Department of Psychology, University of Michigan at Ann Arbor, 530 Church Street, Ann Arbor, MI, 48109, USA.
LEAD-CNRS UMR5022, Université de Bourgogne, Dijon, France.
Psychol Res. 2024 Jun;88(4):1182-1202. doi: 10.1007/s00426-024-01940-1. Epub 2024 Mar 14.
Contingency learning can involve learning that the identity of one stimulus in a sequence predicts the identity of the next stimulus. It remains unclear, however, whether such learning speeds responses to the next stimulus only by reducing the threshold for triggering the expected response after stimulus onset or also by preparing the expected response before stimulus onset. To distinguish between these competing accounts, we manipulated the probabilities with which each of two prime arrows (Left and Right) were followed by each of two probe arrows (Up and Down) in a prime-probe task while using force-sensitive keyboards to monitor sub-threshold finger force. Consistent with the response preparation account, two experiments revealed greater force just before probe onset on the response key corresponding to the direction in which the probe was more (versus less) likely to point (e.g., Up vs. Down). Furthermore, mirroring sequential contingency effects in behavior, this pre-probe force effect vanished after a single low-probability trial. These findings favor the response preparation account over the threshold only account. They also suggest the possibility that contingency learning in our tasks indexes trial-by-trial expectations regarding the utility of the prime for predicting the upcoming probe.
条件学习可以涉及到这样一种学习，即序列中一个刺激的身份预测下一个刺激的身份。然而，目前尚不清楚这种学习是否仅通过降低刺激后预期反应的触发阈值来加速对下一个刺激的反应，或者是否通过在刺激前准备预期反应来加速对下一个刺激的反应。为了区分这些相互竞争的解释，我们在启动-探测任务中操纵了两个启动箭头（左和右）在每个探测箭头（上和下）之后的概率，同时使用力敏键盘监测亚阈值手指力。与反应准备理论一致，两个实验在探测键上探测前呈现出更大的力，探测指向更有可能（而非更不可能）的方向（例如，上与下）。此外，与行为中的序列条件效应相呼应，在单一低概率试验后，这种预探测力效应消失了。这些发现支持了反应准备理论，而不是仅阈值理论。它们还表明，在我们的任务中，条件学习可能会对每个试验的预期进行索引，即启动对预测即将到来的探测的有用性的预期。
