know what it measures and what it doesn't. If you've gotten a score from a test you trust, write it down with the date and the subscale breakdown, treat it as one input among several, and don't let it do more work than it can support.

The number isn't your identity. It's a reading from one instrument, calibrated to one population, on one day of your life. Knowing how to read it correctly is the whole point.

You May Also Like