Abstract There is a long tradition of norming words and other materials for various descriptive properties (e.g., concreteness, emotional valence, imagery, meaningfulness), which allows those properties to be manipulated in memory experiments. We introduce a new approach to norming, in which measurement models are used to norm the underlying retrieval processes that memory items trigger. We report two worked examples of this approach, for the most commonly administered items in false memory experiments: DRM lists. In one project, 36 lists that had been previously normed for their levels of true and false memory were normed for their levels of 3 retrieval processes that control false memory (recollection rejection, phantom recollection, and semantic familiarity) and 3 parallel processes that control true memory (true recollection, erroneous recollection rejection, and semantic familiarity). In the other project, 72 new DRM lists, whose associative and semantic strengths had been systematically varied, were normed for the same retrieval processes. Together, these norms provide maximum likelihood estimates of all 6 retrieval processes for all 108 lists. Analyses of those estimates revealed that list differences in the three false memory processes are mutually dissociated, as are list differences in the three true memory processes. Thus, investigators can vary each retrieval process, without confounding it with other processes, by merely selecting appropriate groups of lists. That allows simple List × Treatment designs to be used to pinpoint the retrieval processes that are responsible for the effects of important memory manipulations (e.g., delay, divided attention, encoding instructions, repetition).

中文翻译：

---

规范检索过程

摘要 长期以来，人们一直在为各种描述性属性（例如，具体性、情感效价、意象、意义）规范词和其他材料，这允许在记忆实验中操纵这些属性。我们引入了一种新的规范方法，其中使用测量模型来规范记忆项目触发的潜在检索过程。我们报告了这种方法的两个工作示例，用于错误记忆实验中最常管理的项目：DRM 列表。在一个项目中，36 个先前已根据其真假记忆水平进行规范的列表，根据其控制错误记忆的 3 个检索过程（回忆拒绝、幻影回忆和语义熟悉度）和控制真假记忆的 3 个平行过程的水平进行规范。记忆（真实的回忆，错误的回忆拒绝和语义熟悉）。在另一个项目中，为相同的检索过程规范了 72 个新的 DRM 列表，它们的关联和语义强度已经系统地变化。这些规范一起为所有 108 个列表提供了所有 6 个检索过程的最大似然估计。对这些估计的分析表明，三个错误记忆过程中的列表差异是相互分离的，三个真实记忆过程中的列表差异也是如此。因此，调查人员可以改变每个检索过程，而不会将它与其他过程混淆，只需选择适当的列表组。这允许使用简单的列表×处理设计来精确定位导致重要记忆操作（例如延迟、注意力分散、