Particular surface features of problems can become associatively linked with effective, wellmastered solution principles, thus modifying their application and producing faster solutions. In 3 experiments, 8 3-bit decision rules were used to solve word problems generated from experimentally restricted feature pools. The authors confirmed that frequency of exposure to surface features speeds problem encoding and categorization. Co-occurrence between particular problem features and a relevant rule (particularization) resulted in outracing of well-learned problem-solving processes by direct linkage between features and action choice. Particularization, an inductive process emerging during rule-guided problem solving, is proposed as a mechanism by which rule-guided decisions become automatized. In this way, memories of specific rule-application experiences supplant analytic processes in problem solving.

We report two experiments about how people estimate the frequency of event properties when they are explicitly (e.g., spinach-GREEN) and implicitly (e.g., spinach) presented. In Experiment 1, verbal reports indicated that, for explicitly presented properties, participants used several retrieval-and impressionbased strategiesand were relatively accurate. Implicitly presented properties led to off-target retrieval, which brought to mind more instances of nontarget than of target properties and degraded estimates. A third group estimated the frequency of taxonomic categories (e.g., furniture) much as the explicit property group did, suggesting that people can use properties to organize remembered events. In a second experiment, estimation time patterns underscored the results of Experiment 1 and eliminated reactive verbal reports as an explanation. Off-target retrieval was both ineffective and slow.

Two studies explored underlying differences between self- and proxy reports and determining whether proxy reporters process and retrieve information the same way self-reporters do. In both studies, participants recorded their own and a target person's incidental purchases for two weeks and then returned, either immediately or one-week after the diary-keeping period, for two recall tests. These tests pertained to self- and proxy purchases. The reported items were scored against the recorded ones. In Study 1, the data indicate that proxies tend to recall more typical events than do self-reporters immediately following the diary-keeping period. However, after a delay of one week, both proxy and self-reporters tended to report typical items. These findings suggest that, at least initially, self- and proxy reporters process the information differently but that after one week they process it similarly. Study 2 employed a method that was identical to that of Study 1 except that participants were given a recognition test. The verbal protocols from the recognition test indicate that participants used different strategies to determine who made the purchase. Whereas participants focused on perceptual details surrounding the event when identifying their own purchases, proxies engaged in various strategies to identify their spouses' purchases.

The authors confirmed E. Z. Rothkopf and M. L. Dashen's (1995) finding that specific problem context, such as thematic surface features, forms associative connections with deep problem features and thus speeds particular decisions (particularization). In 5 experiments, using a 3-bit decision task and pre-memorized decision rules, the authors found that the ability of a situational context to reinstate was decreased by its replacement by another modal surface context. Context reinstatement, as measured by decision speed, depended on both global and recent local densities of specific problem features linked to a particular decision. The authors' results are consistent with J. R. Anderson and L. J. Schooler's (1991) needs/odds analysis and suggest a push-down file model for diverse context influences as a mechanism for responding to changing situational demands.

No abstract