An Exploratory Investigation of Consumers’ Evaluations of External Information Sources in Prepurchase Search

Cheryl Burke Jarvis, Indiana University
ABSTRACT - Limited by opportunity, ability, and motivation, consumers who want to conduct prepurchase external information search to help make a product decision must first choose among types of information sources. An adaptive decision-making perspective using a cost-benefit framework is applied to explain the process of deciding which information sources to consult. An exploratory study using depth interviews, focus groups, and a preliminary written survey investigates the attributes or choice criteria consumers use to evaluate types of external information sources in making a source decision, the enduring images consumers possess about specific types of information sources, and how the application of these evaluative criteria differs under situational contingencies.
[ to cite ]:
Cheryl Burke Jarvis (1998) ,"An Exploratory Investigation of Consumers’ Evaluations of External Information Sources in Prepurchase Search", in NA - Advances in Consumer Research Volume 25, eds. Joseph W. Alba & J. Wesley Hutchinson, Provo, UT : Association for Consumer Research, Pages: 446-452.

Advances in Consumer Research Volume 25, 1998      Pages 446-452

AN EXPLORATORY INVESTIGATION OF CONSUMERS’ EVALUATIONS OF EXTERNAL INFORMATION SOURCES IN PREPURCHASE SEARCH

Cheryl Burke Jarvis, Indiana University

ABSTRACT -

Limited by opportunity, ability, and motivation, consumers who want to conduct prepurchase external information search to help make a product decision must first choose among types of information sources. An adaptive decision-making perspective using a cost-benefit framework is applied to explain the process of deciding which information sources to consult. An exploratory study using depth interviews, focus groups, and a preliminary written survey investigates the attributes or choice criteria consumers use to evaluate types of external information sources in making a source decision, the enduring images consumers possess about specific types of information sources, and how the application of these evaluative criteria differs under situational contingencies.

Technological advances have made vast quantities of information instantly available to consumers at the touch of a button. The media touts that we are in "the information age." However, despite a substantial body of research on consumer behavior with respect to information, do we yet really know what consumers want to know and how they evaluate the information they receive? More specifically, as marketers flood consumers with more information in greater detail and newly available forms, it becomes clear marketers don’t understand why and how consumers choose to consult some information sources and avoid others in the process of making a product purchase decision.

It is this conceptualization of consumer choice regarding information that is lacking in existing consumer behavior researchCthe understanding of consumers’ choice of information sources as a decision-making process similar in form to any product decision process. Marketing researchers have studied consumers’ use of information in prepurchase decision-making, but much of this work has focused on influences on the amount of external information search or differences in search behavior across individuals (e.g., Beatty and Smith 1987; Brucks 1985; Furse, Punj and Stewart 1984; Ozanne, Brucks and Grewal 1992; Urbany, Dickson and Wilkie 1989). Others have investigated consumers’ attention to, comprehension of, and integration of information (e.g., Celsi and Olson 1988; Cohen and Basu 1987; Lim, Olshavsky and Kim 1988; MacInnis and Jaworski 1989; Mick 1992).

However, little research has addressed the processes by which consumers formulate desires for new information from external sources and make choices among alternative information sources. Olshavsky and Wymer (1995) proposed a model of desire for information from external sources, arguing that since information can be viewed as a special type of good, consumer behavior with respect to information can be described in the same terms as behavior with respect to goods.

Conceptualizing external information search behavior as a decision-making process can provide insight into why and how consumers select among prepurchase information sources. Consumers limited by time, ability, and motivation to search for information limit their consultation of sources, making choices among an ever-expanding number of sources. Considerable evidence exists that external information search is extremely limited, even for the most expensive products (Beales et al 1981; Jacoby et al 1975; Capon and Burke 1980). Since few sources are consulted for any particular purchase, it would be valuable to know which evaluative criteria consumers use to distinguish and select among types of information sources. This knowledge could help us understand the enduring images consumers hold about information source types, how these images affect consideration sets of information sources in search, and how contingencies alter the application of evaluative criteria to affect consumers’ judgments of a source’s appropriateness for a decision.

The purpose of this article is threefold: First, a cost-benefit framework of adaptive decision-making is applied to the context of consumer choice of information sources. Second, an exploratory study using depth interviews, focus groups, and a pretest of a written survey attempts to identify the evaluative criteria and enduring images of information source types that consumers apply in prepurchase search. While a few previous studies have included issues of consumer evaluations of information sources in different contexts, this article is the first to place them in a cohesive theory for comprehensive investigation. Finally, the study will begin to investigate situational contingencies’ effects on the application of evaluative criteria to information source decisions.

INFORMATION SOURCE CHOICE AS AN ADAPTIVE DECISION PROCESS

A consumer in the midst of a product choice task may or may not seek out external sources of information to help make that decision. Robust evidence exists to demonstrate that external prepurchase information search is very limited even for expensive products, and practically non-existent for most product decisions (Capon and Burke 1980, Jacoby et al 1975). The suggestion exists that consumers are bad decision-makers if they don’t take advantage of all information available to them. However, an adaptive decision-making framework suggests that consumers weigh the costs and benefits of external search in order to maximize utility in situational contingencies. Rather than focus solely on the amount of informaton search undertaken, as a majority of past studies have done, a more productive research approach would be to delineate the types of sources that are consulted under certain contingencies and understand the evaluative criteria applied to choose among sources.

A distinction should be made between the criteria consumers use to evaluate information versus sources of information. For example, if one judges a salesperson to be "expert" and "trustworthy," then one infers that the information that salesperson gives is "accurate"Cbut the judgment about the information is an inference, not a direct judgment.

This article structures the argument that: (1) when a prepurchase product decision problem generates a desire for external information, consumers evaluate sources based on some critical criteria; (2) consumers hold enduring beliefs about information sources, learned through experience or socialization; (3) situational contingencies affect not only desire for information, but also how evaluative criteria are applied and what information source choice strategy is employed; and (4) consumers maximize utility to adaptively respond to an information source decision problem to cope with environmental and processing limitations.

The Desire for External Information

First must come the desire to seek external information. Olshavsky and Wymer (1995) proposed that consumers must first recognize an information "problem" which involves a comparison of the "desired state" and "perceived actual state" in regard to information. The "desired state" refers to the specific information a consumer must have in order to execute an evoked product preference formation strategy. The "perceived actual state" is the consumer’s perception of information stored in long-term memory and external memory aids (i.e. clipped ads, saved media reports, notes, etc).

When desire for external information is generated, the model proposes, consumers evaluate types of external information sources by comparing the "desired state" with expectations of each type’s ability to provide the needed information. Considered one source at a time, if the desired state exceeds expectations then the consumer may reject that type of source in whole or in part. If no external sources are acceptable, then the desire for information will continue but the choice process may be postponed. If the desired state is less than or equal to the expectations for a source, then desire for information from a type of source or a combination of types of sources will be formed.

Consumer Evaluation of Information Source Types

A few studies have touched upon consumers’ enduring evaluations of source types, but usually only as secondary findings in studies with the traditional objective of investigating antecedents or conditions of search. For example, Beatty and Smith (1987) found that consumers with little product knowledge are more likely to rely on friends and associates for information because they are regarded as more credible than salespeople. They also noted that consumers under time pressure are less likely to use neutral sources because they are regarded as more time-consuming than commercial sources. Robertson (1971) found marketer-controlled sources are more important in early stages of decision-making for information on product alternatives, while opinions of associates become more important closer to the final decision because they are regarded as more trustworthy.

Consumer evaluations of the credibility of commercial information sources have been more directly investigated (Deshpande and Stayman 1994; Friestad and Wright 1994). Calfee and Ringold (1994) reviewed 60 years of national polls and trade data that consistently indicated about 70 percent of consumers think advertising is often untruthful and seeks to persuade people to buy things they do not want, but nontheless provides valuable information.

Thus, some evidence exists that consumers hold enduring beliefs learned through experience or socialization about information source types on many evaluative criteria, and these beliefs affect how consumers choose and use information source types. Decision theory also suggests that situational contingencies would affect the application of these enduring evaluations in an information choice problem, either by reprioritizing the evaluative criteria to fit the decision characteristics or by generating an alternative information source choice strategy.

Much of the research in external information search has focused on developing typologies of information search strategies using nearly 60 variables influencing the amount of information search (e.g. Beatty and Smith 1987; Furse, Punj and Stewart 1984; Punj and Staelin 1983; Schmidt and Spreng 1996). These typologies often include aspects of the environment (e.g. difficulty of the choice task, number of alternatives, complexity of the alternatives), situational variables (e.g. previous satisfaction, time constraints, perceived risk), and consumer characteristics (e.g. education, prior knowledge, involvement). In the adaptive decision-making framework presented here, these antecedents of search serve as task or context contingencies affecting consumers’ evaluations of the appropriateness of source types for specific prepurchase situations.

Payne, Bettman and Johnson (1993) identified task and context variables of decision problems that affect the decision strategy used. Among the task characteristics they specified: (1) task complexity, including number of alternatives, number of attributes, and time pressure; (2) information display characteristics such as concreteness, availability, completeness, and format; and (3) agenda effects, or effects of placing constraints on the order in which elements of a choice set are considered.

It can be hypothesized that task complexity variables affect information source type decisions in the same way as other decision problems. Task complexity in a information source decision can be influenced by how many different types of sources are available for a particular piece of information. When faced with more complex multialternative decision tasks (multiple sources of information), subjects have been found to prefer noncompensatory strategies such as elimination-by-aspects (Tversky 1972) and the conjunctive model (Einhorn 1970). Noncompensatory strategies also are used to simplify task complexity when the number of attributes (evaluative criteria) is increased, for example, when an information source must be judged based equally on accuracy, timeliness, completeness and ease of use. Time pressure might make criteria such as immediate availability or ease-of-use of a source most salient, resulting in the use of a lexicographic heuristic in which the source judged highest on availability is consultedCa salesperson, for example.

The group of task effects gathered under Payne, Bettman and Johnson’s (1993) "information display" category maps directly on to the set of evaluative criteria suggested by previous research on information search. The difference in approach proposed here, however, is in conceptualizing the decision problem as that of using different characteristics to choose which information sources to consult, rather than that of having the characteristics of the information affect the choice of product.

Finally, agenda effects are also relevant in information source decision problems. Because not all sources are equally available at any one time, there are natural constraints on the order in which the sources in the consideration set can be consulted. This could result in a change in the criteria considered in the elimination process, or a change in evaluation strategy (Tversky and Sattath 1979).

In the same way, Payne, Bettman and Johnson’s (1993) "context effects" transfer directly to the source choice task. They identified four context effects: (1) similarity of alternatives, (2) quality of the option set, (3) reference pont effects, and (4) framing effects. Research is needed to determine how consumers perceive the similarity and quality of information source typesCfor example, if commercial sources are considered less credible than noncommercial sources as past research suggests, are all commercial sources considered similar on this dimension? Or does a continuum exist upon which some sources are perceived as more credible than others? Framing of decision problems (Tversky and Kahneman 1981) can change cognitive costs and the desire for accuracy in a decision process, altering the choice strategy and subsequently the information source used.

The Cost-Benefit Framework

Consumer choice of decision strategies has been discussed most frequently in contingent approaches in terms of cost-benefit (or effort-accuracy) frameworks (e.g. Payne, Bettman, Johnson 1993; Smith, Mitchell and Beach 1982). The Payne, Bettman Johnson (1993) effort-accuracy framework is based on five major assumptions: (1) people have a repertoire of strategies and heuristics for solving decision problems; (2) available strategies have differing advantages (benefits) and disadvantages (costs) with respect to an individual’s goals and the constraints of the problem; (3) different tasks have properties that affect the relative advantages and disadvantages of the various strategies; (4) the strategy considered "best" for the task is selected; and (5) the process is top-down, that is, a priori perceptions of the task and strategies determine subsequent behavior. This model seems most appropriate for the context of information source choice discussed here. Among the a priori perceptions of the task discussed in the fifth step are the enduring perceptions of source type characteristics held by consumers.

While the model focuses on effort and accuracy of the decision as the primary cost and benefit, Payne, Bettman and Johnson (1993) point out that other benefits and costs can influence decision strategy selection and subsequently source selection. For example, the need to justify a decision may factor into how a person decides to decide. Tversky (1972) suggests that an advantage of an elimination-by-aspects process may be that it is easy to explain and defend. Therefore, if we know which criteria consumers consider most important for an information source choice in which an EBA strategy is executed because the choice must be justified to others, we should be able to predict which types of information sources are most likely to be consulted.

In a much more basic approach, the choice of information source type itself (rather than the choice of decision strategy for selecting sources) also can be modeled in a cost-benefit framework. Many of the evaluative criteria consumers are likely to apply to external information sources can be placed on a cost-benefit continuum. For example, types of sources that are difficult to use (because of organization, format, availability, etc.) are cognitively costly, whereas sources that are easy to use on those same dimensions may be judged to have an advantage over other sources. Perceived credibility, accuracy and lack of bias in a source type might be considered benefits because they are perceived to lead to better product choices, while sources with the opposite characteristics would be costly to rely upon because of the increased potential to make a bad product choice.

EXPLORATORY STUDY

Objectives and Method

Because information on the criteria consumers use to evaluate information sources to date is sketchy, disjointed and often anecdotal, exploratory research was considered the appropriate first step. The exploratory research described here consisted of three stages. First, three individual depth inteviews were conducted with adult professionals or doctoral students. From these interviews, a focus group moderator’s guide was developed, and three focus groups were conducted using 28 undergraduate business students in a major Midwestern university. Finally, a preliminary written survey was pretested with the focus group participants following the focus groups to attempt to capture more detail than subjects would be able to articulate from top-of-mind discussions.

The goals of this three-staged study were (1) to attempt to identify the criteria consumers use in evaluating types of information sources; (2) to investigate the enduring beliefs consumers possess about information source types, and (3) to begin to explore if and how situational contingency task effects change the application of evaluative criteria in making source choices.

Results and Discussion

Interviews and Focus Groups: Subjects for the individual interviews and focus groups were asked to recall a recent purchase for which they sought external information and to keep that instance in mind throughout the interview. They were asked to recount the external information sources they had consulted in the product decision, why sources were chosen and for specific evaluations of source types, what types of information were gleaned from each source, and reasons for choosing not to use other sources. Subjects then were presented with alternative scenarios based on variables shown to influence amount of search and decision strategies, such as time pressure or prior product knowledge, and were asked to project how their source choices would change in the scenarios described.

While comparisons between the depth interview and focus group subjects must be considered anecdotal at best, because of sample size and methodology differences, interesting disparities were seen that could be hypothesized to result from observed differences in consumption expertise and experience. Additional research is necessary to determine if these differences are consistent across age/experience groups. The adult subjects in the individual interviews related much more in-depth information searches than did the undergraduate student subjects in the focus groups. All three interview subjects were able to recall clearly their path of information source choices and were easily able to articulate their evaluative criteria for choosing certain sources over others. In comparison, the majority of the undergraduate focus group subjects revealed themselves to be relatively naive consumers with limited independent consumption and source evaluation experience. For example, only one out of 28 subjects had ever independently purchased an automobile, and only two others reported observing family members purchase an automobile. Only one had purchased a home computer, surprising in a group of junior college students. But even more telling was their relative lack of information search and lack of experience with several information sources.

The product purchase decision scenarios subjects generated for discussion extended across several categories, but all involved relatively expensive or high-risk products. Examples include a computer, a car, an infant car seat (for the individual interview subjects); musical instruments, a purebred dog, car repair, car tires, electronics, and a vacation trip (for the focus group subjects).

All the interview subjects relied on impersonal, non-marketer-controlled sources as their first information source selection (a car buying guide, a PC publication’s editorial ratings of computer, and a Consumer Reports review of infant car seats) to determine desired features, determine brands and models to enter into the consideration set, and gain perspectives on pricing. Family and friends’ recommendations also were used by all three for the same purposes. Company brochures and catalogs were used by the car and computer shoppers to help specify features, options and price ranges. The computer shopper used the World Wide Web to locate computer manufacturers’ hoe pages for product features and prices. As consideration sets narrowed, print ads were consulted to locate best prices and choose retailers. None reported using salespeople for information other than obtaining a final purchase price. None reported using information from television or radio ads or packaging.

The less-experienced focus group subjects reported consulting a more limited set of sources and in a different order. Twenty-five out of 28 subjects reported that their first step in prepurchase information search was gathering recommendations from friends and family. The remaining three reported starting out at a retail outlet, either viewing displays or speaking to salespeople. Of those who talked to friends/family first, most then proceeded to a retail outlet and made a final purchase decision based on price and availability of models, with some consultation with a salesperson. No other information sources were consulted in the vast majority of the decision problems reported.

Only one of the 28 focus group subjects reported having used Consumer Reports, and all stated they had never used any other independent product rating report (such as PC Magazine’s "Top 10 Buys in Home Computers" or Edmund’s car buying guides). Most claimed they had never seen such reports, and several said they were unaware such independent sources existed.

The focus group subjects’ lack of experience with the World Wide Web also was surprising, considering that as university students they have unlimited free access to the Web and are heavy users of electronic mail. Only two of 28 subjects reported having used the Web for prepurchase information search. Of the remaining subjects, only three said they occasionally "surf" the Web for entertainment purposes, and the rest reported either having never accessed the Web or doing so only when required for a college course. This majority reported unfamiliarity and discomfort with using Web browsers.

None of the focus group members self-generated any form of traditional mass media advertising as an information source. Mass media advertising simply wasn’t in their consideration set as a viable external information source in their product prepurchase situations.

Evaluative criteria generated by both interview and focus group subjects were in line with expectations. Commercial sources in general and salespeople in particular were considered "biased" and "untrustworthy." When subjects reported using neutral third-party sources, it was because they were evaluated as "unbiased," "factual," "credible," "well-researched/documented," and "complete" in that they allowed subjects to compare across brands and models. In contrast, friends and family members were considered valuable information sources because of their "convenience," "trustworthiness," "reliability," and knowledge of subjects’ preferences and situations.

Print advertising, when used, was evaluated as "easily accessible," "cheap," "permanent," (could be referenced whenever needed), and the "only/easiest way to compare price across several retailers at once." Broadcast advertising was called "overhyped," "biased" and "flashy," and subjects didn’t believe it provided any substantive information for a product choice. Broadcast advertising was also considered "inaccessible" and "unusable" despite its permeating presenceCa specific ad couldn’t be recalled at the necessary moment in the product decision process.

Subjects self-generated several situations where their information source choices would differ based on the specifics of the situation. They suggested that certain retail situations provide more expert and credible salespeopleCa computer specialty store salesperson would be considered higher in expertise than a department store salesperson responsible for selling not only computers but a variety of other electronics, for example. Others volunteered that information source choices would change depending on the amount of prior prouct knowledge, while another suggested that the product category made a difference. These contingencies parallel those suggested in past research.

Survey Pretest: Survey questions were based upon past literature and the individual interviews and were pretested with graduate students before the larger pretest with the focus group participants. The first seven questions were sets of seven-point semantic differential scales, designed to determine if any significant differences could be observed between the mean evaluations of information source types, and to pretest scale items that might later be tested in factor analysis to assess whether unidimensional constructs can be isolated from the vast array of evaluative terms that have been suggested in the literature. The information sources and evaluative criteria tested are shown in Table 1. Paired t-tests were used to test for significant differences in the means.

Sources differed significantly on respondents’ evaluations of accuracy. "Consumer groups such as Consumer Reports" were judged significantly more accurate than any other source choice except for "friends and family." The means for salespeople and television advertising were both on the negative side of the accuracy scale, while paid print advertising, packaging and Internet Web sites were almost in the exact middle of the scale. It could be that accuracy is not considered important in evaluations of some or all of these last three sources, or it is also possible that the respondents felt they could not appropriately judge these sources on this criteria because of a lack of familiarity.

The level of bias perceived in various information source types was quite distinct, with tight confidence intervals around each mean. Consumer Reports was judged significantly less biased than any other source. Salespeople and TV ads were significantly more biased than all other sources.

The "trustworthy" criteria showed an interestingly large confidence interval around the mean of print advertising, while other sources all exhibited tight confidence intervals. Meanwhile, Consumer Reports and friends/family were both judged significantly more trustworthy than any other source, while packaging and the Web comprised the next cluster on the scale, followed by salespeople and television on the "not trustworthy at all" end of the scale.

Ratings of credibility show big gaps between means and tight confidence intervals. Traditional commercial sourcesCtelevision, print ads, and salespeopleCwere rated lowest on credibility, significantly lower than Consumer Reports and friends/family. One interesting finding was that the Web was considered significantly more credible than salespeople and television advertising. This could be seen as a signal that subjects don’t consider the information gleaned from Web sites as "commercial" material, and therefore they may not even consciously consider or evaluate the original source of Web site information. Another possible explanation, however, based on the scale-centered mean for this source, is that these particular non-Web-using subjects defaulted to the center of the scale as an implicit "don’t know" response due to lack of familiarity with the source.

Consumer Reports and friends/family, as well as the Web in this instance, showed significantly higher ratings on how "informative" they are perceived to be than salespeople, television advertising, or print advertising. Packaging was not significantly more informative than the other commercial sources, but was significantly worse than both Consumer Reports and the Web.

TABLE 1

INFORMATION SOURCES AND EVALUATIVE CRITERIA

Friends/family and Consumer Reports were considered the most "useful for decision-making," significantly more so than salespeople, television or print advertising, and packaging. The Consumer Reports results may be positively biased by the focus group discussions that preceeded the survey.

Traditional commercial sources were assessed as the least complete sources; while Consumer Reports was signifiantly more complete than print or television advertising or salespeople. Packaging, friends/family, and the Web showed no significant differences and were centered at the middle of the scale. Again, the neutral centering of the mean for the Web may reflect an inability of this unfamiliar sample to judge the Web on this criteria.

Friend and family recommendations were considered the easiest sources to obtain, significantly more so than information from salespeople, Consumer Reports, print advertising, packaging, and the Web. The fact that Consumer Reports and the Web were rated as more difficult to obtain may well be a bias of this particular set of subjects, as discussed earlier. Television advertising was judged the least time-consuming to obtain, and in post-survey discussion subjects explained that they answered based on the fact that television as a medium was not time-consuming to obtain because they all had easy access to the technology and received information from it passively and without conscious effort. Consumer Reports and the Web were considered the most time-consuming, again probably partly due to this particular group of subjects’ lack of familiarity and comfort with these sources. Packaging was also judged highly time-consuming, surprising after the focus group discussion about how most of the subjects’ information-gathering came at the point of purchase. Friends/family and print advertising hovered somewhere near the middle of the scale.

Friends/family were rated the easiest sources to understand, while Consumer Reports was the least easyCagain, probably the result of this sample group’s complete unfamiliarity with Consumer Reports or similar ratings publications. There were no significant differences among the other sources, all of which fell at the middle of the scale. Friends/family were also judged the most able to "be organized to fit my needs" and the most "available when needed," significantly more so than any other source. Salespeople were the next most available, significantly more so than Consumer Reports and television advertising, but not significantly different from the other sources. Consumer Reports earned a significantly better rating than salespeople and television on the basis of the usability of the format of the information, but showed no significant differences between it and any other sources.

The evaluations of how "entertaining" a source is, a potential aesthetics criteria suggested by Olshavsky and Wymer (1995), showed significant difference in means and tight confidence intervals. Television was significantly more entertaining than salespeople, Consumer Reports, print ads, and packaging. Friends/family and the Web were less entertaining than television, but not significantly so.

The criteria of "timeliness" did not generate any significant perceived differences between sources. The means hover around the center of the scale, and the confidence intervals are quite large. This evidence and post-session debriefing interviews with subjects indicate that this item’s wording may have been confusing or open to multiple interpretations by respondents. It needs further testing.

The second set of survey questions was ranking tasks included as a preliminary investigation of the potential for situational contingencies in the product decision task to alter the priority of evaluative criteria, which is one of the hypothesized reasons for why consumers choose different information sources in different types of decision problems. For example, respondents were asked to imagine themselves buying a product needed urgently without much time to make a decision, and then rank the most important criteria for an information source they would consult in this circumstance from a list provided. Frequencies of rankings were investigated.

As would be expected, under conditions of riskCexpensive products, decisions that must be justified to an important third party, low/no prior knowledge of the product categoryCsubjects ranked the criteria relating to the accuracy, trustworthiess, bias and expertise of the source as the most critical and did not consider issues regarding ease of use. Under conditions of time pressure, the ease of obtaining the source, its ease of use, understandability, and timeliness earned more emphasis, even while accuracy and trustworthiness remained important. Without time pressure, the ease-of-use factors practically disappeared from the rankings, returning the emphasis to the accuracy/trust criteria, but the criteria of completeness also entered the picture when time pressure was removed from the scenario. When the product under consideration was an inexpensive, one-use throw-away product, the accuracy/bias/trust criteria almost disappeared, completely in favor of the easy to obtain/use/understand. Most interestingly, when the subjects were given a scenario in which they were very knowledgeable about the product category, their rankings were almost evenly distributed across the various choice criteria, except for a slight emphasis on the "easy to obtain" dimension.

Certainly, these results would seem to provide preliminary evidence that situational differences in the product decision problems could influence the prioritization of evaluative criteria. However, it must be recognized that the methodology of using self-reported measures in hypothetical scenarios may create self-fulfilling prophecies through response biasC"good" subjects applying their own implicit theories in attempting to give the answers to the questions that they believe should be given. In addition, self-reports of the lack of any use of advertising as an information source must be considered suspect in the light of extant research’s strong evidence that advertising is used for brand awareness and global attitude development at the very least.

Limitations: Several additional limitations of this study should be notedCthe most obvious being its exploratory nature and small sample size. While the article provides a general conceptual framework and some preliminary guidance for developing future studies, empirical research with established a priori hypotheses and a controllable and externally valid methodology will be needed before any conclusions or recommendations can be drawn from this topic.

Conceptually, this article and the study were both limited to situations in which the desire for external information exists, in the context of prepurchase information search. Scenarios presented for discussion in interviews and focus groups asked subjects to consider recent actual purchases in which they needed to gather external information, as well as hypothetical purchase situations developed to incorporate antecedents of external search. Other information acquisition contextsCsuch as on-going search related to enduring product involvement, passive information acquisition, information acquisition to guide product use, curiosityCprovide substantial areas for additional research.

Also, the administration of the survey pretest to the focus group participants after the focus group discussion could have biased their opinions on the survey. Testing of the questionnaire independent of group discussion on the issues is necessary for any firm conclusions to be drawn.

Finally, as discussed in the focus group results, the use of undergraduate students as subjects limited the results because of their relative lack of independent product evaluation and purchase experience. While this biased sample provided an initial starting point for this study, future research in this area must incorporate older subjects with more varied independent consumption experience.

CONCLUSION

This exploratory study provides evidence that consumers do hold enduring evaluations of external information sources based a number of criteria, and these evaluations affect choices of information source types in prepurchase product decisions. This is the irst comprehensive effort to investigate and identify the criteria consumers use to evaluate external sources of information in making choices between sources. The framework from the decision-making literature helps explain how consumers contingently decide between information source types based on the preferences and priorities for external information they have formed from their evaluations of sources.

This approach expands the existing information search literature beyond its previous investigations solely into the influences on amount of search, into a discussion of what types of information sources consumers prefer for prepurchase decision making and why. A continuation of this line of study could contribute significantly to our understanding of consumers’ choices of information, leading to improvements both in marketing managers’ communication campaigns and in public policy efforts to provide consumers with the information they desire.

REFERENCES

Beales, Howard, Michael Mazis, Steven Salop and Richard Staelin (1981), "Consumer Search and Public Policy," Journal of Consumer Research, 8 (June), 11-22.

Beatty, Sharon E. and Scott M. Smith (1987), "External Search Effort: An Investigation Across Several Product Categories," Journal of Consumer Research, 14 (June), 83-95.

Brucks, Merrie (1985), "The Effects of Product Class Knowledge on Information Search Behavior," Journal of Consumer Research, 12 (June), 1-16.

Calfee, John E. and Debra Jones Ringold (1994), "The 70% Majority: Enduring Consumer Beliefs About Advertising," Journal of Public Policy & Marketing, 13 (Fall), 228-238.

Capon, Noel and Marion Burke (1980), "Individual, Product Class and Task-Related Factors in Consumer Information Processing," Journal of Consumer Research, 7 (December), 314-326.

Celsi, Richard L. and Jerry Olson (1988), "The Role of Involvement in Attention and Comprehension Processes," Journal of Consumer Research, 15 (September), 210-224.

Cohen, Joel B. and Kunal Basu (1987), "Alternative Models of Categorization: Toward a Contingent Processing Framework," Journal of Consumer Research, 13 (March), 455-472.

Deshpande, Rohit and Douglas Stayman (1994), "A Tale of Two Cities: Distinctiveness Theory and Advertising Effectiveness," Journal of Marketing Research, 31 (February), 57-64.

Einhorn, H.J. (1970), "The Use of Nonlinear, Noncompensatory Models in Decision Making," Psychological Bulletin, 73, 211-230.

Friestad, Marian and Peter Wright (1994), "The Persuasion Knowledge Model: How People Cope with Persuasion Attempts," Journal of Consumer Research, 21 (June), 1-31.

Furse, D.H., G.N. Punj and D.W. Stewart (1984), "A Typology of Individual Search Strategies Among Purchasers of New Automobiles," Journal of Consumer Research, 10 (March), 417- 431.

Jacoby, J., R. Chestnut, K. Weigl, W. Fisher (1975), "Prepurchase Information Acquisition," in Advances in Consumer Research, Vol. 3, B.B. Anderson, ed. Atlanta: Association for Consumer Research, 306-314.

Lim, Jeen-Su, Richard W. Olshavky and John Kim (1988), "The Impact of Inferences on Product Evaluations: Replication and Extension," Journal of Marketing Research, 25 (August), 308- 316.

MacInnis, Deborah J. and Bernard J. Jaworski (1989), "Information Processing from Advertisements: Toward an Integrated Framework," Journal of Marketing, 53 (October), 1-23.

Mick, D.G. (1992), !Levels of Subjective Comprehension in Advertising Processing and Their Relations to Ad Perceptions, Attitudes and Memory," Journal of Consumer Research, 18 (March), 411-424.

Olshavsky, Richard W. (1985), "Towards a More Comprehensive Theory of Choice," in Advances in Consumer Research, Elizabeth C. Hirschman and Morris Holbrook, eds., Association for Consumer Research, 12, 465-470.

Olshavsky, Richard W. and Walter Wymer (1995), "The Desire for New Information from External Sources," in Proceedings of the Society for Consumer Psychology, American Psychological Association, Division 23, Scott B. MacKenzie and Douglas M. Stayman, eds., Printmasters: Bloomington, IN, 17-27.

Ozanne, J.L., M. Brucks and D. Grewal (1992), "A Study of Information Search Behavior During the Categorization of New Products," Journal of Consumer Research, 18 (March), 452-463.

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

Punj, Girish and Richard Staelin (1983), "A Model of Consumer Information Search Behavior for New Automobiles," Journal of Consumer Research, 9 (March), 366-380.

Robertson, Thomas S. (1971), Innovative Behavior and Communication, New York: Holt, Rinehart and Winston.

Schmidt, Jeffrey B. and Richard A. Spreng (1996), "A Proposed Model of External Consumer Information Search," Journal of the Academy of Marketing Science, 24 (3), 246-256.

Smith, J. F., T.R. Mitchell and L.R. Beach (1982), "A Cost-Benefit Mechanism for Selecting Problem Solving Strategies: Some Extensions and Empirical Tests," Organizational Behavior and Human Performance, 29, 370-396.

Tversky, A. (1972), "Elimination By Aspects: A Theory of Choice," Psychological Review, 84, 327- 352.

Tversky, A. and D. Kahneman (1981), "The Framing of Decisions and the Psychology of Choice," Science, 211, 453-428.

Tversky, A. and S. Sattath (1979), "Preference Trees," Psychological Review, 86, 542-573.

Urbany, Joel E., Peter R. Dickson and William L. Wilkie (1989), "Buyer Uncertainty and Information Search," Journal of Consumer Research, 16 (September), 208-215.

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