Associative spread as a mediating variable in the generation effect.

Memory retention is greater for information that is subject-produced rather than experimenter-provided; this enhancement has been labelled the generation effect. One explanation of this phenomenon, the Lexical Activation Hypothesis, assumes that lexical representation is a necessary condition for emergence of the effect. A corollary to this hypothesis, the Lexical Spread Hypothesis (Nairne, Pusen, & Widner, 1985), emphasises associative spread as a mediating variable. This single-factor explanation has relied for its support exclusively on recognition performance. In the present study, four experiments examined the viability of this hypothesis, under recognition and free-recall conditions. A two-factor explanation is proposed to account for the findings of the present study. It appears that associative spread does not mediate the generation effect, when a recognition test is administered, but may mediate free-recall performance, where retrieval networks play a more significant role in memory performance. An alternative explanation, based on the benefits to memory retention obtained from disruptions of familiar orthographic patterns, is proposed to account for the recognition findings.