Joint action changes valence-based action coding in an implicit attitude task.

Recent studies suggest that co-acting with another person induces a problem to discriminate between one's own and the other's actions which can be resolved by emphasizing action features that discriminate best between both persons' actions in a given task context. Mostly, overt action features like the spatial position of responses have been suggested as discriminating action features. In the present study, we tested whether non-externally perceivable, covert action features can be used for resolving the action discrimination problem during joint action. Therefore, we compared task performance between a joint and an individual version of the Go/Nogo Association Task, a task requiring the association of a valence to the response. We found a larger implicit attitude effect in the joint than in the individual setting for person-related (self and other, Experiment 1) as well as for non-person-related attitude objects (fruit and insect, Experiment 2) suggesting that the weight of valence information is increased in the internal coding of responses when valence discriminates between both responses. In contrast, we found a smaller implicit attitude effect in a person present setting than an individual setting (Experiment 3) indicating that the enhanced implicit attitude effect observed in the joint settings of Experiments 1 and 2 is not due to social facilitation. Our results suggest that action discrimination during joint action can rely on covert action features. The results are in line with the referential coding account, and specify the kind of action features that are represented when sharing a task with another person.