Valid population inference for information-based imaging: From the second-level $t$-test to prevalence inference

TL;DR

This paper critiques the use of second-level t-tests for population inference in neuroimaging classification accuracy, proposing prevalence inference as a more valid alternative supported by theory and simulations.

Contribution

It highlights the limitations of t-tests for information-like measures and introduces a prevalence inference method for more accurate population-level conclusions.

Findings

t-tests cannot reliably infer population effects for classification accuracy

Prevalence inference provides a better population-level understanding of effects

The proposed permutation-based method is effective on empirical data

Abstract

In multivariate pattern analysis of neuroimaging data, 'second-level' inference is often performed by entering classification accuracies into a $t$-test vs chance level across subjects. We argue that while the random-effects analysis implemented by the $t$-test does provide population inference if applied to activation differences, it fails to do so in the case of classification accuracy or other 'information-like' measures, because the true value of such measures can never be below chance level. This constraint changes the meaning of the population-level null hypothesis being tested, which becomes equivalent to the global null hypothesis that there is no effect in any subject in the population. Consequently, rejecting it only allows to infer that there are some subjects in which there is an information effect, but not that it generalizes, rendering it effectively equivalent to fixed-effects analysis. This statement is supported by theoretical arguments as well as simulations. We review possible alternative approaches to population inference for information-based imaging, converging on the idea that it should not target the mean, but the prevalence of the effect in the population. One method to do so, 'permutation-based information prevalence inference using the minimum statistic', is described in detail and applied to empirical data.