Stopping an action and discarding another action with(out) feature-overlap

This experiment will combine a stop-signal paradigm (Logan & Cowan, 1984; Verbruggen et al., 2019) with an ‘ABBA’ paradigm (Stoet & Hommel, 1999). In the latter, participants are asked to prepare an action A. Yet, before carrying out that action A another action B is requested. Initiating this first requested action B is typically delayed and more error prone if it does share some features with the planned action A, than when it does not, leading to so called partial overlap costs. These partial overlap costs are traditionally attributed to the incidental retrieval of features already bound for the currently not requested action A. We will study the stopping of highly prepared actions B briefly before execution. Specifically, the role of feature-based processes for stopping will be revealed. We will focus on how the concurrent planning of a feature-overlapping action A impacts the potential to stop an action B. In this second experiment, not only action B needs to be stopped (while action A still needs to be performed), but the planned action A is also to be canceled upon the presentation of a stop signal (i.e., no response is required in a stop trial). In the paradigm by Mocke et al. (2024), we will introduce a stop signal that appears after stimulus B in 25% of the trials and requires to withhold action B. Thus, responding to Stimulus B will be the Go task. The planned action A will be executed upon the presentation of a final go signal after the response limit for action B in stop trials or 500 ms after releasing the keypress B in go trials. This final Go-A signal will allow to distinguish correct actions A from fully overlapping, but unsuccessfully stopped actions B, which can be identified by coming before this final Go-A signal. According to the consensus guideline for stop signal experiments by Verbruggen and colleagues (2019), the following implementations will be made. The stop signal will be a salient stimulus, that is, a red rectangle. Stop signals will be presented in 25% of trials. The stop-signal delay (SSD) will be continuously adjusted via a standard adaptive tracking procedure, with steps of 50 ms. That is, after each unsuccessful stop trial (including premature responses and choice errors), the SSD will be increased by 50 ms (3 frames for 60 Hz screens) and after each successful stop trial decreased by 50 ms. Participants will be instructed to ‘respond as quickly as possible to the go stimulus and not wait for the stop signal to occur’. They will further be discouraged from using a waiting strategy (i.e. informed that the probability of an unsuccessful stop trial will approximate 0.50, and that SSD will increase if they gradually slow their responses). Participants will be given block-based feedback, informing them about their mean RT on go trials, number of go omissions (with a reminder that this should be 0), and p(respond|signal) (with a reminder that this should be close to .50). Various practice blocks will first teach Stimulus A and Stimulus B sets, followed by the classical ABBA task, and finally the stop-signal manipulation. Each participant will provide 768 trials, that is, 48 stop trials per side overlap x duration overlap condition. This will result in a 90 minute experiment.