Do We See It or Not? Sensory Attenuation in the Visual Domain

Sensory consequences of an agent's actions are perceived less intensely than sensory stimuli that are not caused (and thus not predicted) by the observer. This effect of sensory attenuation has been discussed as a key principle of perception, potentially mediating various crucial functions such as agency and the discrimination of self-caused sensory stimulation from stimuli caused by external factors. Precise models describe the theoretical underpinnings of this phenomenon across a variety of modalities, especially the auditory, tactile, and visual domain. Despite these strong claims, empirical evidence for sensory attenuation in the visual domain is surprisingly sparse and ambiguous. In the present article, the authors therefore aim to clarify the role of sensory attenuation for learned visual action effects. To this end, the authors present a comprehensive replication effort including 3 separate, high-powered experiments on sensory attenuation in the visual domain with 1 direct and 2 preregistered, conceptual replication attempts of an influential study on this topic (Cardoso-Leite et al., 2010). Signal detection analyses were targeted to distinguish between true visual sensitivity and response bias. Contrary to previous assumptions and despite high statistical power, however, the authors found no evidence for sensory attenuation of learned visual action effects. Bayesian analyses further supported the null hypothesis of no effect, thus constraining theories that promote sensory attenuation as an immediate and necessary consequence of voluntary actions.