The flexibility of models of recognition memory

尽管经过了几十年的研究，但在一组相互竞争的识别记忆模型中，哪一个模型能提供最充分的数据解释，仍然是一个有争议的问题。造成这种状况的几个因素之一是，所使用的模型选择措施只考虑参数的数量，而不是模型的功能形式。我们打算制定措施的模型的灵活性，考虑功能形式的最小描述长度原则的基础上。由此产生的模型选择措施将是有用的，在未来使用这些模型的所有研究，而不仅仅是在识别记忆领域。在本建议中，我们专注于发展模型选择措施的信心评级数据收集的绝大多数研究领域。在最后一步中，新方法将应用于识别记忆的现有数据的元分析。

Despite decades of research, it is still a contested question which member of a set of competing models of recognition memory provides the most adequate account of the data. One of several factors contributing to this state of affairs is that the used model-selection measures only consider the number of parameters, but not the functional form of the models. We intend to develop measures of model flexibility that take functional form into account based on the minimum-description length principle. The resulting model selection measures will be useful for all research using these models in the future, not only in the field of recognition memory. In the present proposal, we focus on developing the model-selection measures for confidence rating data as collected by the vast majority of studies in the field. In a final step, the new methods are to be applied in a meta-analysis of existing data on recognition memory.