Slide Slam N14
Behavioural mechanisms in response planning during social interaction
Andrew H. Son1, Cindy Hamon-Hill1, Christina Torrealba1, Aaron J. Newman1; 1Dalhousie University
Turn-taking occurs when two people in conversation produce alternating short bursts of speech. Since pauses between two turns are short, some researchers have suggested that response planning occurs as soon as listeners have heard enough to formulate a response (i.e. early response planning); however, others advocate that response planning occurs only near the end of the other person’s speaking turn (i.e. late response planning). The main research question of our study was, when listening to an incoming question uttered by a speaker, do listeners prepare their response as soon as a critical piece of information (i.e. critical words) is given to them in the question, or do listeners wait until the very end of the incoming turn to prepare their response? By combining an interactive quiz paradigm and finger-tapping task, we created an online behavioral experiment measuring tapping performance as a proxy of response planning during turn-taking. The experiment consisted of two tasks: Tap and Respond and Tap and Listen. In Tap and Respond, participants heard spoken stimulus questions and responded verbally to them while continuously tapping a complex 4-key pattern throughout the whole experiment. In Tap and Listen, everything was the same, except participants were asked not to answer the questions. This allowed for a comparison of tapping metrics when participants planned a response and when they only listened. For both tasks, pairs of questions were created. Each question pair consisted of an early condition version (where the critical word was placed in the middle of the sentence) and a late condition version (where the critical word was placed at the end of the sentence). For example, a pair of target questions that was used in the study was: (i) Early condition: What is the fastest animal on this planet? (ii) Late condition: On this planet, what is the fastest animal? The primary dependent measures were tapping rate and accuracy. The two competing models of response planning suggested two alternative hypotheses. The early response planning model predicted that tapping rate and accuracy would decline as soon as the critical word was heard in a question (regardless of whether the critical word came early or late in the sentence), only in the Tap and Respond task. Conversely, the late planning model predicted that tapping performance would decline only at the end of the sentences (regardless of the position of the critical word), but again only in the Tap and Respond task. As of this writing data collection is ongoing, with 32 usable data sets (6 rejected due to overall poor task performance). The results to date show a trend towards slower tapping rates at the ends of sentences in the late condition relative to the early condition, only in the Tap and Respond task. This is consistent with the early planning hypothesis, since late planning would predict similar effects at the end of the sentence regardless of the position of the critical word.