Implicit memory and aging: Generation effect in a word-stem completion test

Two experiments were designed to test the preservation of implicit memory with age and the generation effect in young, middle-aged, and older subjects. During an incidental-learning study phase, subjects read words presented normally (reading condition) or with two reversed and underlined letters (generation condition). In the latter condition, subjects had to mentally exchange the two underlined letters (EXAMELP for EXAMPLE). Implicit (word-stem completion) and explicit (free recall and recognition) tests of memory were administered, in that order. In both experiments, preservation of implicit memory with age was observed, unlike explicit memory as measured by the recognition test. The results of the first experiment, which used a standard no-competition procedure, indicated a generation effect on explicit memory but not on implicit memory. In the second experiment, a new procedure (that favored competition between items) was used. Subjects had to produce four nouns for each word stem. Results indicated a positive generation effect on both the implicit and explicit memory tests. This was the first demonstration of a generation effect in a word-stem completion test. The concept of trace strength was proposed to account for the findings.