Systematic theory–method dependencies in consciousness research

<p>Scientific theories of consciousness are typically evaluated by comparing empirical findings across experimental paradigms. Such comparisons implicitly assume that different paradigms provide commensurable evidence and do not systematically favour particular theoretical interpretations. Here, we test this assumption using the updated ConTraSt database, comprising 518 published consciousness experiments, and Bayesian regression models relating methodological characteristics to theoretical interpretations. We find decisive evidence that experimental methodology predicts whether studies support, challenge, or do not mention major theories of consciousness. These associations are robust across multiple Bayesian model comparison approaches and remain evident across a range of prior specifications. Methodological effects differ substantially across theories, indicating that theories occupy partially distinct methodological niches rather than a common empirical space. While these findings do not imply that any individual theory is more or less valid, they suggest that empirical support is systematically shaped by methodological tradition. Ultimately, it can be challenged if neural correlates of consciousness are “studying the same things” – and if not, what they in that case are actually about. Consequently, direct comparisons between theories may be more constrained than generally assumed. Our results highlight the need for greater methodological diversity in theory testing and suggest that future progress in consciousness science may require approaches that explicitly address the relationship between experimental methodology and theoretical interpretation.</p>