On the retention function for autobiographical memory

College undergraduates were asked to record events from their lives, and then to date those events. Data were collected from groups of subjects using a set of cue words to prompt the events, from individual subjects, for individual cue words, from groups of subjects using no cue words, and from subjects who kept diaries. If it is assumed that the subjects encoded an equal number of events from each day of their lives, the distribution of events recorded as a function of time can be viewed as a retention function. The data from all experiments provided an excellent fit to the single-trace fragility function proposed by Wickelgren to account for more traditional laboratory learning experiments. Taken together these experiments indicate that the retention function is not an artifact of summing different functions produced by individual subjects or cue words and that the episodes recorded are, for the most part, accurately dated memories of actual events. Thus, episodic memory of a naturalistic, autobiographical nature and episodic memory for lists appear to have the same retention properties. © 1982 Academic Press, Inc.