Memory Accessibility and Task Involvement As Factors in Choice

ABSTRACT - This paper extends the current body of research on the importance of memory in consumer choice tasks. Subjects' ability to use prior information and their willingness to do so were manipulated by varying the accessibility of memory information and task involvement, respectively. Accessibility had a consistent impact on measures of memory use and choice outcomes. Involvement interacted with accessibility to affect memory use and time taken to make a choice but not choice outcomes. These findings were interpreted as reflecting subjects' differential attention to retrieval and choice processing caused by varying involvement.


Sarah Fisher Gardial and Gabriel J. Biehal (1985) ,"Memory Accessibility and Task Involvement As Factors in Choice", in NA - Advances in Consumer Research Volume 12, eds. Elizabeth C. Hirschman and Moris B. Holbrook, Provo, UT : Association for Consumer Research, Pages: 414-419.

Advances in Consumer Research Volume 12, 1985      Pages 414-419


Sarah Fisher Gardial, University of Houston

Gabriel J. Biehal, University of Houston


This paper extends the current body of research on the importance of memory in consumer choice tasks. Subjects' ability to use prior information and their willingness to do so were manipulated by varying the accessibility of memory information and task involvement, respectively. Accessibility had a consistent impact on measures of memory use and choice outcomes. Involvement interacted with accessibility to affect memory use and time taken to make a choice but not choice outcomes. These findings were interpreted as reflecting subjects' differential attention to retrieval and choice processing caused by varying involvement.


Recent studies have shown the importance of memory phenomena in consumer decision making (Bettman and Park 1980a, Johnson and Russo 1984, Biehal and Chakravarti 1982). One common aspect of this work has been assessing the impact of different memory representations on consumers' ability to retrieve and process product information in a choice-related task. For example, the levels of processing notion (Craik and Lockhart 1972) suggests that information processed to a greater depth, perhaps in a more elaborated manner, will be more easily retrieved. Hence it should be more likely to figure in subsequent choice processing. However, actual use of memory information in decision making is likely to be a function not only of a consumer's ability to retrieve and use it but also of the consumer's willingness or motivation to use it. Thus both ability and motivation-related factors should affect a consumer's use of memory information in choice tasks.

This paper presents the results of a preliminary investigation of this issue. The "ability" aspect it examined was degree of accessibility of memory information (Tulving and Pearlstone 1966, Tulving and Psotka 1972, Biehal and Chakravarti 1983). Memory information may differ in accessibility tue to the way it was initially encoded. As a result, less accessible information should be harder to retrieve from memory to employ in a consumer choice task. However, a consumer's attempts to overcome memory retrieval difficulties may vary according to his/her motivation to perform the task. Even information which is relatively difficult to retrieve might be accessed during choice if the consumer is sufficiently motivated. The motivation-related factor used in this study, consumer task involvement, provided a useful theoretical foundation in this context. In summary, the purpose of the study was to examine the impact of memory accessibility (two levels) and consumer task involvement (two levels) on consumer choice processing.


Most memory models postulate that concepts, represented as nodes, are joined by associational links into a network (Collins and Loftus 1975). Learning or encoding information involves establishing links between the material to be learned and the existing conceptual network in memory. Retrieval of information from memory requires determination of an "entry point", or retrieval cue, into the network. The retrieval cue used to enter the network may be generated internally by the consumer or taken from the external environment. To the extent that retrieval cues have been previously encoded with the to-be-remembered information, memory retrieval should be facilitated. Otherwise, information stored in memory may be less accessible (Tulving and Thomson 1973). Thus information availability in memory is a necessary but not a sufficient condition for its successful retrieval and subsequent use (Tulving and Pearlstone 1966).

Differences in memory accessibility may be caused by the way information was originally learned. A variety of learning tasks may be performed in a consumer context. One possibility is directed learning, i.e., learning product information is the primary focus of processing (Bettman 1979). Information so acquired may be subsequently recalled and combined with external information to make a choice. Such information should be highly accessible: a large number of potential retrieval cues strongly linked with existing memory structures may be created. Consequently the likelihood that some cues are available at retrieval is enhanced. Other consumer tasks may not have learning as their primary focus. Instead information may be acquired incidentally (Eagle and Leiter 1964). For example, consumers may classify brands based on a comparison of their performance to a known reference level, perhaps their most preferred brand. During classification several processes may occur. e.g., forming overall evaluations or making pairwise comparisons. While the end result is a brand classification, a by-product may be the retention and availability in memory of product information- However, two differences from directed learning are likely to exist. First, incidentally generated links are likely to be weaker. Second, there may be fewer retrieval cues encoded with the information used in classification. Thus information encoded incidentally is less accessible than information encoded by directed learning.

[Defined as the number of occasions the message recipient relates the message content to his/her own personal experiences. In this study the definition was extended to include relating stimulus information to the needs of a "friend" (see Involvement Manipulation).]

In a subsequent choice task that requires combining memory and external information on a set of brands lower accessibility may result in less memory information being retrieved and an associated reduction in the time spent making a choice (latency). Choice outcomes could also be affected: they would be based to a greater extent on the external information when memory accessibility is low, whereas when accessibility is high choice should be based on both memory and external information. Depending on the brand-attribute values of information from the two sources. choice outcomes may differ.

What is the impact of differences in task involvement on these processes? It is widely accepted that higher involvement is positively associated with increased cognitive processing. For example, involvement's impact on connections (Krugman 1966) and cognitive responses (Petty, Ostrum and Brock 1981) has been established. Both processes relate directly to the accessibility discussion because they are founded on the accessibility of existing associations avail. in memory. Thus it seems likely that higher involvement would be associated with more attempts to retrieve memory information (Mitchell 1981). Because more information will be used, making a choice should take longer. However, the extent to which higher involvement enhances choice processing is likely to depend on the accessibility of memory information. When accessibility is high memory information should be retrieved with little effort and differences in involvement should not cause significant choice processing differences. However, when accessibility is lower more effort will be needed to retrieve memory information. In this case differences in involvement are likely to have an impact, with higher involvement leading to increased retrieval and choice processing. Thus in a low accessibility situation highly involved consumers should make significantly more attempts to retrieve and use memory information during choice: their choices should be based on both memory and external information. In contrast, low involved consumers will be less inclined to use low accessible memory information: their choices should be based on external information to a greater degree. Thus choice outcomes should depend on the interaction between memory accessibility and task involvement.

This discussion is summarized in the following hypotheses:

H1: The use of memory information (a) increases with memory accessibility; (b) increases with higher task involvement only when memory accessibility is low.

H2: Choice outcomes are determined by the interaction of memory accessibility and task involvement.

H3: The time taken to make a choice (latency) (a) increases when memory accessibility is higher (2) increases with higher task involvement only when accessibility is low.



Subjects were given information on five brands described on three attributes. Some subjects learned the information (high memory accessibility), while others performed a structured brand classification task (low accessibility). Both groups were then given information on four new attributes for the same five brands. Using experimenter specified attribute importance weights, they chose the best brand based on all the information they had encountered. Some subjects were asked to imagine they were choosing a brand for a valued friend. They were also told they would receive feedback on their choice "quality" (high involvement condition). Other subjects were just told to use the importance weights as a guide (low involvement).

Stimulus Design

The stimulus information is shown in Figure 1. Thirty-five millimeter cameras were chosen as the product of interest for two reasons. First the product needed to be sufficiently complex to warrant a fairly extensive information processing task. Second, a pretest indicated that students were reasonably familiar with the product category and had some knowledge of the attributes relevant to choosing a 35mm camera.



The attributes and their values came from two sources: (1) the 1980 Consumer Reports Buyers' Guide article on 35mm cameras, and (2) a free elicitation which asked students what characteristics they would consider if they were purchasing a 35mm camera. The sever attributes selected included four Consumer Reports attributes listed with high frequency in the free elicitation (exposure control, weight, picture clarity and convenience rating) and three attributes not mentioned in Consumer Reports but which the students listed with high frequency number of additional lenses, durability and the amount of accessory equipment). Brands were labelled A through E to avoid confusion with actual products. The meaning of the attributes and possible values on each attribute were carefully described to subjects during the experimental procedure.

For the study the stimulus matrix was divided into two parts. The first part ("prior information", contained in the dotted lines of Figure 1) contained five brands described on three attributes. Prior information was used to create memory representations that differed in accessibility. The second part ("new information") used the same five brands described on four additional attributes, and a list of importance weights for each of the seven attributes.

Brand attribute values and importance weights were set to maximize the likelihood that the processing phenomena hypothesized to exist would affect choice outcomes. Thus if only new information was used to make a choice, brand B would be the likely choice: it equalled brand D on two attributes and performed better on the remaining two. For prior information the attribute weights and brand-attribute values were set so that: (1) the three attributes all rated high in importance, thus prompting subjects to use them; (2) the attribute importance weights defined a hierarchy, with durability first (score 10), followed by convenience (9) and number of extra lenses (8); and (3) the further the subject processed "down the hierarchy", the greater the likelihood that brand D (defined as the target brand) would be chosen instead of brand B.

Memory Accessibility Manipulation

High accessibility subjects were given the prior information matrix and learned it in anticipation of a recall test (directed learning). To discourage evaluative processing during learning (1) subjects were not given the attribute importance weights, and (2) brand-attribute values were assigned so that no clearly superior brand existed.

In contrast low accessibility subjects performed a classification task using the prior information. Along with attribute values on the five alternative brands, each was given a "prototype brand" which was rated on the same three attributes as the alternatives. The task required subjects to compare each brand in the matrix with the prototype brand and to classify it into one of three groups depending on the number of values it received which were "greater than" the prototype's values. This task was constructed with four considerations in mind. First, the classification rules were such that they required the subject to examine every brand-attribute value in the matrix to perform the task correctly. Thus any difference in the subsequent use of prior information could not be attributed to non-exposure to the information. Second, the groups were labelled 1,2, and 3 to minimize any connotation that one group was superior to another and to eliminate premature evaluations and retrieval cues which might have resulted from groups labelled "satisfactory" and "unsatisfactory". Third, based on a pre-test it was determined that by repeating the classification task (with different prototypes but the same stimulus matrix), the percentage of recall of information for low accessibility groups would be (1) markedly less than for high accessible groups but (2) at a level above zero recall such that meaningful use of prior information could be expected. Two classifications were judged sufficient to achieve this discrimination. [In pretests the average percent recall accuracy after two classifications was 36.6%, compared to 87.52 for directed learning.] Finally, to lessen the chance that these classifications yielded overall evaluations that could be used as retrieval cues in the subsequent choice task, the prototypes were defined so that brands were assigned to different groups in the two classifications.

Task Involvement Manipulation

The involvement manipulation occurred just prior to the brand choice task. It was designed to increase the degree to which subjects made connections during choice. High involvement subjects were told to think of a good friend or very important person in their lives and to choose a 35mm camera for that person. They were instructed to imagine that this person had provided them with the importance weights on all seven attributes, so that they would be able to choose a best brand based on their friend's importance weights or preferences.

Subjects were asked not to consider their own personal importance weights or preferences, but to try to choose the best brand for their "friends". Also, they were informed that after they had arrived at a decision, they would receive feedback on their "choice quality", i.e., whether they had actually chosen the brand which best fit their friend's preferences. Low involvement subjects were given neither of these instructions. They were simply told to make a brand choice based on the given importance weights. No mention of feedback was made

These procedures were followed for two reasons. First, the suggestions of a best friend or important person was intended to help subjects visualize the choice task in terms of personal importance or connections (Leavitt et al. 1981). It was hoped that this would personally involve or interest the subject more so than (1) a choice with no recipient in mind, and (2) importance weights which were given no additional justification or relevance beyond their function as a decision "guide." Second, the feedback manipulation was added after a pretest found greater involvement for subjects who knew that they would be getting immediate feedback about the "quality" of their decision compared to those who did not.

Experimental Procedure

At each stage of the experimental procedure subjects were given written instructions, and any questions or problems were handled immediately by the experimenter.

Subjects first read and signed an informed consent form. Then they were given a warm-up for the concurrent verbal protocol procedure required during choice. To do this subjects talked into a tape-recorder about the last time they made a housing choice. When subjects appeared comfortable verbalizing, they were given a sheet of paper introducing them to the product category used in the study. This sheet described the three attributes contained in the prior information, together with possible brand values on each attribute.

After reading this material, the encoding manipulation followed. High accessibility subjects were given the prior information and asked to learn it, whereas low accessibility subjects performed the classification task. There was no time limit for either encoding task. After completing the encoding all materials were removed.

Subjects then received additional material that described the four new attributes and possible values. After reading this information subjects were given a sheet containing the full matrix (Figure 1), except that prior information values were missing. They were then asked to use all the information they had encountered, together with the importance weights, to choose a best brand. At this point they were given the involvement manipulation. Then they were instructed "to describe out loud your thoughts and reactions, much like you did before with the housing decision."

After making their choice, subjects completed a recall task. The questionnaire asked them to recall all five brand scores on the seven attributes. A five minute time limit was set for this task. The recall test was followed by a true-false recognition test. For the 35 brand attribute values, 17 were true and the remainder false. All subjects received the same recognition task, which was scored for the percentage of correct responses. Finally, subjects completed a debriefing questionnaire. On average, the experimental procedure took about 40 minutes to complete.


The concurrent verbal protocols were first transcribed, then phrases were coded [Coding was done by only one of the authors (GB). Studies using more complex tasks and protocol coding schemes (e.g., Bettman and Park 1980a, Biehal and Chakravarti 1982) report intercoder reliability scores of 78-90%.] for two types of processing: (1) connections, and (2) use of prior (memory) information. To measure connections three codes were defined, based on Krugman's (1966) original guidelines: (1) attempts to visualize how stimulus information related to a third party's [High involvement subjects made a choice for a "friend". Low involvement subjects were given no context.] or the subject's lifestyle, attitudes, occupation or other personal characteristics; (2) attempts to visualize ownership, use, acquisition or knowledge of the products by a third party or the subJect; (3) statements that identified a specific third party and/or a situation for product use. Each subject's three scores were then summed into a measure of total connections, which was used as a manipulation check.

To measure memory use two measures were developed. The first contained four codes, taken from Bettman and Park's (1980b) coding scheme, that measured the extent to which each of the three memory attributes were used during choice: (1) statements of original brand-attributes values for a single brand, e.g., "Brand C had 50 additional lenses", (2) "recoded" statements of original brand-attribute values for a single brand, e.g., "Brand C had a lot of lenses", (3) general statements about how selected brands performed on an attribute, e.g., "Brands C and D were best in additional lenses" 9 and (4) statements about the range of values on an attribute across the set of brands, e.g., "The additional lenses ranged from 5 to 50." Subjects' counts on the four codes were summed for each of the three memory attributes. In addition a count was made of the number of brands out of the five available for which some memory information was used to make a choice, and the number of memory attributes that were retrieved and used.

The second memory use measure was a more general categorization of retrieval behavior in choice. After coding for memory use, the protocols were re-examined for one of the following types of behavior: (1) subject tried and succeeded in retrieving information on a specific attribute; (2) subject tried but did not succeed in retrieving prior information on a specific attribute; (3) subject tried but did not succeed in retrieving prior information in general [This code was assigned when subjects made statements like: "I can't remember the old information", i.e., retrieval for non-identified attribute was inferred to have occurred.]; (4) subject made no attempt based on the protocol to retrieve prior information or there was no mention of memory information. Responses were scored as 0-1 dummy variables for all eight items. [Items (1) and (2) were scored for each of the 3 memory attributes, i.e., six scores were created, plus the two general items, (3) and (4).]


The 48 subjects were students in the first author's upper elective marketing course, given at a major southwestern university. Students received class participation points for their help. Pretests of experimental materials were made using student volunteers in other marketing sections, none of whom provided the primary study data. Subjects were randomly assigned to the four experimental groups, with 12 subjects per group. Most of the subjects were women (52.1%), and their mean age was 23 years.


Task Perceptions

Several debriefing items were used to measure subJects' perceptions of the experimental procedures and task. All subjects considered both the five camera "brands" and the choice task to be quite realistic. Consistent with experimental procedures, all subjects reported they based their choices on importance weights provided to them: their own personal preferences for the attributes did not affect their final choices to a great extent. Finally, all subjects considered that the attributes used to define the hypothetical brands were clearly presented and consequently understood. These perceptual data were thus consistent with the methodological needs of the study.

Manipulation Checks

The effectiveness of the accessibility manipulation was assessed by (1) self report data and (2) subjects' performance on the recall and recognition tasks. High accessibility subjects scored significantly higher in response to a question about how much use was made of prior information (F(1,44)=42.60, p < .0001). They also reported that prior information was easier to use when making a choice (F(1,44)=36.18 p < .0001) and that it helped them select their best brand to a greater degree (F(1,44)=75.44, P < .0001) Finally subjects were asked to indicate if their final choice depended more on prior, new or a combination of the two information sources. For the high accessibility group 62.5% said it depended more on prior information and 29.2% on both sources. Low accessible subjects stated that their final choice depended more on external information (87.5%).

The effectiveness of the accessibility manipulation was also assessed by examining the increment in retrieval performance for recall versus recognition (Tulving and Psotka 1971). Because the accessibility effect is due to retrieval failure (an appropriate retrieval cue is less likely to be available) subjects' retrieval performance should improve if a suitable cue is presented. This was the pattern of results obtained (Table 1).



For high accessible subjects the increment from recall to recognition for prior information averaged 7.8%. When accessibility was low the increment averaged 36.1%, or approximately five times higher. An analysis of the difference scores showed a significant main effect for accessibility (F(1,44)=22.76, p < .0001). The interaction was not significant (F(1,44)=1.06, p > .20). These results, together with the perceptual data indicated that the accessibility manipulation yielded theoretically needed differences in memory representations. [Because low accessible subjects had recognition rates lower than would be expected by chance (i.e., 50%) it is possible that the manipulation partially confounded the amount of information in memory and its accessibility. This possibility had been considered in the manipulations pretest, which showed significantly higher recall for two repetitions (36.6%) compared to the level in the study (7.8% on average). The difference in the two scores could not be explained.]



The involvement manipulation was also supported, but less clearly (Table 2). High involvement subjects on average made a total of 1.58 connections, compared to 0.63 for low involvement subjects (F(1,44)=4.64, p < .10). Table 2 also shows how the two groups compared on the three types of connections coded in the study. The first two items were in the expected direction but the first was not significant (F(1,44)=1.09, p >.10) and the second only marginally so (F(1,44)=2.69, p=.10). Finally, the last item was in the predicted direction and statistically significant (F(1,44)=4.63, p < .05).

Use of Memory Information in Choice

Hypothesis H1 stated that the use of memory information during choice would increase as a function of accessibility of memory information. However, the increase in memory use with higher involvement levels would occur only when accessibility was low. Table 3 shows the total use of each prior (memory) attribute in choice, the average number of memory attributes used and the total number of brands that were referenced in the context of memory processing. Higher accessibility was positively associated with subjects' processing of information on durability (F(1,44)=5.59, p < .05), the number of additional lenses (F(1,44)=27.31, p < .001) and the number of brands (F(1,44)=15.06, p < .001). Two weak interactions were also fount: for durability and convenience information the data showed that at low levels of accessibility the involvement level was positively associated with memory use. This was consistent with the hypothesis. Unexpectedly, when memory information was highly accessible, the relationship was reversed. For durability and convenience the interactions were significant (F(1,44)=3.27, p < .10 and F(1,44)=3.99, p < .10, respectively) but not for the additional lenses attribute (F(1944)=0.0, p >.20). In the latter case higher involved subjects made greater use of lens information than low involved subjects (F(1,44)=3.03, p < .10). Finally, Table 3 shows that the average number of memory attributes used was significantly higher for highly accessible information (F(1,44)=21.70, p < .0001) but the accessibility-involvement interaction was not significant (F(1,44)=1.50, p > .20).



Table 4 shows subjects' retrieval behavior based on the general categorization of memory use inferred from the concurrent verbal protocol. The table is broken down into three levels showing the number of subjects who (1) attempted retrieval of brand values on one or more attributes (either successfully or unsuccessfully); (2) attempted retrieval in general, i.e., no attribute was identified or (3) made no retrieval attempts. More subjects used information on durability, convenience and additional lenses when memory information was accessible (estimated chi-squared from a log-linear model equalled 11.88, 9.05 and 11.00, respectively, p < .001 in all cases, for one degree of freedom). Conversely, more subjects made no attempt at all to use memory information (line C) when it was less accessible (estimated chi-squared-5.17, df=19 p < .05). Also, when involvement was higher the number of subjects who made no attempt to retrieve memory information tended to be lower (estimated chi-squared=2.65, p=.10). In summary, these results gave strong support for the main effect of accessibility on memory use stated in Hypothesis H1. However, support for the interaction hypothesis was less clear.



Choice Outcomes and Choice Latency

Hypothesis H2 stated that choice outcomes would reflect an interaction between memory accessibility and task involvement. To test this hypothesis choices were divided into two groups, depending on if the subject chose or rejected the target brand. A log-linear model was then used to test for the hypothesized effect. Table 5 shows the frequency of target brand choices by condition.



These data do not support the hypothesized interaction between the two treatments (estimated chi-squared=0.12, df=1, p >.20). The main effect of encoding was significant, however (estimated chi-squared=13.74, df=1, p<.001) . Hypothesis H2 must therefore be rejected.

Hypothesis H3 predicted that the time taken to choose (choice latency) would increase with higher memory accessibility but that the increase caused by higher task involvement would occur only for low accessibility. Table 5 shows that choice latency, averaged across involvement, increased from 3.33 minutes for low accessibility to 3.50 minutes for high accessibility subjects, but this was not significant (F(1,44)=0.13, p >.20). The increased choice latency at low memory accessibility from 2.58 to 4.08 minutes conformed to predictions, but the unexpected decline at high accessibility, from 3.53 to 3.17 minutes, did not. The treatment interaction was significant (F(1,44)=5.29, p<.05). Taken together these results only partially support Hypothesis H3.


It had been expected that at high accessibility, involvement would not affect several choice processing measures. Instead some declines occurred, i.e., when memory information was more accessible, higher involved subjects' overall processing was less. The reason for this finding was hard to determine. It is possible that depending upon memory accessibility, involvement affected the emphasis placed on different choice-related processes. Thus low accessible subjects found that important memory information was hard to retrieve and use during choice because of its initial encoding. Without a strong retrieval effort, induced by higher involvement, their choices would be based more on readily available but less important external information. However, if higher involvement enhanced efforts to retrieve important memory information, more would be available and used during choice. Hence the increased processing across involvement when accessibility was low.

In contrast, subjects with more accessible memory information would have less need to focus on retrieval processes. Instead they could direct their attention to other processes, e.g., choice structuring and simplification. (Johnson and Russo (1984) report a positive correlation between prior knowledge and the use of phased decision strategies). When involvement was higher, more effort may have been applied to simplifying because subjects were more concerned about making a good choice. As a result high involved/accessible subjects would simplify the task more effectively than less involved/accessible subjects and hence need to do less processing. At the same time, because more memory information was available they would still do more processing of it than highly involved/less accessible subjects. This latter group may have spent more time on retrieval and reconstructive processing of memory information with the result that their choice latency exceeded that of high involved/accessible subjects. Thus these results may reflect involvement's differential impact on subjects' attention to retrieval and choice simplification processes caused by variations in memory accessibility (cf. Mitchell 1981. Petty, Cacioppo and Schuman 1983).

These issues clearly warrant further research. Future work could examine the impact of several levels of memory accessibility instead of the two used here. Also, a "stronger" involvement manipulation should be developed. For two reasons manipulating involvement is often problematic in lab situations: (1) because it is often a novel experience, study participating may overwhelm experimentally induced involvement processes; (2) the involvement manipulation procedures may be transparent to subjects and hence not credible. Regardless of these de sign problems, the procedure used in this study to check the involvement manipulation - searching for evidence of involvement-created "traces" during choice (Batra and Ray 1983) - seems more promising than the often sole reliance on self report measures. Acute, often quite transitory, consumer differences should provide an interesting extension to our understanding of memory factors in consumer decision making.


Batra, Rajeev and Michael L. Ray (1983), "Operationalizing Involvement as Depth and Quality of Cognitive Response," in Advances in Consumer Research Vol. 10, eds. Richard P. Bagozzi and Alice M. Tybout, Association for Consumer Research, Ann Arbor, MI, 309-13.

Bettman, James R. (1979), An Information Processing Theory of Consumer Choice, Reading MA: Addison-Wesley.

Bettman, James R. and C. Whan Park (1980a), "Effects of prior Experience and Phase of Choice Process on Consumer Decision Processes: A Protocol Analysis," Journal of Consumer Research, 7 (December), 234-48.

Bettman, James R. and C. Whan Park (1980b), "Implications of a Constructive View of Choice for Analysis of Protocol Data: A Coding Scheme for Elements of Choice Processes," in Advances in Consumer Research Vol. 7, ed. Jerry C. Olson, Association for Consumer Research, Ann Arbor, MI, 148-53.

Biehal, Gabriel J. and Dipankar Chakravarti (1982), "Information Presentation Format and Learning Goals as Determinants of Consumers' Memory-retrieval and Choice Processes," Journal of Consumer Research, 8 (March), 431-41.

Biehal, Gabriel J. and Dipankar Chakravarti (1983), "Information Accessibility as a Moderator of Consumer Choice," Journal of Consumer Research, 10 (1), 1-14.

Collins, Allan M. and Elizabeth F. Loftus (1975), "A Spreading Activation Theory of Semantic Processing," Psychological Review, 82 (6), 407-28.

Craik, Fergus I.M. and Robert S. Lockhart (1972), "Levels of Processing: A Framework for Memory Research," Journal of Verbal Learning and Verbal Behavior, 11, 671-84.

Eagle, Morris and Eli Leiter (1964), "Recall and Recognition in Intentional and Incidental Learning," Journal of Experimental Psychology, 68 (1), 58-63.

Johnson, Eric J. and J. Edward Russo (1984), "Product Familiarity and Learning New Information," Journal of Consumer Research, Vol. 11 (1), 542-50.

Leavitt, Clark, Anthony G. Greenwald and Carl Obermiller (1981), "What is Low Involvement Low In?" in Advances in Consumer Research Vol. 8, ed. Kent B. Monroe, Association for Consumer Research, Ann Arbor, MI, 15-19.

Mitchell, Andrew A. (1981), "The Dimensions of Advertising Involvement," in Advances in Consumer Research Vol. 8, ed. Kent B. Monroe, Association for Consumer Research, Ann Arbor, MI, 25-30.

Petty, R.E., J.T. Cacioppo, and D. Schuman (1983), "Central and Peripheral Routes to Advertising Effectiveness: The Moderating Role of Involvement," Journal of Consumer Research, Vol. 10 (2), 135-46.

Petty, R.E., Thomas M. Ostrum and Timothy C. Brock (1981), Cognitive Responses in Persuasion, Hillsdale NJ: Lawrence Erlbaum Associates.

Tulving, Endel and Zena Pearlstone (1966), "Availability versus Accessibility of Information in Memory for Words," Journal of Verbal Learning and Verbal Behavior, 5, 381-91.

Tulving, Endel and Joseph Psotka (1971), "Retroactive Inhibition in Free Recall: Inaccessibility of Information Available in Memory Store," Journal of Experimental Psychology, 87 (1), 1-8.

Tulving, Endel and David M. Thomson (1973), Encoding Specificity and Retrieval Processes in Episodic Memory," Psychological Review, 80, 352-73.



Sarah Fisher Gardial, University of Houston
Gabriel J. Biehal, University of Houston


NA - Advances in Consumer Research Volume 12 | 1985

Share Proceeding

Featured papers

See More


Format Neglect?: How Different Rank Claim Formats Influence Preference

Julio Sevilla, University of Georgia, USA
Mathew S. Isaac, Seattle University
Rajesh Bagchi, Virginia Tech, USA

Read More


E7. Pronouns in Fundraising Appeals – The Impact of I vs. S/He on Donations

Amir Sepehri, Western University, Canada
Rod Duclos, Western University, Canada
Hamid Elahi, Western University, Canada

Read More


How Temporal Separation in Budgeting Affects Spending Behavior

Yuna Choe, Texas A&M University, USA
Christina Kan, Texas A&M University, USA

Read More

Engage with Us

Becoming an Association for Consumer Research member is simple. Membership in ACR is relatively inexpensive, but brings significant benefits to its members.