Response to comment on 'Criterion placement threatens the construct validity of neural measures of consciousness'.

Criterion shifts—changes in how people decide whether they have seen something—can distort neural measures of consciousness, causing simultaneous over- or under-estimation of both conscious and unconscious processing. While it was already known that subjective reports are vulnerable to such biases, new simulations and empirical work show that these effects have concrete consequences when analyzing real brain data. The authors argue that their findings are novel and that the experimental manipulations they used are not unrealistic compared to natural variations across experiments.

Study at a glance

Abstract

In Fahrenfort et al., 2025 we show the influence of non-perceptual criterion shifts on neural measures of consciousness. We fully agree (and point out in our article) that it was already known that subjective measures are sensitive to criterion confounds, and we are happy to read that this is acknowledged by Sandberg and Overgaard in their comment (Sandberg and Overgaard, 2025). However, we contest that the main findings of our simulations and empirical studies had already been demonstrated. Several findings from our studies are novel, such as the fact that criterion effects reveal themselves as over- (or under-) estimations of both conscious and unconscious processing in tandem, and that this has tangible implications when analyzing real neural data. We also challenge the suggestion that our experimental manipulations are (too) radical compared to signal-to-noise variations that occur naturally between experiments.