How a co-actor's (Un-) reliability modulates goal selection in a novel joint goal-setting paradigm

Sociomotor theory - an extension of ideomotor theory - suggests that actions can also be represented in terms of the effects they elicit from others. But what if those others violate one's action effect anticipations? Here, we introduce a novel joint goal-setting paradigm to investigate effects of co-actors' occasional and overall unreliability on an individual's goal selection. In a first step, the participant moved a target halfway from the bottom center to the top left or right corner of the computer screen. In the second step, the co-actor moved the target to its final left or right position. In a learning block, the co-actor always continued the participant's target movements. In the test block(s), the co-actor produced congruent action effects in 50% (unreliable) vs. 80% (reliable co-actor) of the trials. Experiment 1 consisted of one (between-participants), Experiment 2 and 3 of two (within-participants) test blocks; in Experiment 3, the co-actor changed between blocks. Results of Experiments 1 and 3 reveal that participants repeated their corner choice more often after incongruent trials, but only when the co-actor was generally reliable. Implications in terms of sociomotor action control and joint action are discussed.