MPT模型在事件性前瞻记忆研究中的应用

MPT Model in Event-Based Prospective Memory

前瞻性记忆是指对将要进行的活动或事件的记忆。前瞻记忆中,包含了前瞻成分和回溯成分。前人研究中,缺乏对前瞻成分和回溯成分有效的分离手段,使得对前瞻记忆机制的探讨缺乏深入挖掘。本文将MPT模型与Cohen等人的研究范式对比,分析了该模型在事件性前瞻记忆研究中的优势。文章对模型的理论基础,模型的主要内容,模型的数据计算方法,模型的效度以及模型的应用等几个方面进行了介绍,同时对模型使用中需要注意的问题进行了讨论。

Remembering to perform an action in the future is called prospective memory （PM）. It contains time-based prospective memory （TBPM） and event-based prospective memory （EBPM）. TBPM refers to remembering to perform an action at a specific time or after a certain amount of time has elapsed. EBPM refers to what must be performed when a certain event occurs. Both of these memories constitute a crucial form of memory use in our daily lives. This paper mainly introduces EBPM. PM contains two components, the prospective component and the retrospective component. Remembering that you have to do something is the prospective component, whereas remembering what you have to do and when you have to do it is considered the retrospective component. Unfortunately, if a variable affects prospective memory, we cannot determine how the variable affects each of the two different components using traditional accuracy measures. We cannot disentangle the two components clearly, and then we cannot explore the latent cognitive processes of prospective memory deeply in previous studies. This article presents a detailed discussion and application of a methodology by comparing the paradigm of Cohen to others in EBPM, called Multinomial process tree （MPT） models. In these models, it is assumed that there are discrete cognitive states that participants attain with certain probabilities during task performance. These probabilities are represented as model parameters that can be estimated from the observed raw data via the maximum likelihood parameter estimation. The fit of the resulting model to the empirical data can be evaluated via goodness-of-fit tests. The MPT model is relatively uncomplicated, does not require advanced mathematical techniques, and has certain advantages over other more traditional methods of studying cognitive processes. Smith and Bayen first introduced the MPT model for the measurement of EBPM. The model is based on the preparatory attentional processes and the memory processes （PAM） theory, which proposes that the prospective component involves processes that draw on our limited resources, and, thus, that these processes are not automatic. Now this theory is supported by many studies. The researchers first explained why we apply the model in EBPM, and then introduced the model in detail including the theoretical basis, the main content, the calculation method of the data, the validity and the application of the model. At the same time, the limitations in using the model of the problem are also discussed. Nonfocal tasks are those in which the PM cue is not part of the information being extracted in the service of the ongoing task. By contrast, focal tasks are those in which the ongoing task involves processing the defining features of the PM cue. In nonfocal tasks, prospective remembering is thought to require executive attentional resources to engage in extra monitoring for the cue to perform the intended action. And at the end of the article, we have a summary in full text and put forward some opinions on future research： for example, how we distribute the limited resources between ongoing task and PM task. Our goal in the work presented here was to develop and evaluate a formal mathematical model for the investigation of EBPM.