Ostergaard A L
Department of Psychiatry, School of Medicine, University of California, San Diego, La Jolla 92093-0603.
J Exp Psychol Learn Mem Cogn. 1992 Mar;18(2):413-20. doi: 10.1037//0278-7393.18.2.413.
A large number of reports have been published on stochastic independence between implicit and explicit measures of memory. This is often taken to imply that different memory systems mediate implicit and explicit memory performance. In these cases, stochastic independence is inferred from contingency analysis of overall success rates in two memory tasks when performance in one or both of the tasks is, to a large extent, mediated by factors other than memory. Typically, the difference between performance with studied and nonstudied items is not large in implicit memory tasks. It is argued that this must be taken into account when evaluating the contingency analysis. A method is presented for estimating the relevant joint and conditional probabilities, assuming that the aspects of performance in the two tasks that are related to memory are dependent to the maximum possible extent. The method is applied to a number of published studies, and it is shown that the difference between these estimated probabilities and those given by stochastic independence is too small to allow any conclusion to be drawn about memory systems from contingency analysis of data reported in these studies.
关于内隐记忆和外显记忆测量之间的随机独立性，已经发表了大量报告。这通常被认为意味着不同的记忆系统介导了内隐记忆和外显记忆表现。在这些情况下，随机独立性是从两个记忆任务总体成功率的列联分析中推断出来的，而其中一个或两个任务的表现很大程度上是由记忆以外的因素介导的。通常，在内隐记忆任务中，学习过的项目和未学习过的项目之间的表现差异不大。有人认为，在评估列联分析时必须考虑到这一点。提出了一种估计相关联合概率和条件概率的方法，假设两个任务中与记忆相关的表现方面在最大可能程度上是相关的。该方法应用于一些已发表的研究，结果表明，这些估计概率与随机独立性给出的概率之间的差异太小，无法从这些研究报告的数据的列联分析中得出关于记忆系统的任何结论。
