Jun 2, 2026·IQ Scores & Interpretation

What Is Full Scale IQ? How to Read the Overall Score

Full Scale IQ explained: what the overall IQ score means, what it can miss, and how to read it alongside index and subtest scores.

Dr. Russell T. WarneChief Scientist

What Is Full Scale IQ? How to Read the Overall Score

If you've ever received a score report after taking an IQ test, you know the moment: there's a page of index scores, subtest breakdowns, percentiles — and then, usually near the top, one number that everyone's eyes go to first. That's the Full Scale IQ. It's the number people remember, repeat, and sometimes worry about. But it's also the most frequently misunderstood number on the entire report.

I want to walk through what the Full Scale IQ actually is, how it's constructed, what the number does and doesn't tell you, and how to read it the way a psychometrician would rather than the way a headline would.

What the Full Scale IQ Actually Is

The Full Scale IQ (FSIQ) is not a raw measurement of anything. It's a composite measure of cognitive ability calculated from performance across multiple subtests — not just one task, but a statistical summary of how someone performs across language comprehension, visual reasoning, memory, and mental speed, all collapsed into a single number on a standardized scale.

The construction process is layered, not flat. Each subtest is scored, the subtest scores combine into index scores, and the index scores combine into the FSIQ — so the test yields not one result but a layered profile of abilities. Put differently, the FSIQ sits at the top of a pyramid. Beneath it are the broad index scores (verbal comprehension, visual-spatial reasoning, fluid reasoning, working memory, processing speed, depending on the battery), and beneath those are the individual subtests that feed into each index.

Every subtest starts with a raw score — the actual count of correct items or the time bonus achieved. That raw score gets converted to a scaled score using age-appropriate normative tables, where scaled scores have a mean of 10 and a standard deviation of 3. The subtest scaled scores then get summed within each index, and that sum converts to an index score on the familiar 100/15 metric. The index sums are then added together, and the manual's table yields the Full Scale IQ. There's no shortcut — the table lookups are non-linear because subtest variances differ.

That non-linearity matters. It's part of why two people can't simply average their index scores by hand and expect to land on the published FSIQ — the conversion tables account for how subtest scores correlate with each other across the standardization sample.

Why 100 Is the Center and 15 Is the Unit

Once the FSIQ is calculated, it's reported on a scale that has become so familiar it's easy to forget it's a convention rather than a natural law. The modern IQ scale is anchored by two numbers: a mean of 100 and a standard deviation of 15. Both are conventions, not discoveries. The mean is fixed at 100 because it's a round, easy-to-read center point. The SD is 15 because that's the value Wechsler chose in 1939 and the field consolidated around it.

That convention is what makes the FSIQ interpretable as a relative position rather than an absolute quantity. Raw test scores are converted to standard scores where the average IQ is 100 and the standard deviation is usually 15. Intelligence follows the normal distribution, which means it has a bell curve shape. If we apply the empirical rule to the IQ distribution, 68% of people have an IQ between 85 and 115; 95% fall between 70 and 130; and 99.7% fall between 55 and 145.

This is also why percentile rank — not the raw number — is the more meaningful figure for interpretation. A percentile means the share of the norm sample scoring at or below a score. If your result is at the 84th percentile, you scored as high as or higher than about 84 out of 100 people in the comparison group. And the relationship between score and percentile isn't linear. The gap from IQ 100 to IQ 115 is 34 percentile points (50th to 84th), while the gap from IQ 115 to IQ 130 is only about 14 percentile points (84th to 98th) — even though both spans are exactly one standard deviation. This is a property of the bell curve, not a quirk of any particular test.

The Classification Bands — and Their Limits

Every FSIQ report includes a classification label: Average, High Average, Superior, and so on. The Wechsler scale classifies scores as follows: 130 and above is Very Superior, 120–129 is Superior, 110–119 is High Average, 90–109 is Average, 80–89 is Low Average, 70–79 is Borderline, and 69 and below is Extremely Low.

It's worth being honest about where these labels come from. Because IQ scores are forced to follow a normal distribution by the way they're scaled, the classification bands map onto rigid percentile cutoffs. The labels themselves, though, are conventional. The labels — gifted, superior, average, and so on — are conventional but vary slightly between test publishers. The percentile rank is more meaningful and is what should drive any interpretation.

This is one reason I'm cautious about treating any single classification boundary as a hard line. These cutoffs are not based on any natural boundaries in human cognition. Cognitive abilities exist on a continuum. There is no magical point where someone suddenly becomes a "genius." A score of 119 and a score of 120 sit on opposite sides of a published label boundary, but they represent nearly identical levels of ability. The number is more informative than the label attached to it.

Why a Single Score Is Always a Range

One of the most consequential things I can tell someone reading an IQ report is that the FSIQ is not a single point — it's the center of an interval. Every IQ test has measurement error. The WAIS-IV reports its standard error of measurement (SEM) as about 2.5 IQ points, which means a measured score of 120 corresponds to a 95% confidence interval of roughly 115 to 125 — the "true" score could fall anywhere in that range.

This matters more for shorter or online assessments. Online tests have larger SEMs — typically 5 to 8 IQ points, depending on test length and item quality. A 60-question online IQ test might give a 95% confidence interval of plus or minus 10 to 15 IQ points around the measured score. This is why repeat-testing produces apparently different scores even when ability is unchanged. The practical implication is that a single score should always be reported as a range.

I'd add one more layer to this: the confidence interval isn't a flaw in the test. It's the test being honest about its own precision. A report that gives you a single number with no interval is giving you false precision — and that's a red flag for the quality of the instrument, not a sign of confidence.

What the FSIQ Doesn't Tell You

Here's where I want to be direct, because this is the part most popular discussions of IQ get wrong. The FSIQ is often treated as a direct stand-in for "general intelligence," or g — the statistical factor that explains why performance on different cognitive tasks tends to correlate. But the FSIQ and g are not the same thing.

The Wechsler FSIQ does not directly or exclusively measure g across the full range of the population distribution of intelligence. The general factor of intelligence results from the positive correlation of all human cognitive abilities — but the FSIQ is a composite that depends on which subtests are included and how much each one is weighted. In practice, that means the FSIQ is a very good approximation of g for most people — but it's an approximation built from a specific set of tasks, and it can behave differently when someone's abilities are unevenly distributed across domains.

This is exactly the scenario where the FSIQ can become misleading on its own. By definition, the Full Scale IQ is a score attempting to measure intelligence, established by calculating the average performance across a number of subtests. It therefore depends on the subtests used and the influence specific cognitive functions have on those performances. Treating intelligence as a singular, homogeneous construct can produce misleading interpretations of cognitive outcomes — especially in clinical populations.

Concretely: if someone has a very high verbal comprehension index but a notably lower processing speed index, those two numbers might average into an FSIQ that accurately represents neither. The FSIQ is most informative when the index scores that feed into it are reasonably consistent with each other. When there's a large spread between indices, the index profile tells you more than the single composite does.

What the Number Doesn't Tell You About a Person

I'll close on something that applies regardless of where someone's FSIQ falls. Intelligence test scores estimate a construct using interval scales and have meaning only relative to other people of the same age. Someone with an IQ score of 130 is not "30% smarter" than someone with a score of 100 — a score of 130 puts a person in the highest 2% of the population, while a score of 100 sits at the 50th percentile. The number is a comparison, not a magnitude.

And the FSIQ — like any single cognitive score — says very little about character, temperament, or life outcomes beyond the cognitive domain. A score in the superior range signifies strong reasoning and problem-solving capabilities, but it dictates very little about a person's character, temperament, or emotional life. What it does tell you, reliably, is where someone's overall pattern of performance on a specific set of cognitive tasks falls relative to a representative sample of their peers — which, read correctly, is genuinely useful information. Read as anything more than that, it starts to mislead.

The Takeaway

The Full Scale IQ is a real, carefully constructed composite score — not an arbitrary number, and not a mystical one either. It's built layer by layer from subtest scores through index scores, standardized against a representative sample, and reported with a known margin of error. Reading it well means treating it as a range rather than a point, treating the classification label as a convenience rather than a verdict, and looking at the index profile underneath it whenever the indices themselves diverge.

If you want to see your own profile — index scores, FSIQ, and the confidence interval around it, reported the way a professional report would present them — the RIOT was built to give you that full picture, not just a single number.