How IQ Scores Are Calculated: The Deviation IQ Explained

The Basic Idea Behind an IQ Score

An IQ score is not a raw count of correct answers. It is a statistical comparison: your performance on a set of reasoning tasks is measured against the performance of a large reference group, often called a norming sample. The result is a single number that tells you roughly where you fall relative to that group, not an absolute measure of "how smart" someone is in any objective sense.

This approach is known as the deviation IQ, and it has been the standard method used by psychologists and test developers for decades. It replaced an older, cruder method that divided a person's "mental age" by their chronological age. The deviation IQ is more consistent across age groups and produces scores that behave predictably in a statistical sense.

Mean 100, Standard Deviation 15

When a test is normed, raw scores from the reference sample are converted onto a standardized scale with two fixed anchor points: a mean of 100 and a standard deviation of 15. The mean represents the mathematical center of the distribution — the average performance in the norming sample. The standard deviation describes how spread out the scores are around that center.

Because cognitive test scores tend to follow a roughly normal distribution (the familiar bell curve), this fixed mean and standard deviation let us describe, with reasonable precision, what proportion of people fall into each score range:

• About 68% of people score between 85 and 115 (within one standard deviation of the mean)
• About 95% of people score between 70 and 130 (within two standard deviations)
• About 99.7% of people score between 55 and 145 (within three standard deviations)

A score of 100 does not mean "perfect" or even "the target score" — it simply means average relative to the norming group. A score of 130, roughly two standard deviations above the mean, places someone in a small minority, statistically speaking, of the population that was tested.

What "Norming Against a Population" Actually Means

Norming is the process of administering a test to a large, carefully selected sample of people and using their results to build the scale that everyone else's scores will be compared against. The quality of a norm depends heavily on how representative that sample is — ideally covering a broad range of ages, backgrounds, and education levels, and large enough that the resulting distribution is statistically stable.

This has an important consequence: an IQ score is always relative to the group it was normed on. A score is meaningful only in the context of the reference population and the specific test used to generate it. Different tests, built and normed independently, are not perfectly interchangeable, even if both report scores on a mean-100, SD-15 scale.

It also explains why population norms need to be periodically updated. Average performance on cognitive tests has shifted over generations in a well-documented pattern known as the Flynn effect. Test publishers periodically re-norm their instruments so that a score of 100 continues to represent the current population's average, rather than the average from decades earlier.

Percentiles: Translating IQ Into "How Many People Score Lower"

Because the distribution is approximately normal, any IQ score can be converted into a percentile — the percentage of the reference population expected to score at or below that point. A score of 100 sits at roughly the 50th percentile, meaning about half the population scores lower. A score of 130 sits at roughly the 98th percentile, which is why that threshold is often cited as the entry requirement for high-IQ societies such as Mensa: it marks the top 2% of a normed distribution.

Percentiles are often easier to interpret intuitively than raw IQ numbers, since they directly express relative standing rather than an abstract scale value.

What This Scale Does and Doesn't Capture

The deviation IQ scale is a tool for organizing and comparing performance on reasoning tasks; it is not a direct measurement of a physical quantity, the way a thermometer measures temperature. Researchers often discuss a general reasoning factor, sometimes labeled g, that tends to correlate across different types of cognitive tasks, alongside more specific abilities such as fluid intelligence (reasoning with novel problems) and crystallized intelligence (accumulated knowledge and vocabulary). A single IQ score is a summary across whatever mix of tasks a given test includes, and different tests can weight these components differently.

It's also worth being clear about the limits of a free, online test like the ones on IQTesta. Our score is indicative, not a clinical assessment. A licensed psychologist administering a professionally validated instrument under standardized, controlled conditions produces a result with a different level of rigor and documentation than a self-administered online test taken at home. This distinction matters especially when evaluating children or assessing giftedness: an online score should never be treated as a diagnosis, and any concerns about a child's cognitive development or educational needs should be discussed with a qualified professional.