The Better-Than-Average Effect

What it is

The effect is a gap between what people say about their own standing and where the group's standing sits. The clearest evidence this publication holds comes from the 2020 study by Gilles Gignac and Marcin Zajenkowski. They had 929 adults from the community rate their own intelligence on a single scale and then sit a standard reasoning test scored against age norms. Self-rated IQ averaged about 124 on an IQ-like scale, against a measured average of about 102. The two measures correlated only modestly.

That gap is the better-than-average effect in one sample, and it is large. It matters in the Dunning-Kruger debate because of what happens when the sample is sorted into quarters by measured score. If everyone's self-rating is shifted upwards and the self-rating is only loosely related to the measured score, the bottom quarter will show the largest gap between belief and measurement, and the top quarter the smallest, with no deficit of insight needed to produce it.

Joachim Krueger and Ross Mueller named the pairing of this tendency with statistical regression in the title of their 2002 critique of the original Dunning-Kruger paper.

In effect

Gignac and Zajenkowski put numbers on the argument. Sorted into quarters, the gap between self-assessment and measurement in their data shrank from about 33 points in the bottom quarter to about 13 in the top, the familiar Dunning-Kruger shape. Two tests that do not depend on grouping people into quarters found no pattern, and a simulation produced the quartile pattern from data in which none had been built in. They concluded that the effect is mostly, though not entirely, a statistical artefact.

A replication by Curtis Dunkel and colleagues on a nationally representative American sample of 13,977 people found a significant but small effect even with these confounds in view. The better-than-average effect, then, does not erase the Dunning-Kruger pattern. It accounts for a large part of how that pattern looks in grouped data.

What it does not say

It does not say that everyone overrates themselves on every skill. The evidence on this page concerns a single measure of intelligence in one sample. Gignac and Zajenkowski noted that other abilities or experimental settings might behave differently.

It does not explain why people rate themselves above average. This page makes no claim about motive, ego protection or information, and the vault has not reviewed that literature. The note on this concept is marked unreviewed and the originator of the term is not established here.

Nor does it make the self-rating worthless. The two measures in the 929-person study correlated, modestly, so self-assessment carries some information about ability.

---