The role of stimulus-response binding and retrieval in instrumental learning

该项目系统地探索了情景捆绑和提取过程在工具性学习中的作用。我们在第一个资助期的研究结果表明，用练习法则(Thorndike，1898/1911)描述的短期权变学习效应可以用基于新近的情景提取过程(新近法则；Giesen，Schmidt，&Rothermund，2020)完全解释。在第二个资助期，我们将调查情景提取过程对长期工具性学习(颜色词权变学习，反应序列的内隐学习)的贡献。在第一个工作包中，我们将使用一种新的范式(交替块范式；Rothermund&Kapinos，2021)来系统地评估颜色词权变学习任务中的长期学习效果。这一范式允许我们区分基于情景提取(基于频率的提取)和非情景学习形式(例如，联想形成、统计学习；Miller等人，2019；Theeuwes，2019)的长期学习解释。在第二个工作包中，空间连续反应时间任务(SRTT；Nissen&Bullemer，1987)将被用来研究情景反应-反应(RR)提取(Moeller&Frings，2019a，b)在内隐序列学习中的作用。SRTT的一个新的变种(Rothermund，Moeller，&Sprengholz，2021)允许我们评估基于最近和基于频率的情景RR提取过程对反应序列的内隐学习的贡献。该项目的最高目标是进一步明确这个研究单位的总体模型(BRAC模型)。具体地说，我们介绍了基于最近和基于频率的检索的区别，并将研究短距离和长距离SR和RR情节提取过程的定性特征。计划中的研究对于将BRAC模型的范围扩展到长期学习现象(权变学习、内隐序列学习)也是至关重要的。除了理论意义外，我们的研究对于我们理解习惯也具有显著的实践意义。

This project systematically explores the role of episodic binding and retrieval processes for instrumental learning. Our findings from the first funding period revealed that short-term contingency learning effects that have been described in terms of the law of exercise (Thorndike, 1898/1911) can be fully explained by processes of recency-based episodic retrieval (the law of recency; Giesen, Schmidt, & Rothermund, 2020). In the second funding period, we will investigate the contribution of episodic retrieval processes to long-term instrumental learning (color-word contingency learning, implicit learning of response sequences).In a first work package, we will employ a novel paradigm (the alternating blocks paradigm; Rothermund & Kapinos, 2021) to systematically assess long-term learning effects in a color-word contingency learning task. This paradigm allows us to distinguish between explanations of long-term learning that are based on episodic retrieval (frequency-based retrieval) vs. non-episodic forms of learning (e.g., association formation, statistical learning; Miller et al., 2019; Theeuwes, 2019).In a second work package, a spatial serial reaction time task (SRTT; Nissen & Bullemer, 1987) will be used to investigate the role of episodic response-response (RR) retrieval (Moeller & Frings, 2019a, b) in implicit sequence learning. A newly developed variant of the SRTT (Rothermund, Moeller, & Sprengholz, 2021) allows us to assess the contribution of recency-based and frequency-based episodic RR retrieval processes to an implicit learning of response sequences.The superordinate aim of the project is to further specify the overarching model of this research unit (BRAC model). Specifically, we introduce a distinction between recency-based and frequency-based retrieval, and we will investigate the qualitative characteristics of episodic retrieval processes for SR and RR episodes across short and long distances. The planned studies are also crucial for extending the scope of the BRAC model to phenomena of long-term learning (contingency learning, implicit sequence learning). Besides the theoretical implications, our research is also of eminent practical relevance for our understanding of habits.