Research on Consumer Information Acquisition

John L. Swasy, The Pennsylvania State University
ABSTRACT - The three papers in this session focus upon a wide range of issues related to consumers' information search. They illustrate several important conceptual and methodological aspects which are important for future research to consider.
[ to cite ]:
John L. Swasy (1984) ,"Research on Consumer Information Acquisition", in NA - Advances in Consumer Research Volume 11, eds. Thomas C. Kinnear, Provo, UT : Association for Consumer Research, Pages: 256-258.

Advances in Consumer Research Volume 11, 1984      Pages 256-258

RESEARCH ON CONSUMER INFORMATION ACQUISITION

John L. Swasy, The Pennsylvania State University

ABSTRACT -

The three papers in this session focus upon a wide range of issues related to consumers' information search. They illustrate several important conceptual and methodological aspects which are important for future research to consider.

INTRODUCTION

A variety of issues are presented in the three studies in this session. The papers illustrate survey and experimental methodologies. They employ self report and behavior data. Also, they investigate a range of variables -- cognitive personality traits, sex, education, etc.--which are believed to influence consumers search behavior. The following comments are offered as suggestions for future research.

Swartz and Stephens

The Swartz and Stephens study proposes to investigate the information search for services by individuals age 55 and older. They point out that despite the importance of this consumer segment, very little research has addressed their distinct information needs and search behavior. Also, as Swartz and Stephens note, few studies have distinguished the differences in consumers search behavior for services and tangible products. This background framed their research goals:

(1) to examine whether the amount and type of information search undertaken by an older consumer will be the same for services as for tangible products;

(2) to determine the types of sources employed by older individuals when selecting services; and

(3) to identify the determinants of how many different source types are consulted.

Future research may improve and build upon several aspects of the Swartz and Stephens investigation. One issue is the development of a theoretical perspective of consumers' information acquisition processes and behavior that would support the search for expected differences between "tangible products" and "services." The rationale for expecting differences or similarities might be developed in more detail. Future research might explore various concepts raised by Shostack (1977), in her description of services marketing. These include: an experiential definition of a service's reality, the determination of consensus realities, the marketing practice of enhancing products via associations with abstract concepts versus the marketing practice of enhancing services by association with tangible, peripheral cues, and the importance of consumers' need to "experience" the service's environment. A closer examination of how consumers deduce the reality of a service, what peripheral cues are important and how consumers' assess cue credibility appear to be important determinants of consumers information search behavior for services. A consideration of these issues might also lead to the articulation of new dependent variables for describing the nature of consumer search. Indicators such as number of sources and number of source types may not adequately describe the processes which may underlie these general outcomes.

Beyond this first step it also appears likely that great differences in search behavior may exist among various types of services and products. Portions of Swartz and Stephens' results support this intuitive notion. The barber/beauty shop and financial institution decision data are quite different. Because of this, care must be exercised when formulating conclusions about consumers search behavior for services in general. In the Swartz and Stephens study three services were examined while in a related study by Klippel and Sweeney (1974), two products were used. Future research might examine multiple services representative of various categories of services. Aggregating over multiple observations within meaningful categories of services might improve the validity of future research findings.

In line with the need for more theoretical development pertaining to differences between products and services, is the need for theory development concerning differences between the search behavior of older and younger consumer segments. A recent review by Phillips and Sternthal (1977) might be helpful in directing extensions to the research by Swartz and Stephens (1981) and Klippel and Sweeney (1974).

It should be noted that the sample used by Swartz and Stephens was recent newcomers to a southwestern city. Although demographic information was not reported it seems reasonable to suggest that the abilities and information needs of these more mobile elderly may differ quite considerably from the less mobile elderly. These differences may be of importance when considering the generalizeability of these results and their implications for public policy.

Several methodological issues are important for future research to consider. In the Swartz and Stephens study subjects rated the importance of each information source they used. From these responses mean importance ratings were compared using multiple mean difference tests and conclusions about the relative importance of various sources are presented. Limitations on the interpretation of these differences (or lack of differences) should be noted. Because all subjects did not use many of the sources, these unweighted means may overstate the importance of various sources. For example, in Table 2 Swartz and Stephens report that the mean importance rating for "print ads" was 6.5 and not different from the mean of 6.2 for "personal sources" in the doctor selection search. Because only 9 respondents used print ads as an information source and 53 used personal sources, this comparison may be inappropriate. Future research might consider assessing relative importance using only subjects who report using the same set of information sources.

A second methodological issue of concern is the use of post-process self report data as a basis for investigating the determinants of consumer search. Specifically asking subjects to recall the evoke set size which they had at the beginning of the search process appears to be problematic. In the Swartz and Stephens study subjects were asked to recall evoke set size for decisions which may have occurred as long as two years prior to the study. The observation that this measure was the most important predictor of the amount of search may stem from respondents' consistency in reconstructing past events rather than reflecting a true causal relationship.

Schaninger and Buss

This study proposes to investigate the influence of task structure variables, general individual-differences, and the interaction between general and product-specific individual differences on the extent of consumers information search. Future research might improve upon several aspects of this study.

The primary shortcoming of this study is that it is not clearly positioned with respect to an earlier reported study by Schaninger and Sciglimpaglia (1981). These two studies employed the same data base and used approximately the same set of variables (socio-demographic variables were also included in the earlier study). In the initial study a canonical correlation analysis was used to examine the relationships between a set of demographic and individual difference predictor variables, and a set of criterion variables reflecting depth of search. In the Schaninger and Buss study the same individual difference variables and dependent variables are used. The distinction appears to be that instead of investigating a linear composite of main effects due to various sources the Schaninger and Buss study is examining the effects of the two-way and three-way interactions among these various predictor variables. It is not clear how the partially significant relationships observed by Schaninger and Sciglimpaglia (1981) between the abstract concepts representing the predictor and criterion sets of variables warrant further investigation of univariate relationships. A quote from Franklin Evans describes this concern:

"The degree of correlation that can be found in a set of data is limited only by the ingenuity and persistence of the statistician" (Evans and Roberts 1963).

The second major shortcoming of this study is that, regardless of the earlier study's findings, the rationale for investigating these higher order interactions is not clear. As partial support, the authors cite the work of Brody and Cunningham (1968) and Blake et al. (1973). A close examination of these studies, however, indicates that their findings are somewhat suspect. In fact, in the Blake et al. (1973) study the authors report their inability to replicate their findings in a second sample using a different set of stimulus Products.

Instead of simultaneously assessing a multitude of various cognitive and product-specific individual difference variables, future research might focus on a smaller number of specific concepts and carefully consider which interactions merit testing. In the Schaninger and Buss study the authors report approximately 400 correlation coefficients. From these they suggest that the patterns support their hypotheses. Very few of the reported correlations are statistically significant and even fewer are of reasonable magnitude. Without further conceptual support it appears doubtful that these patterns have any meaningful implications, even if they are reproducible in a follow-up study.

Lastly, the assessment of the relationship between individual difference constructs and behavior has been a topic of on going interest in psychology and particularly personality research. Numerous researchers have commented on the consistently low correlations between individual difference measures and behavior which have been reported in past studies. In a review of this literature, Epstein (1979, 1980) makes several suggestions which appear to be equally applicable to consumer research investigating the effects of individual difference variables on consumer search. These suggestions center around the need to consider the benefits derived from aggregating observations over either subjects, stimuli, situations, trials, occasions, or measures. Aggregation accomplishes several purposes. It reduces the error of measurement and broadens the range of generalization; in so doing it provides a procedure for establishing replicable generalizations (see Epstein 1980 for a detailed explanation). As illustrated by the Schaninger and Buss study consumer researchers have tended to rely upon single setting observations of search behavior for single products. These have then been compared to single observations of specific traits measured once. Given the multiple source of "error" present in this approach the insignificant relationships which have been reported are not surprising.

Price and Feick

The purpose of the Price and Feick study was to examine the role of interpersonal sources in consumers external search. It specifically addresses several theoretical aspects of social comparison theory and the costs and benefits of external search.

The authors' identification of source similarity, expertise, and accessibility as possible determinants of source selection is an interesting approach and one which provides needed theoretical advances for better understanding consumer search. The authors also illustrate the importance of considering the role of interpersonal sources at various stages in the decision process, other than just at the initial search stage. Several issues might be explored further in future research.

The concept of source accessibility appears to apply to other sources in addition to interpersonal sources. Obviously perfect information about the market is not available to consumers. Future research might investigate consumers' perceptions of what sources exist and their perceived accessibility. This information might provide a starting point for exploring the usefulness of the accessibility construct.

Another aspect of the use of interpersonal sources which reflects an informational rather than normative motive might be suggested. One might hypothesize that an interpersonal source might be used as a "sounding board" for the consumers decision making process. However, rather than using the person as a "source" of information, the consumer may use the person as an aid in articulating and crystallizing one's search strategy. This process might reflect very little "exchange" of new information but may be a necessary step in certain instances (i.e., the initial steps in the search for a new product or service for which the product performance and appropriate search strategy are highly subjective and not high in risk).

Lastly, the Price and Fieck framework might be expanded upon by considering a study by Bither and Wright (1977). In this research preferences for alternative types of individuals sought out as product consultants in a new product evaluation task were studied. In the choice situation, subjects evidenced a strong preference for consultants similar in certain ability dimensions. Following from this study one might hypothesize that for certain types of informational needs (i.e., factual in nature) objectively superior, more expert sources might be most influential. On the other hand for more subjective judgments, less expert, more similar sources may be important.

SUMMARY

These studies illustrate that in order to improve our understanding of consumers information search behavior theoretical and conceptual advances will be necessary. Along with this effort researchers will have to reconsider related methodological issues such as the need for data aggregation.

REFERENCES

Bither, S. and P. Wright (1977), "Preferences Between Product Consultants: Choices vs. Preference Functions," Journal of Consumer Research, 4, 39-47.

Blake, Brian F., Perloff, Robert, Zenhausern, Robert, and Heslin, Richard (1973), "The Effect of Intolerance of Ambiguity Upon Product Perceptions," Journal of Applied Psychology, 58, 239-243.

Brody, Robert P. and Cunningham, Scott M. (1968), "Personality Variables and the Consumer Decision Process," Journal of Marketing Research (February), 50-57.

Epstein, S. (1979), "The Stability of Behavior: I. On Predicting Most of the People Much of the Time," Journal of Personality and Social Psychology, 37(7), 1097-1126.

Epstein, S. (1980), "The Stability of Behavior: II. Implications for Psychological Research," American Psychologist, 35(9), 790-806.

Evans, F. B. and H. J. Roberts (1963), "Ford, Chevrolet and the Problem of Discrimination," Journal of Business, 36 (April), 242-249.

Klippel, R. Eugene and Sweeney, Timothy W. (1974), "The Use of Information Sources by the Aged Consumer," The Gerontologist, (April), 163-166.

Phillips, L. and B. Sternthal (1977), "Age Differences in Information Processing: A Perspective of the Aged Consumer," Journal of Marketing Research, November, 444-457.

Schaninger, Charles M. and Sciglimpaglia, Donald (1981), "The Influence of Cognitive Personality Traits and Demographics on Consumer Information Acquisition Behavior," Journal of Consumer Research, 9, (September).

Shostack, G. Lynn (1977), "Breaking Free From Product Marketing," Journal of Marketing, 40, 73-80.

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