Still no Dunning-Kruger effect: A reply to Hiller

Abstract

The primary criticism raised by Hiller (2023) pertained to the transformation of the self-assessed intelligence (SAI) scores applied by Gignac and Zajenkowski (2020), an investigation that suggested the Dunning-Kruger effect (DKE) may be primarily a statistical artefact. Hiller recommended an alternative transformation. Our re-analysis of Gignac and Zajenkowski's data with Hiller's recommended transformation failed to find evidence in favour of the DK hypothesis. Hiller also recommended SAI measurement with a percentile-based approach. Based on a review of the literature, one investigation has employed a percentile-based approach to SAI measurement, and it largely failed to support the DK hypothesis, when analysed with the non-linear regression approach recommended by Gignac and Zajenkowski. We conclude by encouraging researchers to continue to derive novel approaches to testing the DK hypothesis, especially approaches that overcome known methodological challenges.