The IQ Bell Curve: Mean, Spread and Reference Scores

A bell curve is a visual model of how values can be distributed. In a normal distribution, values near the center are more common than values far from it. The center and the spread are separate settings.

Two numbers define the reference model

The mean locates the center. The standard deviation describes the spread. A familiar IQ reference model uses mean 100 and standard deviation 15. Changing the mean moves the curve sideways; increasing the standard deviation makes it wider.

The height of the curve is a density, not the percentage of people with exactly that score. To discuss proportions, consider an area under the curve across an interval. Confusing height with percentage is a common chart-reading error.

Follow a marker from the axis

On starp's homepage, the example marker reads You · 114. Start at 114 on the horizontal axis, then follow the marker upward. It sits to the right of the reference mean of 100.

The marker is an illustration available before you take a test. It is not a score assigned to you. It also does not show the empirical distribution of starp's calibration results.

A bell shape is not a validation study

A website can draw a smooth curve from a mathematical formula without collecting any assessment responses. To interpret it as evidence about actual people, you need to know how the observations were collected and whether the model fits them.

Even a real dataset with a mean of 100 need not be normally distributed. It might have two peaks, many tied scores or an unusual concentration near the endpoints.

Read charts with three questions

• Is this a mathematical reference or observed data?
• What do the horizontal values and vertical height represent?
• Which population or sample, if any, supports the comparison?

These questions apply beyond IQ charts. They help you distinguish a useful explanatory picture from evidence for a specific claim about a test.

Explore the curve with the IQ percentile calculator, or compare two IQ scores to see why equal score gaps can cover different amounts of the distribution.