Information Processability and Restructuring: Consumer Strategies For Managing Difficult Decisions

Eloise Coupey, Virginia Polytechnic Institute and State University
Carol W. DeMoranville, Northern Illinois University
ABSTRACT - Our research is focused on consumer decision making when the available information is difficult to process. To better understand what constitutes processability, we proceed as follows: 1) examine consumers' strategies for managing decisions that are difficult because information is not easily comparable, and 2) assess the effect of display processability on aspects of consumer decision making, including outcome quality, recall, restructuring, and perceptions of the decision. The results of a study provide support for hypotheses about the effects of processability due to information comparability on the processes and outcomes of consumers' choices and judgments.
[ to cite ]:
Eloise Coupey and Carol W. DeMoranville (1996) ,"Information Processability and Restructuring: Consumer Strategies For Managing Difficult Decisions", in NA - Advances in Consumer Research Volume 23, eds. Kim P. Corfman and John G. Lynch Jr., Provo, UT : Association for Consumer Research, Pages: 225-230.

Advances in Consumer Research Volume 23, 1996      Pages 225-230


Eloise Coupey, Virginia Polytechnic Institute and State University

Carol W. DeMoranville, Northern Illinois University


Our research is focused on consumer decision making when the available information is difficult to process. To better understand what constitutes processability, we proceed as follows: 1) examine consumers' strategies for managing decisions that are difficult because information is not easily comparable, and 2) assess the effect of display processability on aspects of consumer decision making, including outcome quality, recall, restructuring, and perceptions of the decision. The results of a study provide support for hypotheses about the effects of processability due to information comparability on the processes and outcomes of consumers' choices and judgments.

Consumers are often faced with decisions that are difficult. Unclear preferences, too little (or too much) product information, and complex available information are characteristic examples of person, task and context factors that may increase decision difficulty (Payne, Bettman, and Johnson 1993). There is an abundance of research designed to further our understanding of the effects of these factors on consumer decision making. Most of this research, however, has focused upon the evaluative processes (e.g., choice heuristics) that consumers use to process presented information. In addition, the stimuli characteristic of this research are brand-attribute matrices in which the pieces of information are well-organized and comparable in form (Ford, et al. 1989). Less attention has been focused upon display characteristics that make decisions difficult, or on the processes by which consumers construct or alter displays to make difficult decisions easier (Coupey 1994). The explosion of information available to consumers in new media (e.g., interactive technologies, such as the Internet) emphasizes the need to understand how consumers manage decisions when information from marketers is readily available but is presented in different formats. For instance, if certain forms of information are easier to process than others, then search for similar information may receive disproportionate attention, weight, and processing. We examine consumers' use of information displays when decision making is difficult because characteristics of the information available in a display decrease the processability of that information.

In many real-life decisions consumers collect information from several sources about several brands. When they try to use the gathered information, they may find that not all brands are described in the same manner (e.g., warranty in weeks, months, or years), thus reducing the ability to compare brands attribute by attribute, a process often used for making choices (Biehal and Chakravarti 1982). Decision making can be further complicated if the product under consideration is unfamiliar to the consumer or if the available information is complex and hard to understand. Our research provides insights into the processes by which consumers attempt to make displays processable, and the effects of these processes on outcomes (e.g., preference judgments and choices), perceptions of the decision, and memory for decision information.

Understanding how the processability of information affects decision making is important for consumers, marketers, and public policy makers. Perceptions of difficulty due to low processability may influence the type and amount of information consumers use, as well as the manner in which they use it. By affecting the amount of effort consumers are willing to invest in decision making, processability may also influence the quality of the decisions and the amount, type, and organization of information stored in memory.

Marketers can use knowledge of systematic responses by consumers to variations in display processability to develop product information (e.g., brochures, product displays, packaging, and on-line displays). Public policy makers, working to maintain an equitable balance between consumers and marketers, can use knowledge obtained from basic research on adaptive display construction and use to maximize the benefits of information presentation for consumers.


Why Is Processability Desired?

Information that is easily processable, such as a list of unit prices, can help consumers make better decisions (Russo 1977). Processability can reduce the costs, such as cognitive effort, of acquiring and using information. It may also increase the benefits of consumer decisions, such as saving money or making choices of better quality.

An effort/accuracy framework has often been used to explain how processability affects information use and other aspects of decision behavior such as outcome quality (e.g., Kleinmuntz and Schkade 1993). An effort/accuracy rationale for the appeal of processable displays suggests that because processable information requires fewer operations to transform information to make comparisons, cognitive resources are saved in the acquisition and preparation of information. These resources could simply be conserved or they could be reallocated to another part of the decision process. For example, reallocation may be done to keep available a larger amount of information for further processing or to employ an evaluative strategy that offers greater outcome quality.

Two seemingly contradictory processes for information use have been proposed by decision researchers: 1) concreteness, the tendency to use information as presented (Slovic 1972), and 2) restructuring, the tendency to alter a display (Coupey 1994). These different explanations of behavior can be reconciled within an effort/accuracy framework in which task goals (e.g., judgments or choices) emphasize different levels of processability.

Dimensions of Processability

Russo, Krieser and Miyashita (1975) proposed that information must be both available and processable in order to receive maximum usage by consumers. Russo (1977) demonstrated that providing consumers with lists of brands and their unit prices increased consumers' use of unit price information and their savings. This research suggests that one dimension on which processability is effected is the comparability of information, such as the availability of cost in unit prices for all brands on a list. Consider the following scenario:

Early one morning while you are not yet fully awake, you turn on the shower as the first step in your pre-eight o'clock-class ablution. You drag yourself into the shower only to hop out hastily as the water temperature seems barely above freezing. You are now fully awake, so only a cursory check is needed to confirm your suspicion that the water heater died quietly in the night and a new one must be purchased.

You evaluate several water heaters, but because you have far more experience with the function of a water heater than with its features, you are not sure of the meaning of information on the yellow labels on the water heaters. The salesperson tells you that lower numbers are better, so you choose the water heater with the lowest numbers on the yellow label.

As you skim through Consumer Reports the next week, you notice that water heaters are reviewed. You are pleased to learn that the model you selected is the one favored by Consumer Reports, although you are still not sure what the numbers on the yellow labels mean.

This example illustrates the idea that it is possible for consumers to make optimal choices without understanding the information used as the basis for comparison. In other words, a consumer could be very confident that he or she has chosen the best brand, but have no idea of why it is good, or how it works.

Another dimension of processability is suggested by Bettman, Johnson and Payne (1991). They state that information use increases when the displays are well-organized and appropriately formatted. "Appropriately formatted' can be construed to include several conditions, such as the similarity of attribute value information, the availability of similar attribute information across brands, and/or the match between the organization of information in a display and how information is stored in memory.

That information use can be influenced by memorial information underscores another facet of processability: the degree to which it can be comprehended. As Russo's (1977) results indicate, merely making information readily comparable may often be a sufficient condition for an optimal choice. However, if the consumer must understand what the information means in order to successfully complete a task, then the concept of processability must be expanded to include characteristics of the decision maker, as well as the information. Suppose the consumer in the above example is now purchasing an automobile. The information on the sticker, such as mpg, may be more processable than that of the water heater because the consumer understands what the numbers mean. This suggests that processability may be influenced by both comparability and understandability.

While the two dimensions of processability may have similar effects on restructuring, decision outcomes, and perceptions of the decision, we limit our investigation in this paper to the effects of varying levels of comparability under conditions of low understandability. This constraint is imposed because we are interested in the effects of restructuring on memory for displayed information. Thus, we wish to avoid confounding subjects' current knowledge structures with information comparability effects. Future research should investigate the effects of varying levels of understandability and any interactions between the two dimensions of processability.

We suggest that although greater processability is preferred to lesser processability, the extent to which consumers are motivated to alter information displays to make information more processable, as well as the processes by which they do so, depends strongly upon the goal for which the information is used (e.g., choice versus judgment).

Judgment and Choice Effects on Restructuring

Consumers often make product decisions that take one of two forms: choices or judgments. While a choice is merely the selection of one brand, there are several types of judgments. For example, consumers may rank brands in order of desirability, rate brands to reflect their desirability or similarity, or provide attribute values to equate brands. Judgment and choice tasks lead to an often-replicated result; strategies for processing display information typically differ as a function of the task goal (Biehal and Chakravarti 1982). Process tracing data tend to suggest that consumers use within-attribute strategies for evaluating display information to make brand choices, but within-brand strategies for making judgments (Billings and Scherer 1988; Schkade and Johnson 1989).

We suggest that task goal not only influences processing strategies, but that it also affects the level of processability desired by consumers. As a result, consumers' actions to modify or restructure an information display are predicted to differ as a function of the task. As in the water heater example, choices can be made with information that is complex and hard to understand C provided the information is displayed in a form that promotes comparisons of brands on an attribute-by-attribute basis. Therefore, restructuring a complex display is expected to take the form of making attribute information more comparable. In contrast, in judgment tasks when complex information must be comprehended and retained, mere comparability of attribute information may not suffice. To engender feelings of confidence, consumers may have to perform additional restructuring operations on the display to implement processing strategies that maximize information use or effective storage (i.e., ready accessibility) of brand information into memory for future decisions. In this case, restructuring is expected to take a form of making brands comparable on a more abstract level, as by ranking displayed brands to reflect their relative overall performance, thus increasing processability.

Effects of Processability on Recall and Perceptions

The processability of display information may affect more than just the immediate decision (i.e., restructuring behaviors and outcomes). Information used in the decision may be stored in memory in different cognitive structures, depending upon display processability, and may be more or less accessible. In addition, consumers' perceptions of how difficult it was to use the display information, the effort they felt their decisions required, and their confidence in their decisions may vary with processability and the operations needed to improve processability.

Marketers would like consumers to remember the information they provide about their products. The strength of a memory trace, or what can be recalled, may depend upon the amount of processing and the type of processing carried out to encode and store information (Craik and Lockhart 1972). Because processing to make a choice makes comparability more important than understandability, lower levels of recall are expected for a choice task than for a judgment task. In addition to the task goal, the use of restructuring to increase processability, either by making information more comparable or by decreasing inherent difficulty, is expected to influence memory for presented information. If consumers restructure information to make a choice, memory for that restructured information is expected to be better than that for non-restructured information. Therefore, brand information presented in a form that must be restructured should be more available in memory than information that does not have to be restructured. In judgment tasks, however, we predict more brand-based evaluations and less restructuring for attribute comparability, relative to choice tasks. As a result, the difference in the amount of product information recalled between consumers who make judgments with information that is easy to compare and consumers who make judgments with information that is hard to compare may be smaller, in general, than the difference in recall between consumers who make choices with the same types of information displays.


Differences in consumers' perceptions of the difficulty of decisions and of effort and confidence in making these decisions, may be affected by the processability of available information. Because choices are often made with attribute-based strategies, decreased processability may lead to higher ratings of decision difficulty and lower ratings of confidence than when information is presented in comparable forms. Even though a decision may be difficult because the information is inherently difficult, effort perceptions are expected to increase only when consumers who make choices attempt to restructure information into a more comparable format.

When consumers make judgments, the use of overall brand evaluations is expected to decrease the impact of low processability due to lack of comparability between brands, relative to that expected for choices among brands. Consumers may feel that making judgments when information is low in comparability is more difficult, but because it does not inhibit use of a brand-based evaluation strategy, this form is not expected to result in significantly different perceptions of effort. If consumers restructure to make information comparable, confidence may be higher than for decisions made when the same information is not restructured.


Subjects and Procedure

Forty-one subjects participated in the experiment, which took, on average, twenty-one minutes to complete. The subjects were undergraduates enrolled in an introductory marketing course at a southeastern university. For their participation, subjects received extra credit in the course and a meal of pizza and soft drinks. Each subject received a booklet with the stimuli, instructions, and scales. In the task, subjects indicated their familiarity with eighteen products, and completed four choice [judgment] decisions and related scales, and a recall task.


The stimuli used to elicit subjects' choices and judgments were four brand/attribute matrices, one for each of four products. The products were water softeners, binoculars, sergers, and health insurance policies. These products were selected on the basis of pretests which indicated that similar subject populations were unfamiliar with these products. Unfamiliarity reduced the likelihood that subjects would use information other than that presented in the stimuli booklet to make their decisions. The use of extraneous information could have reduced our ability to gauge the effect of display comparability on outcome quality and other behaviors.

The stimuli were structured so that the quality of choices or judgments could be assessed with respect to display comparability. Each product matrix consisted of four brands and four attributes. One brand was constructed to dominate the other brands on all attributes. Two brands were midrange in quality. A fourth brand was inferior to the other brands on all attributes. Subjects were not asked to restructure the display information. They were told, however, that they could write in the booklets if they wished. In this respect, restructuring was essentially spontaneous.

Independent Variables

To examine the effects of display processability on decision making, we manipulated two factors: task goal and information comparability. Task goal was manipulated between subjects, so that twenty subjects chose one brand from each product matrix, while twenty-one subjects rated the quality of each brand in each product matrix.

Information comparability was manipulated within subjects. We held the amount of information constant within each matrix display and manipulated the comparability of the display information to create four matrix types for each brand. In one matrix type, all brands were comparable in display form; that is, all attribute values were in the same units. In two matrices, the units that described the attribute values for one brand differed from those used to describe the other three brands in the matrix. In one of these matrices, the dominating brand (Target 1) was different (matrix 2); in the other, one of the average brands (Target 2) was different (Matrix 3). In the fourth matrix, both of the target brands (dominant and average) had attribute units that differed from the other two brands (Matrix 4). Table 1 contains the schema.

This schema enabled us to present all four products to each subject, using each type of matrix one time. Although matrix type was linked to a particular product for each subject, the link was varied across subjects by counterbalancing both the product order and the matrix type.

Dependent Variables

The dependent variables consisted of outcome, processing, and perceptual measures. Outcome variables included decision quality and recall. In the choice task, we measured decision quality by determining whether the subject chose the dominant brand. In the judgment task, we measured decision quality by determining whether the subject gave the highest rating to the dominant brand and the lowest rating to the inferior brand. Recall was assessed by counting the amount of information recalled for the water softener and whether the recalled information was accurate. Dependent variables for processing consisted of the amount and type of restructuring. We measured the amount of restructuring by counting the number of transformations the subject had written in the stimulus booklet. Those transformations were then coded as either standardizing to increase attribute comparability or relabeling to increase brand comparability (Coupey 1994), or some other type of restructuring. For example, changing a brand attribute value to match other attribute values is standardizing (e.g., writing '1.5 hours' next to '2 cycles' or '3 sq. ft.' next to '342 sq. in.'). Writing rankings by attribute values is relabeling (e.g., '1' next to '484 gms,' '2' next to '539 gms,' and '3' next to '681 gms'). The type of restructuring was coded whether or not the information was changed correctly. The perceptual measures consisted of four seven-point Likert items which assessed the difficulty to compare information, difficulty to understand information, effort to make a decision, and confidence in the decision.



Our results indicate that subjects have different strategies for managing contextually driven decision difficulty, and that the selection of strategies for handling display information is contingent upon the task required of subjects. Subjects' behaviors can be explained as the result of effort/accuracy tradeoffs, in which subjects use different amounts of restructuring operations to achieve different goals suggested by the different tasks. We find that subjects tend to restructure more when doing judgment tasks than when making choices. Contrary to our expectations, however, subjects tend to do relatively more standardizing than relabeling for judgments. However, this finding makes sense when it is interpreted from an effort/accuracy perspective and is consistent with previous research (e.g., Montgomery 1983).

Decision quality and perceptions of the decision are influenced by the amount of restructuring that must be done to make the displays processable. Recall is also affected, although not as strongly.

Manipulation Checks and Other Nuisances

The average rating of familiarity with the eighteen products we assessed was 3.38 on a seven-point scale, where 1 indicated high familiarity and 7 indicated low familiarity. Each of the four products we used in our matrices was significantly less familiar to subjects than the average of the eighteen products (water softener: mean=5.85, p<.0001; binoculars: mean=3.72, p<.04; insurance: mean=4.11, p<.0001; sergers: mean=6.01, p<.0001). The selection of products used in the stimuli was deemed acceptable.

To avoid misinterpretation of the data due to procedural biases, we examined the effect of product and order of problem presentation on the primary dependent measures. There was no significant effect of product or problem order on accuracy (product: (2=3.06, p<.38; problem order: (2=1.72, p<.63). The order of presentation also had no significant effect on perceptions (ANOVA: difficult to compare, p<.51; easy to understand, p<.65; effort, p<.88; confidence, p<.66), recall (ANOVA, p<.62), or restructuring (ANOVA, p<.91). As expected, the product did affect perceptions of how easy the information was to understand (ANOVA, p<.003) and the effort the decision required (ANOVA, p<.04), but Duncan's multiple range tests indicated that the only significant differences were between sergers and binoculars. Sergers were seen as harder to understand and more effortful than binoculars.

In addition, there were significantly more accurate answers than inaccurate ones (125 versus 39: (2=45.10, p<.0001). This suggests that subjects did take the study seriously.

Hypotheses Tests

Outcome Variables. We examined the effect of information comparability on decision quality and recall. For decision quality, the effect of matrix types on accuracy was significant ((2=13.28, p<.0041). Subjects were less likely to be accurate when the dominant brand information was in a different form from the other brands in the matrix. The percentages of accurate responses were: all comparable matrix (EASY): 93%; dominant different (T1): 63%; average different (T2): 83%; both targets different (T1T2): 66%. The task also affected outcome quality ((2=4.89, p<.02). Subjects who made choices were more accurate (84% correct) than subjects who made judgments about the brands (69% correct), but this may be a function of the more stringent criteria for assessing judgment accuracy. [Recall that accuracy in choice was reflected by selection of the dominant brand, while in judgment it was reflected by determining the best and worst brands. Although the measure is arguably not comparable across tasks, the alternative measures would have obviated the task manipulation. For example, we could have asked choice subjects to list the best and worst brands; this essentially turns the choice task into a judgment task, thus limiting our ability to observe process differences when they exist.]

Contrary to expectations, the amount of recall was not affected by the matrix type or by the task. This non-effect may indicate a floor effect; subjects were asked to recall twenty pieces of information, but the median amount correctly recalled was three items. [The failure to observe a task effect on recall may be due to the type of judgment we requested. As noted in the discussion of process variables, the evaluations we elicited in the judgment task appear to have caused subjects to restructure the display by ranking the performance of brands on each attribute. As a result, the attribute value information receives little attention after the display is restructured.]

Process Variables. To assess the effect of task and matrix type on the amount of restructuring, we used a count of all restructuring operations done by each subject on each problem as the dependent measure. The amount of restructuring (all types) was influenced by the task (ANOVA: p<.0001). As expected, subjects did more restructuring for judgments than for choices (judgment: mean=8.82; choice: mean=3.89). Matrix type exerted no significant influence on the total amount of restructuring, but it did affect the types of restructuring.

Types of restructuring were significantly influenced by both of the independent variables. Matrix type influenced the number of standardizing operations (ANOVA: p<.0001). Subjects only standardized information when brands were not in comparable units. Matrix type did not affect the number of relabeling operations (ANOVA: p<.99). Subjects tended to relabel information by ranking attribute value information for each brand. This type of operation was equally effective for choice and judgment decisions. Subjects used both relabeling and standardizing operations more frequently for judgment tasks than for choice tasks (relabeling mean: judgment=4.4, choice=2.5; standardizing mean: judgment=3.1, choice=1.3). The number of other operations, such as summarizing brands, or adding attribute information to the display, also differed by task (ANOVA: p<.01; means: judgment=1.4, choice=.1).

We constructed a relative measure of restructuring to directly observe the differences in amounts and types of restructuring [(RELABEL - STANDARDIZE) / (RELABEL + STANDARDIZE)]. Main effects were found for task and matrix type (both significant at p<.0001). In the choice task, subjects tended to relabel more than standardize. In the judgment task, subjects tended to standardize more than relabel (index means: choice=.35; judgment=-.15). In the EASY matrix, subjects did relatively more relabeling than they did in any of the other three matrix types.

We examined the effects of restructuring on outcome quality and recall. Several effects were notable. When the target brand was comparable, (EASY, T2) less restructuring was done than when the target brand was not comparable (ANOVA, p<.02). As expected, a similar effect was observed for the number of standardizing operations (ANOVA: p<.0003) and the relative amount of standardization (ANOVA: p<.003). To examine the effect of restructuring on outcome quality, we recoded restructuring as a dichotomous variable (i.e., presence or absence of restructuring). We used a dichotomous variable to determine whether restructuring at all C regardless of the amount or type C affected outcome quality. Subjects who made evaluations and who restructured rated the dominant brand more highly than did subjects who did not restructure (ANOVA: p<.04; mean without restructuring=2.83, mean with restructuring=1.4, where lower values indicate better brand ratings). The effect of restructuring on choice quality was not significant.

We had predicted that recall for brand information would differ as a function of the type of restructuring, which itself was proposed to differ depending upon task. The relative amount of different types of restructuring (i.e., standardizing or relabeling) did have a marginal effect on recall (linear regression: p<.07). The more standardizing subjects did, compared with relabeling, the more information they recalled.

Perceptual Variables. An ANOVA to assess the effect of matrix type on: 1) perceptions of difficulty to compare display information, 2) ease of understanding display information, 3) effort to complete a decision, and 4) confidence in a decision was completed. For each scale the matrix with all brands comparable (EASY) differed significantly from the other three matrices, with all p-values less than .007. Subjects uniformly perceived the EASY matrix as less difficult, easier to understand, and less effortful. They were most confident with the EASY matrix.

Analyses of variance also revealed significant effects of the task on all four perceptual measures. Information in the choice task, compared to the judgment task, was deemed less difficult to compare (p<.001) and easier to understand (p<.02). Subjects who made choices were also more confident about their responses ( p<.0002) and felt that their decisions required less effort ( p<.001) than subjects who evaluated all brands.

We had expected that restructuring would affect subjects' perceptions of the decisions, so that completing task-appropriate operations, such as relabeling for judgment and standardizing for choice would increase confidence and decrease difficulty. Perceptions of effort were only expected to differ if subjects restructured noncomparable displays. Contrary to our expectations, confidence was influenced by restructuring so that the more relabeling subjects did, the less confident they became (linear regression, beta=-.37, p<.002). Confidence was not affected by the amount of standardizing. The amount of restructuring also had no effect on perceptions of difficulty to compare information, ease of understanding information, or effort to decide.

To examine the joint influence of task difficulty and restructuring on confidence in greater detail, we completed two linear regressions. Controlling for the effect of task difficulty (i.e., matrix type), we observed a marginally significant effect of restructuring on confidence (p<.06). With the perception of difficulty as a covariate, a regression revealed a significant effect of restructuring on confidence (p<.02).


Our research accomplished three objectives: 1) we examined the proposition that what constitutes a processable display can differ as a function of the task (e.g., choice vs. judgment), 2) we explored the types of restructuring subjects used to make displays more processable in different tasks, and 3) we assessed the effects of varying amounts of comparability and of restructuring on outcome quality, recall and perceptions. Our results indicate that decision difficulty is influenced by display comparability and by task. Subjects not only made better decisions, but felt more confident about their decisions when information was presented in a comparable form. These results were more pronounced for choice than for judgment.

Restructuring was often done to make displays more processable. We demonstrated that subjects adapt their restructuring to demands of the display, allocating more effort to restructure superior brands than inferior brands. This result is consistent with Coupey's (1994) finding of adaptive restructuring. In addition, our research indicates that the strategic use of restructuring operations changed contingent upon the task. Subjects appear to complete multiple display revisions when evaluating brands; they typically standardize all of the available information and then relabel it to reflect relative quality differences. Subjects who make choices tend to restructure only until one brand clearly dominates the others. In this type of task they tend to do more relabeling than standardizing, because relabeling is necessary to determine the best brand on each attribute. Standardization is used when the dominant brand is not in comparable units.

The relationship between restructuring and perceptions of decision difficulty and effort is provocative. We found the amount of restructuring done to a display had little effect on subjects' perceptions of difficulty and effort. Although this result was unexpected, we conclude that the lack of significant differences may indicate that being able to restructure makes difficult decisions less effortful, perhaps by simplifying evaluative aspects of the task.

The effects of display comparability on outcome quality can be explained in terms of effort/accuracy tradeoffs. In addition, differences in choice quality can be related to the difference in processing behaviors that occurs when subjects use information as-is, as suggested by the concreteness principle (Slovic 1972), and when subjects restructure the display (Coupey 1994). Because subjects attempt to maximize outcome quality and minimize decision effort, they process the displayed information opportunistically. If one brand clearly dominates most of the other brands, they choose it C even if another, noncomparable brand remains unexamined. This finding is consistent with the results of research by Slovic and MacPhillamy (1974), who observed that subjects often discard noncommensurate information from a decision.

We suggest that the desire to make good decisions limits the amount of information subjects are willing to eliminate from the decision, and that this elimination process is contingent upon the quality and amount of comparable options. When the only brand that differed was the dominant brand, subjects faced a situation in which two of the remaining brands were very close in overall quality. To make a choice, they restructured the fourth brand. In addition, when the dominant brand and one average brand were different, subjects faced a situation in which using information as is would have reduced the choice set to two brands, thus reducing the expected outcome quality below acceptable levels. To increase their options, they restructured the display.

The effort/accuracy interpretation of display processability effects also provides a plausible explanation for the differences in outcome quality between tasks. Recall that choice subjects only had to choose one brand from a display, while judgment subjects had to rate each brand. This suggests that when it is harder to determine which brand dominates (i.e., in the T1 matrix and the T1T2 matrix), choice subjects lose their advantage; they must restructure to make the target brand comparable and detectable. When a dominant brand can be readily identified (i.e., in the EASY matrix and T2 matrix), choice subjects can make choices more easily than judgment subjects can rate all brands. This interpretation is supported by the absence of significant differences in outcome quality between tasks for matrices in which the dominant brand was different (T1, T1T2), and by the significant difference between tasks when a dominant brand could be detected without altering the display (EASY, T2) ((2= 4.02, df=1, p<.04, choice more accurate).

We have focused solely upon how consumers manage decisions made difficult by external factors, such as the task and the processability of the display. Further research is needed to examine the link between relevant information in memory, such as a reference brand, and the strategies used to increase display processability. Because the extent to which consumers are willing to restructure displays may depend heavily upon the structure and accessibility of prior knowledge about a product category, research to investigate the effects of understanding presented information on processability is also needed.


Bettman, James R. (1979), A Functional Analysis of the Role of Overall Evaluations of Alternatives in Choice Processes, Advances in Consumer Research, Vol. 7, Provo, UT: Association for Consumer Research.

Biehal, G., and D. Chakravarti (1982), Information-Presentation Format and Learning Goals as Determinants of Consumers' Memory Retrieval and Choice Processes, Journal of Consumer Research, 8 (March), 431-441.

Billings, R. S., and Scherer, L. M. (1988), The effects of response mode and importance in decision making strategies: Judgment versus choice. Organizational Behavior and Human Decision Processes, 49, 258-281.

Coupey, Eloise (1994), Restructuring: Constructive Processing of Information Displays in Consumer Choice, Journal of Consumer Research, 21, 83-99.

Craik, Fergus I. M., and Lockhart, Robert S. (1972), Levels of Processing: A framework for memory research, Journal of Verbal Learning and Verbal Behavior, December, 671-684.

Ford, J. Kevin, Neal Schmitt, Susan L Schechtman, Brian Hults, and Mary L. Doherty (1989), Process Tracing Methods: Contributions, Problems, and Neglected Research Questions, Organizational Behavior and Human Decision Processes, 43, 75-117.

Kleinmuntz, Don N., and David A. Schkade (1993), Information Displays and Decision Processes, Psychological Science 4 (4) July, 221-227..

Montgomery, Henry (1983), Decision rules and search for a dominance structure: Towards a process model of decision making. In P. C. Humphreys, O. Svenson, and A. Vari (Eds.), Analyzing and aiding decision processes, 343 - 369. North Holland: Amsterdam.

Payne, John. W., James. R. Bettman, and Eric. J. Johnson (1993), The Adaptive Decision Maker. Cambridge University Press: Cambridge.

Russo, Jay Edward (1977), The Value of Unit Price Information, Journal of Marketing Research, XIV(May), 193-201.

Russo, Jay Edward, G. Krieser, and S. Miyashita (1975), An Effective Display of Unit Price Information, Journal of Marketing, 39 (April), 11-19.

Schkade, David A., and Johnson, Eric J. (1989), Cognitive processess in preference reversals. Organizational Behavior and Human Decision Processes, 44, 203-231.

Slovic, Paul (1972), From Shakespeare to Simon: Speculation - and some evidence - about man's ability to process information. Oregon Research Institute Bulletin, 12 (3).

Slovic, Paul, and Douglas MacPhillamy (1974), Dimensional Commensurability and Cue Utilization in Comparative Judgment, Organizational Behavior and Human Performance, 11(April), 172-194.