Description
Responses to bimodal signals are typically faster than responses to unimodal signals, aphenomenon known as the redundant signal effect (RSE). Traditionally, the RSE is studied using tasks that require participants to perform the same motor action when detecting targets in either modality. The RSE can be explained either by pooling models, proposing sensory evidence is integrated in a single decision unit with a shared response threshold to trigger the motor response, or by race models, proposing parallel processing in two decision units with separate thresholds. Distinguishing between these accounts has proven challenging, and particularly with respect to race models it would be beneficial to know whether a response to a bimodal signal was triggered by one or the other modality. To address this, we developed a new version of the task, requiring participants to press distinct buttons corresponding to the modality they detected first. In addition, we employed a signal onset asynchrony (SOA) manipulation as typically used in RSE experiments. In bimodal trials, we found that the proportion of button presses corresponding to a specific modality was predictably modulated as a function of SOA, confirming adherence to task instructions. Notably, the two-button task yielded RSE patterns consistent with predictions from the race account. Even though modalities were mapped to separate responses, responses violated Miller’s bound, a finding often interpreted as evidence for pooling. Our results suggest that introducing a two-button instruction can clarify the mechanisms underlying the RSE.
| Period | 18 Jul 2025 |
|---|---|
| Event title | 23rd International Multisensory Research Forum |
| Event type | Conference |
| Location | Durham, United KingdomShow on map |
| Degree of Recognition | International |