Go-activation endures following the presentation of a stop-signal: Evidence from startle

  • Neil M Drummond School of Human Kinetics, University of Ottawa
  • Erin K Cressman School of Human Kinetics, University of Ottawa
  • Anthony N Carlsen School of Human Kinetics, University of Ottawa

Abstract

Logan & Cowen (1984) proposed that in a stop-signal task (SST), independent go- and stop-processes "race" to control behaviour. If the go-process wins, an overt response is produced, while if the stop-process wins, the response is withheld. One prediction that follows from this proposal is that if the activation associated with one process is enhanced, it would be more likely to win the race. We looked to determine whether response outcome could be manipulated by using a startling acoustic stimulus (SAS), as a SAS has been shown to provide added response activation, resulting in the early release of a response. In the present study participants were to respond to a visual go-stimulus (green); however, if a subsequent stop-signal appeared (stimulus turned red) they were to inhibit the response. Participants completed a SST of 100-trials, including 25 trials where a stop-signal was presented at a delay corresponding to a probability of responding of 0.4 (determined from a previous baseline block of trials). On stop-signal trials a SAS was presented either simultaneous with the go-signal or stop-signal, or 100, 150, or 200ms following the stop-signal. Results showed that presenting a SAS during stop-trials led to an increase in probability of responding at all SAS presentation times after the go-signal. The latency of SAS responses at the stop-signal+150ms and stop-signal+200ms time points suggest that these responses would have been voluntarily inhibited but instead were involuntarily triggered by the SAS. Results demonstrate that go-activation endures even 200ms following a stop-signal and remains accessible well after the response has been inhibited.

Acknowledgments: Supported by the Natural Sciences and Engineering Research Council of Canada