Factors influencing strategy choice in memory-based multi-attribute decisions

研究多属性决策的主要范式是研究人员向参与者提供所有必要的信息。然而，许多决策需要从长期记忆中检索相关线索信息。只有少数研究考察了这些基于记忆的决策和所使用的信息整合策略。结果好坏参半：一些人发现忽略某些信息的简单词典启发式占优势，而另一些人则发现主要是考虑所有加权线索信息的补偿策略。在研究中，这可以通过改变信息的表示格式来操纵，但在研究之间，这种差异的原因仍然难以捉摸。此外，在最近的一项研究的复制中发现了基于范例的推理，该研究最初没有发现范例效应。然而，所有的研究都使用了不同的材料、范式、线索格式等。该项目旨在阐明可能影响策略选择的几个认知因素的作用。从文献的结果中提取了一个初步框架，产生了对潜在重要变量的可测试假设，这些变量在6个实验中进行了其他条件的检验。

Multi-attribute decisions have mainly been studied with a paradigm in which researchers provide participants with all necessary information. However, many decisions require retrieval of relevant cue information from long term memory. Only a handful of studies examined these memory-based decisions and the information integration strategies used. The results have been mixed: Whereas some found a preponderance of simple lexicographic heuristics that ignore certain information, others found mainly compensatory strategies which take all weighted cue information into account. Within studies, this can be manipulated by varying the representation format of information, but between studies, the reasons for this discrepancy remain elusive. Furthermore, exemplar-based reasoning has been found in a recent replication of a study that did not find exemplar-effects originally. However, all studies used different materials, paradigms, cue formats etc. The project aims to illuminate the role of several cognitive factors that may influence strategy selection. A preliminary framework was extracted from the results in the literature yielding testable hypotheses about potentially important variables that are subjected to ceteris paribus tests in 6 experiments.