Memory-Based Product Judgments: Effects of Presentation Order and Retrieval Cues

Jong-Won Park, Korea University
ABSTRACT - This paper investigated the way in which the presentation order of information and a retrieval cue influenced product judgments when they were memory-based. Subjects in an experiment received a list of informational items with instructions to comprehend each item. The items were separately presented one at a time and the order of the presentation was varied. After receiving the information and a short delay, subjects were requested to form an overall evaluation of the product. At this time, some subjects received an informational item as a retrieval cue and others did not. Results suggested that the presentation order and the cue appeared to interactively influence recall of product information and hence, product judgments. Implications of these results for product impression formation and judgment are discussed.
[ to cite ]:
Jong-Won Park (1995) ,"Memory-Based Product Judgments: Effects of Presentation Order and Retrieval Cues", in NA - Advances in Consumer Research Volume 22, eds. Frank R. Kardes and Mita Sujan, Provo, UT : Association for Consumer Research, Pages: 159-164.

Advances in Consumer Research Volume 22, 1995      Pages 159-164

MEMORY-BASED PRODUCT JUDGMENTS: EFFECTS OF PRESENTATION ORDER AND RETRIEVAL CUES

Jong-Won Park, Korea University

ABSTRACT -

This paper investigated the way in which the presentation order of information and a retrieval cue influenced product judgments when they were memory-based. Subjects in an experiment received a list of informational items with instructions to comprehend each item. The items were separately presented one at a time and the order of the presentation was varied. After receiving the information and a short delay, subjects were requested to form an overall evaluation of the product. At this time, some subjects received an informational item as a retrieval cue and others did not. Results suggested that the presentation order and the cue appeared to interactively influence recall of product information and hence, product judgments. Implications of these results for product impression formation and judgment are discussed.

Consumers often make a product judgment some time after they have acquired product information. Thus, their judgment is memory-based, i.e., based on product information available in memory (Lynch and Srull 1982; Alba, Hutchinson, and Lynch 1991). It has been suggested that consumers are likely to retrieve only a subset of information available in memory (e.g., Wyer and Srull 1989). This is often attributed to consumers' lack of ability to access all of the information they have acquired and stored in memory, and to their lack of motivation to do so. When incomplete retrieval takes place, it tends to show recency effects, i.e., retrieval is likely to be biased toward information that was received relatively recently (e.g., Anderson and Hubert 1963). On the other hand, retrieval of information is often facilitated when an appropriate cue is provided (e.g., Keller 1987). Then, it seems reasonable to suppose that judgments based on retrieved information would be influenced by both the presentation order of information and retrieval cues at the time of judgments. However, the exact pattern of such effects is not clear. The present research provides some insights into this matter.

BACKGROUND

Effects of Presentation Order of Information in Judgments

When consumers form a product judgment based on memory, the evaluative implications of information that is accessed from memory plays an important role (e.g., Kisielius and Sternthal 1984, 1986; Reyes, Thompson, and Bower 1980). Since information retrieval is unlikely to be exhaustive, the relative accessibility of product information in memory is critical (Feldman and Lynch 1988). Easier access to negative information could lead to an unfavorable evaluation of the product, whereas easier access to positive information could result in a more favorable evaluation. This means that the same set of information of a product in memory might create different evaluations about the product depending on the relative accessibility of positive and negative informational items in the set.

One of the well-recognized phenomena related to the information's accessibility and judgments is the recency effect (Feldman and Lynch 1988; Wyer and Srull 1989). It refers to the possibility that a judgment is more heavily influenced by recent information in memory than by earlier one (e.g., Anderson and Hubert 1963; Lichtenstein and Srull 1987). Such effect is typically observed when a judgment is formed some time after information is received and thus has to be based on the evaluative implication of information in memory (e.g., Lichtenstein and Srull 1987). According to a general memory model proposed by Wyer and Srull (1986, 1989), recency effects are due to the relative retrieval advantage (or easier accessibility) of recent information compared to earlier one (see also Feldman and Lynch 1988 for a similar idea). When the processing goal at the time information is initially received is to simply learn the information (comprehension goal), each piece of information is encoded and stored as a separate unit into a "bin" in memory from the bottom up. If the contents of the target bin are searched at some later time, the search is assumed to proceed from the top down. However, memory search typically terminates without retrieving all items available in memory. This limited search is likely to result in a biased recall in favor of those near to the top of the bin (i.e., informational items that are received relatively recently). Consequently, judgments that are based on recalled information are likely to be primarily determined by the implications of those recent items (i.e., recency effects).

Recency effects in judgments are perhaps best documented in the study by Lichtenstein and Srull (1987). In their study, subjects received a list of both favorable and unfavorable behavior descriptions about a target person with an objective either to form an overall impression about the person or to comprehend the information. Some subjects received favorable information first and unfavorable information next, while others received the same set of information in the reversed order. After some delay, all subjects were asked to judge overall likability of the target person. Subjects who had received information with the comprehension objective showed strong recency effects in their likability judgments: the target's overall likability was made more favorably when favorable information was presented last than when unfavorable information last. However, such effects did not occur when subjects had the goal of impression formation at the time they received the information. Since there is often a delay between the time consumers acquire product information and the time they make a product judgment or purchase decision, as well as consumers typically learn various aspects of a product over time, recency effects are likely to occur in product judgment situations (Park and Hastak, forthcoming).

Effects of Retrieval Cues in Judgments

Realizing that information retrieval tends to be incomplete, marketing researchers have been interested in the way in which retrieval is assisted. One that has been widely investigated is the role of an external cue. The premise is that consumers may retrieve more information upon exposure to an external stimulus which activates memory nodes for product information. Several consumer researchers have found that such cues as verbal and visual information contained in the original ad facilitated retrieval of product information (e.g., Burke and Srull 1988; Costley 1992; Dick, Chakravarti, and Biehal 1990; Keller 1987, 1991; Lynch, Marmorstein and Weigold 1988). In the study by Keller (1987), for example, subjects were shown target ads in a competitive ad environment. Thus, the accessibility of the target information was not readily accessible. After a delay, subjects' memory for specific aspects of the target products were examined. Results showed that subjects who were provided with a retrieval cue produced a superior recall than those who were not given a cue.

Theoretically, a memory cue activates corresponding nodes in memory. From the activated memory nodes, some form of search for associated information through the network of memory traces takes place. This enhanced search by a cue results in an increased recall. However, it should be noted that cues' effectiveness is not always apparent. First, it may depend on the accessibility of the information to be recalled (c.f., Alba and Chattopadhyay 1985; Keller 1987). If the information is readily accessible, an external cue may offer little additional help because there is no need for the cue. However, a cue can successfully facilitate recall if the information is relatively difficult to access. In other words, cueing difficult-to-access information will have a facilitation effect on recall, whereas cueing easy-to-access information may not. Second, a cue may interfere with retrieval of some of information in memory. Perhaps the best known example of such interference effect is the phenomenon of part-category cueing effects (e.g., Alba and Chattopadhyay 1985; Rundus 1973; Slamecka 1968, 1969). It refers to a possibility that external cues for information pertaining to a category in memory may inhibit recall of other information in that category. Further, interference can also occur between categories when they compete each other for recall. That is, an external cue for one category may interfere in the recall of members in the other category. If the accessibility of each category differs, however, the interference might be asymmetrical. Specifically, cueing a more accessible category inhibits the recall of members in a less accessible category, whereas cueing a less accessible category has little influence on the recall of the members in a more accessible category (e.g., Alba and Chattopadhyay 1985).

The above discussion of presentation order of information and a retrieval cue has some interesting implications for product judgments. Suppose that subjects receive a set of positive information and then a set of negative information of a product (positive-negative order) with an objective of comprehending the information. When they are later asked to form an evaluation of the product, they will have to retrieve information from their memory to make an evaluation. Then, the negative information (which is encoded relatively recently) is more likely to be retrieved and used for judgments, compared to the positive information. Thus, evaluations are likely to be made unfavorably (recency effects). If a retrieval cue is available at the time of evaluation, however, different evaluations may be generated. If the cue is relevant for the information that was encoded relatively early (i.e., positive information), it will facilitate the recall of early information. Consequently, evaluations are likely to be more neutral. If the cue is relevant for relatively recent items (i.e., negative information) instead, it may not facilitate the recall of recent items because the items are readily accessible without the cue's help. However, evaluations would still be affected by the cue (i.e., becoming more negative in this case) because the cue may inhibit the recall of positive items. In short, both early- and recent-cue alter recency effects in judgments, but through different effects on recall. This possibility was examined in the study to be reported below.

METHOD

Overview

Subjects received a large number of informational items about an automobile and were asked to comprehend this information. The information presented was the same in all conditions. However, the order with which the information was presented was experimentally manipulated. That is, some subjects received negative items first and positive items later (positive-last conditions), while others received the information in the reversed order (negative-last conditions). After receiving this information and a brief delay, subjects were asked to form an overall (memory-based) evaluation of the automobile. At this time, another manipulation (retrieval cue factor) was introduced. Specifically, about half of the subjects were reminded of one of the informational items that they had received: either one of the items presented early (early-cue conditions) or one of the items presented later (recent-cue conditions). The remaining subjects did not receive such a retrieval cue item (no-cue conditions). Immediately after subjects formed and reported their evaluations, they were asked to recall as many informational items about the automobile they had received as possible.

Predictions

Since subjects were instructed to comprehend the informational items at the time they received the information, they presumably encoded and stored the items separately in memory without spontaneously forming an overall product evaluation (see Srull and Wyer 1989; Wyer and Srull 1989 for this assumption). Thus, when they were later asked to evaluate the automobile, they would have to retrieve the information from memory to use it as a basis for generating an evaluation. When there was no cue, we expected that information retrieval would be likely to be biased toward the "recent" items, and hence judgments to be subject to recency effects. That is, subjects in the positive-last conditions will produce more favorable evaluations of the product than those in the negative-last conditions. We also expected that providing a retrieval cue would alter this possibility through affecting retrieval. We first speculated that an early-cue would facilitate the recall of early items, without affecting the recall of recent items. Thus, recency effects in judgments would be mitigated. In contrast, a recent-cue would inhibit the recall of early items, without facilitating the recall of recent items. Consequently, recency effects would be strengthened. To state these predictions more formally:

H: When there is no cue available, product judgments would be influenced by presentation order (recency effects). If there is a cue which activates memory for early information, the judgments would be less affected by presentation order (less recency effects). This is through the facilitated recall of early information due to the cue. If a cue is relevant for activating memory for recent information, the judgments would be more heavily influenced by presentation order (stronger recency effects). This is through the inhibited recall of early information due to the cue.

Subjects and Design

One hundred and twenty-two male and female MBA students at a large university participated in the experiment as a part of a course requirement. Subjects were randomly assigned to each of the presentation order conditions (positive-last versus negative-last). Among them, 68 subjects received either an early-cue or a recent-cue, and the remaining subjects received no retrieval cue. All subjects received the same set of information which contained fifteen items total. These stimuli are described below.

Stimulus Materials

An automobile product was selected as the product to be evaluated. A large number of informational items were created on the basis of magazine advertisements and articles in Consumer Reports. The informational items described important attributes of an automobile such as power, comfort, safety, and economy. Based on a pretest, we finally selected seven positive items (e.g., "the anti-lock breaking system provides an excellent breaking power even at a slippery road" "a silent shaft reducing idling noise provides more interior quiet," etc.), five negative items (e.g., "the gas mileage is below average compared to other similar brands," "the interior space for the driver's seat is somewhat narrow," etc.), and three neutral items (e.g., "the industry average warranty period is provided," etc.).

Presentation Order

Subjects received a total of fifteen informational items. These items were presented in separate sheets with an interval of 15 seconds. The presentation order was systematically varied across subjects. First, we used ten items (five positive items and five negative items) to manipulate the presentation order. Some subjects received the positive items first and the negative items later (negative-last condition), while others received the information in the reversed order (positive-last conditions). In either case the serial position at which a given item was presented was counterbalanced across subjects. Thus, pooled over all subjects, the number of times a given item appeared at a particular serial position was the same for all of the ten items. Finally, we interspersed five filler items (two positive and three neutral items) before, after, and during the course of presenting ten test items. This was deemed necessary to make the presentation order manipulation less obvious (c.f., Lichtenstein and Srull 1987).

Retrieval Cue

After receiving the information and before making an overall evaluation of the product, subjects engaged in a 5-minute distraction task which was designed to eliminate a short-term memory effect. Then, an overall evaluation was requested and the manipulation of retrieval cue conditions was introduced. Specifically, some subjects were not provided with any retrieval cues at the time an overall evaluation of the product was requested (no-cue conditions). However, subjects in the cueing conditions received a retrieval cue. They were simply reminded of one particular item that was initially presented. Selection of the item was systematically varied depending on cueing conditions. For some subjects, the cue was one of the items that had been presented early (early-cue conditions). For others, it was the one that had appeared relatively recent (recent-cue conditions). However, selection of an item as the cue was counterbalanced across all conditions. In the cases of early-cue conditions, the cue always was the second item among those presented early. Similarly, the cue always was the forth item among those presented recent in the cases of recent-cue conditions. This ensured counterbalanced selection of an item as the cue because as explained previously, the serial position at which a given item was presented had been counterbalanced across subjects. Thus, pooled over subjects, the number of times a given item was used as the cue was the same for all of the ten items.

Procedure

The experiment was run during class time. At the beginning, subjects were told that (a) we were interested in how comprehension of product information would be influenced by the way in which the information was worded, (b) they would see product feature descriptions of automobile brands, (c) the brands were currently available in the market, and (d) that the feature descriptions were obtained from an article in Consumer Reports. We pretended that the informational items to be presented were about several brands of automobile. This was deemed effective to discourage subjects from forming an overall evaluation of the product on-line. Then, subjects were asked to rate each informational item in terms of how easily they could understand it. These instructions were followed by a list of the product information items. We presented the 15 informational items in the manner described earlier, one item at a time with an interval of 15 seconds.

After this and a 5-minute task unrelated to the present study, subjects were informed that the informational items they had received were actually about a single brand of an automobile, and then were asked to evaluate this brand in terms of overall quality. At this point, a retrieval cue was provided only for those in the cueing conditions. Subjects reported their evaluations on two 9-point scales ("bad"-"good", "dislikable"-"likable"). A recall task immediately followed this. Subjects were asked to recall as many of the informational items about the automobile as they could remember. Finally, subjects were debriefed and dismissed.

Scoring recalled items

Recalled items were scored according to a gist criterion by two independent judges. That is, items were scored as correct if they conveyed the same idea as the original item regardless of wording. In addition, partial credit was given when subjects recalled just the attribute name without specifying its evaluative content. Inter-judge agreement was 82.3%. (Disagreements were resolved through discussion.)

RESULTS

Overall Product Evaluations

We expected recency effects (or presentation order effects) on overall evaluations for no-cue conditions and recent-cue conditions, but not for early-cue conditions. To test this prediction, overall evaluations were analyzed as a function of presentation order and cueing.

Since the two scales used for measuring overall evaluations were highly correlated (r=.82), the average score was used for the analysis to be reported. Table 1 shows mean overall evaluation scores as a function of presentation order and cueing. A 2 by 3 ANOVA (presentation order by cueing) on these scores revealed a main effect for presentation order (F(1,115) =7.20, p <.01). As expected, subjects in the positive-last conditions reported more favorable evaluations than those in the negative-last conditions (5.63 versus 4.83). The presentation order by cueing interaction was not significant (F<1). However, comparisons which examined the effect of presentation order under each cueing condition separately provided support for our expectation. First, when no cue was provided, product evaluations were more favorable in the positive-last conditions than in the negative-last conditions, showing a marginal recency effect (t(115)=1.73, p<.10). Second, this recency effect was strengthened when a recent-cue was provided, confirming our expectation, (t(115)=1.99, p<.05). Finally, as expected, this recency effect was removed when an early-cue was provided, (t(115)=.868, p>.39). In sum, these results are consistent with the hypothesized effects of presentation order and retrieval cue on product evaluations.

Immediate Recall

Previous research suggests that immediate recall is a good indicator of the information that is retrieved and used for judgments especially when the judgments are computed based on information in memory (e.g., Lichtenstein and Srull 1985; see Hastie and Park 1986 for a review). Thus, we examined recall data to gain an insight into retrieval processes underlying recency effects we observed. Overall, relatively recent items were expected to be recalled better than earlier items. However, a retrieval cue would influence the recall. Specifically, it was expected that an early-cue would facilitate recall of early items without affecting the recall of recent items, but a recent-cue would inhibit recall of early items without helping recall of recent items to a meaningful extent.

TABLE 1

OVERALL PRODUCT EVALUATIONS AS A FUNCTION OF PRESENTATION ORDER AND CUEING

TABLE 2

IMMEDIATE RECALL SCORES AS A FUNCTION OF RECALL TYPE AND CUEING

In order to look at these hypothesized effects, we coded the recall of early items and that of recent items separately. The positive items recalled in the negative-last conditions and the negative items recalled in the positive-last conditions were counted for the recall of early items. The recall of recent items was calculated by counting the negative items recalled in the negative-last conditions and the positive items recalled in the positive-last conditions. Then, the recall scores were analyzed as a function of presentation order (positive-last/negative-last), cueing (no cue/recent-cue/early-cue), and recall type (early items/recent items). The last factor was a within-subject variable.

A 2 x 3 x 2 mixed ANOVA revealed three noticeable effects. First, a significant effect for recall type (F(1, 116)=11.17, p<.001) indicated a recency effect in recall. As expected, subjects recalled recent items more than early items (2.41 versus 1.80). Second, there was a recall type by presentation order interaction effect (F(2, 116)=28.16, p<.001). Early items were recalled better in negative-last conditions than in positive-last conditions (2.25 versus 1.36), whereas recent items recalled poorer in negative-last conditions than in positive-last conditions (1.36 versus 2.93). Interpreting differently, subjects recalled more positive items than negative items regardless of the presentation order. Finally, and more importantly, cueing by recall type interaction effect approached a significance (F(2, 116)=2.97, p <.06). Table 2 shows recall scores as a function of cueing and recall type. To further analyze this interaction, several mean comparisons were conducted on the recall of early items and that of recent items separately. First, providing a recent item as a cue did not significantly increase the recall of recent items (2.4 versus 2.5, t(119)<1), confirming our expectation. On the other hand, the recent-cue significantly suppressed the recall of early items (2.0 versus 1.2, t(119)=2.51, p<.05). This also confirmed our expectation. Finally, results in the early-cue conditions were mixed. That is, an early-cue did not suppress the recall of recent items as expected (2.4 versus 2.3, t(119)<1), it did not facilitate the recall of early items either (2.0 versus 2.1, t(119)<1), disconfirming our expectation. However, the difference in recall of early versus recent items was negligible in the early-cue conditions (F<1).

Further Analysis

The previous analysis of overall evaluations revealed that recency effects (or presentation order effects) occurred in no-cue and recent-cue conditions, but not in early-cue conditions. It was assumed that these effects were mediated by retrieval of information during judgments. To further assess this mediation process, the difference score between the recall of positive items and that of negative items was calculated to be used as a covariate. If our assumption about the mediation is correct, recency effects on evaluations observed in the previous analysis should be eliminated when the covariate is introduced in the analysis. In fact, this was the case. First, an ANCOVA on overall evaluations as a function of presentation order and cueing with the covariate revealed a significant effect of the covariate in no-cue conditions (F=18.25, p<.001), in recent-cue conditions (F=3.70, p<.06), and in early-cue conditions (F=10.893, p<.005). More importantly, when the covariate was introduced, the presentation order effects (or recency effects) that were previously observed were successfully removed in no-cue conditions (F=1.72, p>.19) and in recent-cue conditions (F<1). These results indicate that as expected, recency effects on evaluations were mediated by the information recall during judgments.

DISCUSSION

Overall, the study provides reasonable support for the proposition that presentation order and retrieval cues would influence retrieval of information and hence, product judgments. This proposition was supported in analyses for overall product evaluations and in the mediation test. That is, recency effects emerged in no-cue conditions and were strengthened in recent-cue conditions. As expected, however, cueing early items in memory successfully eliminated recency effects. Finally, ANCOVA results suggested that recency effects in evaluations were mediated by the net valence of the information retrieved during judgments.

The recall data provided support for the hypothesized effects of a recent-cue. As expected, a recent-cue did not help recall of recent items but inhibited the recall of early items to a meaningful extent. Consequently, subjects in the recent-cue conditions showed a stronger recency effect in evaluations, as compared to those in no-cue conditions. In contrast, an early-cue did not have such inhibition effect on the recall of recent items. These results are consistent with previous studies. For example, Alba and Chattopadhyay (1985) found that cueing a member in a dominant (or easier-to-access) category inhibited the recall of members in a dominated (or difficult-to-access) category, whereas cueing a member in a dominated category had little influence on the recall of the members in a dominant category. In our study, the recent items and the early items constituted a relatively dominant (or easier-to-access) category versus a dominated (or difficult-to-access) category in memory. This was because the evaluative implications of early items and recent items were oppositely valenced, thus encoded differentially, although the items within each category may not have been clustered in a conventional network form (c.f., Klein and Loftus 1990). Consequently, cueing the dominant category (recent information) suppressed the recall of the members in the dominated category (early information), whereas cueing the dominated category did not decrease the recall of members in the dominant category.

Unfortunately, results do not provide unambiguous evidence for the effects of an early-cue on recall and judgments. As expected, an early-cue successfully eliminated the recency effects in evaluations. This result paralleled the negligible difference in recall of early versus recent items. Further, the cue did not suppress the recall of recent items, consistent with the notion of asymmetric interference effects. However, it did not facilitate recall of early items either, failing to support our expectation. Perhaps, the power to detect this effect was weak. Nonetheless, future research is warranted to resolve this ambiguity.

Finally, it is worth noting that our results have interesting implications for marketing managers. For instance, when marketers want consumers to retrieve as many features of their products as possible, they need to carefully select retrieval cues. If the cues selected by marketers happen to trigger memory node for highly accessible information, they might actually impair the overall recall by consumers. This is because the cues may only suppress less accessible information, while not facilitating highly accessible information. Thus, it is suggested to use such cues that might activate memory for relatively less accessible information. Another implication is related to competition in a retrieval set of brand names. Marketers certainly want their brands to be included in consumers' retrieval set. Our results suggest that both major and minor brands can be benefitted from such marketing practice as a point-of-purchase display (but for different reasons) in increasing their inclusion in consumers' consideration set. Minor brands can be benefitted from such marketing practice because it increases the chance that the brands will be retrieved for consideration during a purchase decision. However, major brands can also be benefitted from it because it is likely to decrease the probability that other brands will be retrieved for consideration.

REFERENCES

Alba, Joseph W. and Amitava Chattopadhyay (1985), "Effects of Context and Part-Category Cues on Recall of Competing Brands," Journal of Marketing Research, 22 (August), 340-349.

Alba, Joseph W., J. Wesley Hutchinson, and John G. Lynch (1991), "Memory and Decision Making," in Handbook of Consumer Behavior, eds., Thomas S. Robertson and Harold H. Kassarjian, Englewood Cliffs, NJ: Prentice-Hall, 1-49.

Anderson, Norman H. and Stephen Hubert (1963), "Effects of Concomitant Verbal Recall on Order Effects in Personality Impression Formation," Journal of Verbal Learning and Verbal Behavior, 2 (5-6), 379-391.

Burke, Raymond, and Thomas Srull (1988), "Competitive Interference and Consumer memory for Advertising," Journal of Consumer Research, 15 (June), 55-68.

Costley, Carolyn L. and Merrie Brucks (1992), "Selective Recall and Information Use in Consumer Preferences," Journal of Consumer Research, 18 (March), 464-473.

Dick, Alan, Dipankar Chakravarti, and Gabriel Biehal (1990), "Memory-Based Inferences During Consumer Choice," Journal of Consumer Research, 17 (June), 82-93.

Feldman, Jack M. and John G. Lynch, Jr. (1988), "Self-Generated Validity and Other Effects of Measurement on Belief, Attitude, Intention, and Behavior," Journal of Applied Psychology, 73 (August), 421-435.

Hastie, Reid and Bernadette Park (1986), "The Relationship Between Memory and Judgment Depends on Whether the Judgment Task is Memory-Based or On-Line," Psychological Review, 93 (July), 258-268.

Keller, Kevin L. (1987), "Memory Factors in Advertising: The Effect of Advertising Retrieval Cues on Brand Evaluations," Journal of Consumer Research, 14 (December), 316-333.

Keller, Kevin L. (1991), "Memory and Evaluation Effects in Competitive Advertising Environment," Journal of Consumer Research, 17 (March), 463-476.

Kisielius, Jolita and Brian Sternthal (1984), "Detecting and Explaining Vividness Effects in Attitudinal Judgments," Journal of Marketing Research, 21, 54-64.

Kisielius, Jolita and Brian Sternthal (1986), "Examining the Vividness Controversy: An Availability-Valence Interpretation," Journal of Consumer Research, 12, 418-431.

Klein, Stanley B. and Judith Loftus (1990), "Rethinking the Role of Organization in Person Memory: An Independent Trace Storage Model," Journal of Personality and Social Psychology, 59 (3), 400-410.

Lichtenstein, Meryl and Thomas K. Srull (1985), "Conceptual and Methodological Issues in Examining the Relationship Between Consumer Memory and Judgment," in Psychological Processes and Advertising Effects: Theory, Research and Applications, eds., Linda Alwitt and Andrew A. Mitchell, Hillsdale, NJ: Earlbaum, 113-128.

Lichtenstein, Meryl and Thomas K. Srull (1987), "Processing Objectives as a Determinant of the Relationship Between Recall and Judgment," Journal of Experimental Social Psychology, 23 (March), 93-118.

Lynch, John G., Jr., Howard Marmorstein, and Michael F. Weigold (1988), "Choices from Sets Including Remembered Brands: Use of Recalled Attributes and Prior Overall Evaluations," Journal of Consumer Research, 15 (September), 169-184.

Lynch, John G., Jr. and Thomas K. Srull (1982), "Memory and Attentional Factors in Consumer Choice: Concepts and Research Methods," Journal of Consumer Research, 9 (June), 18-37.

Park, Jong-Won and Manoj Hastak (forthcoming), "Memory-Based Product Judgments: Effects of Involvement at Encoding and Retrieval," Journal of Consumer Research.

Reyes, Robert M., William C. Thompson, and Gordon H. Bower (1980), "Judgmental Biases Resulting from Differing Availabilities of Arguments," Journal of Personality and Social Psychology, 39 (1), 2-12.

Rundus, Dewey (1973), "Negative Effects of Using List Items as Recall Cues," Journal of Verbal Learning and Verbal Behavior, 12 (1), 43-50.

Slamecka, Norman J. (1968), "An Examination of Trace Storage in Free Recall," Journal of Experimental Psychology, 76, 504-513.

Slamecka, Norman J. (1969), "Testing for Associative Storage in Multitrial Free Recall," Journal of Experimental Psychology, 81, 557-560.

Srull, Thomas K. and Robert S. Wyer, Jr. (1989), "Person Memory and Judgment," Psychological Review, v.96, No.1, 58-83.

Wyer, Robert S., Jr. and Thomas K. Srull (1986), "Human Cognition in Its Social Context," Psychological Review, 93 (3), 322-359.

Wyer, Robert S., Jr. and Thomas K. Srull (1989), Human Memory and Cognition in Its Social Context, Hillsdale, NJ: Earlbaum.

----------------------------------------