Scholarlink

<p>Eliciting judgments of learning (JOLs) from participants has been shown to enhance memory recall on paired-associates tasks in some contexts, a <i>reactivity effect</i>. However, little is known about the effect of JOLs on forms of learning where the task requires generalizing beyond the training set, such as in category learning. Here, we explore the effect of JOLs on relational rule discovery using a categorization paradigm. In Experiments 1 and 2, where both a relational rule and visual stimulus features can be effectively used for categorization, we show that JOLs impair rule discovery but do not affect memorization of visual features. In Experiment 3, we modified the task such that only a relational rule could be used to categorize stimuli and observed no evidence of reactivity. We explain these findings using a conservative strategy-shift account of reactivity, which proposes that eliciting JOLs causes participants to shift strategies to utilize the more obviously rewarded strategy. Specifically, when participants have multiple viable strategies, JOLs shift participants’ categorization strategy away from rule discovery and instead encourage a more obvious strategy based on memorization of visual features.</p>

错误：搜索内容不能为空，请输入英文关键词
错误：关键词超出字数限制，请精简
高级检索

Judgments of learning impair rule-based discovery

  • 期刊论文

  • 出版模式：

    开放获取

  • 发表日期：

    2025年6月4日

查看全文
  • Kit S. Double,
  • Dominic Tran,
  • Micah B. Goldwater

摘要

Eliciting judgments of learning (JOLs) from participants has been shown to enhance memory recall on paired-associates tasks in some contexts, a reactivity effect. However, little is known about the effect of JOLs on forms of learning where the task requires generalizing beyond the training set, such as in category learning. Here, we explore the effect of JOLs on relational rule discovery using a categorization paradigm. In Experiments 1 and 2, where both a relational rule and visual stimulus features can be effectively used for categorization, we show that JOLs impair rule discovery but do not affect memorization of visual features. In Experiment 3, we modified the task such that only a relational rule could be used to categorize stimuli and observed no evidence of reactivity. We explain these findings using a conservative strategy-shift account of reactivity, which proposes that eliciting JOLs causes participants to shift strategies to utilize the more obviously rewarded strategy. Specifically, when participants have multiple viable strategies, JOLs shift participants’ categorization strategy away from rule discovery and instead encourage a more obvious strategy based on memorization of visual features.