How IQ Tests Are Scored: From Raw Answers to a Deviation IQ

An IQ score can look deceptively simple: one number, often somewhere near 100. But quite a lot happens between the moment you finish answering questions and the moment a figure appears on your screen. Understanding those steps makes the result easier to interpret and, just as importantly, easier to take with appropriate caution.

This article walks through the scoring process in plain terms, from raw answers to the final number, and explains why that number should always be read as a range rather than a precise verdict.

From Raw Score to a Normed Score

The starting point is the raw score: the count of items answered correctly, sometimes adjusted for item difficulty or the time taken. On its own, a raw score means little. Getting 32 out of 40 tells you nothing until you know how other people performed on the same items under the same conditions.

That comparison group is called the norming (or standardization) sample. Test developers administer the test to a large group meant to represent the relevant population by age, and sometimes by other factors, then record the distribution of raw scores. Your performance is interpreted relative to that sample. The quality of a score depends heavily on how recent, how large, and how representative this sample is. A test normed decades ago, or on a narrow group, gives less trustworthy comparisons.

The Deviation IQ and Percentiles

Modern tests use a deviation IQ. Raw scores from the norming sample are rescaled so the average is set to 100 and the standard deviation (a measure of spread) is set to 15. Your score then describes how far you sit from the average of your age group, in standardized units, rather than counting correct answers directly.

Because of this design, scores tend to follow a familiar bell-shaped distribution. That makes percentiles a more intuitive companion to the IQ number. A percentile tells you the share of the comparison group that scored at or below your level:

• A score of 100 sits at roughly the 50th percentile, near the middle of the sample.
• About two-thirds of people score between 85 and 115 (within one standard deviation of the mean).
• Around 95% score between 70 and 130 (within two standard deviations).
• Scores far from 100 are, by definition, much less common, which is why small differences in raw answers near the extremes can shift the number noticeably.

Why One Number Is Misleading

No test measures perfectly. Fatigue, mood, distractions, familiarity with the format, and ordinary chance all nudge a score up or down on any given day. Psychometricians capture this with the standard error of measurement, which is used to build a confidence range around your score.

In practice, that means a result is better read as a band than a point. A reported 112 might reasonably be expressed as roughly 106 to 118 at a common level of confidence. Two people with the same headline number can differ within that band, and the same person can land a few points apart on two sittings without anything meaningful having changed.

A few other limits are worth keeping in mind:

• A single composite IQ blends different abilities, such as verbal and spatial reasoning, that may not be equally strong in one person.
• Scores reflect performance on this test, on this day, against this sample, not a fixed, lifelong trait.
• Many things that matter for life and work, including motivation, knowledge, and circumstances, are not captured by the number at all.

If you are curious to see how this works in practice, you are welcome to try our test. It will give you a score with its percentile and a confidence range, presented as one snapshot of your reasoning on the day, nothing more and nothing less. Treat it as a starting point for reflection rather than a final word, and it can be a genuinely interesting thing to explore.