Consumer Information and Confidence: Moderating Effects of Perceived Comprehension and Risk

Eric R. Wendler, University of Southern California
ABSTRACT - Past research in consumer information processing and on effects of comparative, evaluative produce information provision has indicated that increases in confidence and other affective states in response to information provision (including consumer information [CI]) may occur without comprehension of the information. This may, however, be a phenomenon limited to low-risk, low-involvement products and situations. The study required subjects to make a product choice for a low-and a high-risk/ involvement product . Appropriate CI was provided to one group only. Analysis showed that confidence was more related to comprehension perceptions when risk and involvement were greater.
[ to cite ]:
Eric R. Wendler (1983) ,"Consumer Information and Confidence: Moderating Effects of Perceived Comprehension and Risk", in NA - Advances in Consumer Research Volume 10, eds. Richard P. Bagozzi and Alice M. Tybout, Ann Abor, MI : Association for Consumer Research, Pages: 364-369.

Advances in Consumer Research Volume 10, 1983      Pages 364-369


Eric R. Wendler, University of Southern California


Past research in consumer information processing and on effects of comparative, evaluative produce information provision has indicated that increases in confidence and other affective states in response to information provision (including consumer information [CI]) may occur without comprehension of the information. This may, however, be a phenomenon limited to low-risk, low-involvement products and situations. The study required subjects to make a product choice for a low-and a high-risk/ involvement product . Appropriate CI was provided to one group only. Analysis showed that confidence was more related to comprehension perceptions when risk and involvement were greater.


Recent years have seen growing concern by consumer advocates and policy makers with the amount of product and brand information provided to consumers in the marketplace. This study is an attempt to clarify some issues surrounding the effects of increased consumer information provision and the implications of such effects for policy. The focus is on product information that is evaluative and standardized, allowing cross-brand and cross-product comparisons. For example, calorie and nutrient data on a can of peaches permit comparisons to various other brands or peaches and/or to similarly labeled canned pineapple. It is assumed that such messages can cut across different sources and channels; that is, a manufacturer or an independent consumer group might present information via, respectively, broadcast advertising or a consumer magazine. For convenience, this type of information will be termed consumer information (CI) throughout this paper.

Programs which encourage or require the provision of comparative product information will be at least implicitly directed toward some outcomes. The need to carefully delineate the intended goals of CI programs has been asserted by a number of authors (Armstrong, Kendall and Russ 1976; Wilkie and Gardner 1974). Many product information programs in the past, however, ave been instituted with few or no stated specific goals in mind. When rationales for consumer information disclosure requirements are given, they rest on the assumption of certain effects, be they individual outcomes such as knowledge gain or aggregate-level effects such as increased market efficiency (Day 1975, 1976; McNeill and Wilkie 1979). The actual outcomes on consumers of such information then should be measurable against the yardstick of intended outcomes if the disclosures are to be justified.


Consumer confidence can be placed in many things, including a product, a particular brand or manufacturer, an industry, business in general, a piece of information, or a decision. Some have argued that one measure of adequacy of information may be consumer satisfaction with, or confidence in, the purchase decision (Day 1974; Thorelli 1975). Business is often concerned with the confidence, or trust, that members of the public have in the company and its products, but confidence can have wider implications. Confidence plays an important role in one of the most influential models of consumer behavior which has been formulated, by Howard and Sheth (see Farley, Howard, and Ring 1974). In the Howard and Sheth scheme, confidence in decision making, reflecting certainty of a correct decision, is most dependent upon brand comprehension, defined as buyer understanding of brand features (which can include CI-type features).

Howard and Sheth (1969) formulated a nomological network that contains confidence as the central element. Simply put, they view confidence as occupying a central, equilibrating role in buyer behavior. In their scheme, low confidence in judgment of brands motivates search effort; the resulting information acquisition reduces search motivation as confidence rises, and "informational equilibrium" results when enough information is received to cause confidence to reach a perceived level of adequacy, leading to cessation of information seeking. The central importance of confidence to information seeking has also been articulated by others (Brody and Cunningham 1968; Sheth and Venkatesan 1968). As Harrell (1979) put it, "... the confidence construct is tied not only to prediction of purchase consequences but also provides a possible framework within which to integrate information research" (pp. 136-137). If these views are accurate, an understanding of confidence may be crucial if information-seeking behaviors are to be understood and stimulated. Therefore, even if the goal of a given CI program is only awareness, confidence may still be an important state to take into account if active information seeking by the consumer is needed to effect awareness. Confidence, then, can be a potential and desirable outcome of consumer information processing, as well as a central determinant of information seeking.

Confidence and other outcomes. Mayer and Nicosia (1976) have stated that enhanced consumer confidence can yield positive benefits for the seller and society as a whole. They cite Day (1974), who, in discussing the effects of information disclosures, said that "the important conclusion (about information disclosures) is ... that more information enhances confidence (in the decision process, the brand, the product, and the seller)." It is important, then, to know the relationship of confidence to other outcomes that may be assumed or desired effects of CI provision.

The classification of responses used here follows the common three-component attitude scheme (Ray 1973). Cognitive elements include awareness, attention, and comprehension. Affective responses are evaluative, and include satisfaction and confidence. Conative, or behavioral, elements are action responses. Research on the responses of shoppers to food labeling hints at the enhancement of positive affective states without comprehension or actual use (Lenahan et al. 1977; Miller 1978, Monroe and LaPlaca 1972). Lenahan et al. (1972) directed attention to confidence and other affective, "non-use" benefits or nutrition labeling, indicating that these results might be more prevalent and longer lasting than cognitive and behavioral outcomes. It appears that even though a consumer may not understand or even think he or she understands a given piece of product information, just knowing that it is there may increase his or her confidence in the decision process, the store, and so on. This phenomenon has been touched on by a number of theorists (Daly 1976; Mayer and Nicosia 1976), but there have been few if any real direct tests for the presence of increased positive affect without understanding. However, Jacoby and his associates (Jacoby, Speller, and Kohn 1974; Jacoby, Speller, and Kohn Berning 1974), in dealing with product information of all types (not only the evaluative, comparative type which is the focus here), have repeatedly found in a laboratory setting that people can evidence enhanced "subjective" states such as decision certainty and satisfaction in response to increased information levels while at the same time evidencing increased dysfunctional behavioral outcomes (i.e., poorer choices). For the remainder of this paper, confidence will refer to a consumer's subjective certainty that he or she has made the decision that is best for him or her. This is how it will be operationalized in item wording in the experiment that follows. It therefore resembles Jacoby and associates' decision certainty variable, which they refer to as a "subjective state."

From a policy perspective, actual consumer understanding of CI is obviously important. But a person's perceptions of his or her understanding may also deserve attention. This is a cognitive state: a belief regarding understanding. Real understanding of CI can affect the behavioral outcomes of "efficiency" and "accuracy" of choice, which the consumer may never be aware of anyway, but perceived understanding has at least the clear potential to influence affective states (confidence, satisfaction, etc.), and therefore subsequent information seeking and/or future use of this type of information. So from a policy perspective perceived comprehension may be as important an outcome as actual comprehension. Comparative evaluative product information that is made available to consumers, then, can engender increased confidence in decision making and therefore satisfaction with the decision. If this can occur in the absence of actual and/or perceived understanding of the information and use of it, then the criterion-based effectiveness of disclosure policies can be questioned.

Perceived Risk and Involvement

The Jacoby studies demonstrated a lack of positive relationship between affective states and real understanding in response to different levels of product information. However, these studies used as their foci products such as detergent and prepared dinners. These are all low-cost items which would present relatively low levels or risk and involvement to the decision makers, especially in a laboratory situation.

A number of investigators point to greater processing effort and closer examination of information at higher levels of risk and involvement (Bettman 1979; Chaffee and McLeod 1973; Robertson 1976; Sheth and Venkatesan 1968). It may be that the low-risk, low-involvement situations of the Jacoby studies engendered little desire to really examine the information. Thus there would have been little chance of revelation of noncomprehension in such situations, and a resulting confidence increase by sole virtue of the presence of the information. In contrast, if the situation were to involve high risk by virtue of the product or the situation, there should be greater attention, more processing and evaluation of the information, and a consequently stronger connection between understanding and confidence.

Involvement. Lastovicka and Gardner (1979) postulate that for a low-involving product there is little perceived difference between alternatives, and a consequent low desire to seek out information pertaining to alternatives, since uncertainty is low. (Notice the use here of uncertainty, a component of perceived risk.) Wright (1974) similarly associates high involvement with more active information seeking, and goes on to postulate that the low need for high levels of information can make information overload more possible in low than high involvement situations. Bettman (1979) has said that low-involvement situations yield less (or even no) processing of incoming information, Chaffee and McLeod (1970), in discussing consumer information processing, make a further distinction between high and low involvement. They assert that in a situation of low involvement, rationality is "a negligible factor" in evaluating information presented to the consumer; conversely, for high involvement, there is much greater motivation to process incoming information. This mirrors Day's (1970) characterization of involvement as leading to "active, purposeful information search."

This study investigated the effects of CI on confidence. Perceived risk and involvement (reflecting motivation to process information), believed to have been consistently low in past studies that have demonstrated confidence without comprehension, were entered into the research design as possible moderators of the effects of CI comprehension on decision confidence. The effects of actual use of consumer information in decision making were also investigated. It was expected that once the CI was understood, greater levels of use would be related to higher confidence.

The major hypotheses tested, then, were:

In a consumer choice situation, confidence increase may result from presentation and awareness of comparative, evaluative product information (CI) in the following ways:

1. In a low risk and involvement situation, confidence will increase regardless of comprehension and use of the CI.

2. In a high risk and involvement situation, confidence will increase only to the extent that the CI is understood and used.



Subjects were 120 college students enrolled in four communications and business courses at two universities. Of the 115 subjects responding to an age question, about 925: were between 16 and 25 years old, with an approximately even split between the 16-20 and 21-25 ranges. More than half were over 21. The respondents were fairly evenly distributed across year in college, skewed somewhat toward the junior and senior end. Forty-six percent were men, and fifty-four percent women.


Involvement and risk were manipulated by having choice decisions made for two different products. The products chosen were light bulbs and stereo cassette decks. These were expected to greatly differ in perceived risk by virtue of their price difference and in the level of involvement they engender through their highly different centrality to the subjects' (college students) egos. The latter point has been demonstrated (with stereo speakers and light bulbs, among a number of products) by Lastovicka and Gardner (1979). Cassette decks and light bulbs both have standardized, evaluative rating data associated with them, and were expected to be within the experience of college students.

Summary of Procedure

Subjects were randomly assigned to one of two groups-with or without CI. For each of two products, they received written product information for six alternatives (brands or motels) and were asked to choose the one they would select if really buying. The situation was represented to them as a catalog shopping study. Before and after each of the two choice decisions, subjects completed questionnaires measuring the variables of interest, including perceptions of the situation, confidence, comprehension of the CI (if presented), and importance of information types to the decision. Most items were in the form of seven-point Likert-type scales. Order of products--decks, bulbs or bulbs, decks--was randomly counterbalanced within conditions.

Subjects in both conditions received information of the type normally available when shopping by catalog or in person. This included brand, price, and a short product description. The with-CI group additionally received the evaluative, comparative information which is the present focus. In the case of the bulbs, which were all 100-watt, this was lumens (a brightness measure) and expected lifetime, and for the cassette decks this was frequency response, signal-to-noise ratio, and wow and flutter. Information was real, taken -from actual current catalogs and presented in formats similar to the originals. It represented (discounting the CI manipulation) the actual information available to many consumers. (Sample product materials are presented in the Appendix.


The items asking the subjects about their previous experience with the products showed that 86% had shopped for light bulbs at least once (36% often), and 75% had had at least some experience shopping for stereo cassette decks (19"o often). There were no sex differences in light bulb experience, but men in the sample had significantly more experience with cassette decks than women.

Success or Risk and Manipulation

The means for corresponding pre-decision items across produces indicated that the subjects approached the two decision tasks with significantly different levels of certainty, perceived "risk," and feelings of the severity of consequences of bad decisions. These single items, however, do not singly represent the construct or risk that is one of the dimensions of interest in this study. The procedure of multiplicatively combining a measure or uncertainty with one of expectation of consequences was used here (Cox 1967). The items used were the following, each with a 7-point Likert-type response format or "not at all" to "very": "How certain are you that you will be able to choose the model of (cassette deck/light bulb which is the best for you?" and "How bad can the consequences be if one chooses the wrong (cassette deck/light bulb)?" The inverse of the former item constituted the measure of uncertainty. It was not thought feasible in this situation to be able to adequately control real expectations of consequences in an obviously laboratory setting. Therefore the item dealing with the broader implications of a poor decision for each product was used as an approximation of expectations in a real situation. The resulting risk scores upheld the picture of subjects considering the cassette deck decision to be much more risky (bulbs, X=7.7; decks X=15.1; t= -7.40, p < .05).

Post-Decision Outcomes With CI Presented

When asked, "Do you remember seeing information given in the product descriptions about (frequency response/ lumens/etc.)", a greater percentage of subjects remembered seeing the CI for the decks (frequency response, 71% signal-to-noise, 56%; wow/flutter, 55%) than for the bulbs (lumens, 48%), in spite of the greater amounts of total information presented for the decks. Because responses to all three cassette deck CI dimensions were highly correlated and to simplify analysis, the most-remembered CI, frequency response, was used as the single cassette deck dimension in subsequent analyses.

To test the hypotheses directly, an analysis of variance was performed on responses to each product situation. The dependent variable was confidence, measured by the post-decision confidence item: "'How confident are you that you made the best choice for you or a (cassette deck/light bulb)?" The respondents were divided into independent variable groups: The first group consisted of those who did not receive the CI. The remaining respondents were assigned to groups 2 and 3 on the basis of their perceived comprehension of the appropriate CI. Comprehension was measured on a 7-point Likert-type scale ("not at all" to "very") as a response to the question: "If you remember seeing (the information type), how well do you think that you understood (the information)?" Those who indicated awareness and relatively low -comprehension of the lumens information (below the median rating) were placed in group 9; those with above-median perceived comprehension were placed in group 3. The same procedure was used with the decks responses for frequency response comprehension (each ANOVA was performed independently of the other, so group assignments were not necessarily the same for both products).

The ANOVA tables and cell comparisons for both products, with awareness a necessity for inclusion, are given in Table 1. In the case of the light bulbs, the overall F-test showed no significant differences in confidence. The results for the decks were quite different. The overall between-groups F-test was significant at better than p'.01; the post -hoc group comparisons (using the relatively conservative Scheffe method) indicate that the confidence differences between groups 1 and 3 and between groups 9 and 3 were both significant, The groups 1 and 2 comparison was not significant. Figure 1 illustrates the bulb and deck confidence levels graphically.





Multivariate Regression Analyses

The experimental ANOVA approach permitted examination or the effect of perceived comprehension on confidence gain in the presence of the CI. However, a more complete understanding of the effects of CI availability on confidence in the decision requires a multivariate analysis. Besides awareness (kept constant in the A-NOVAs) and comprehension, actual use of the CI in the decision process is expected to play a part in the final confidence level. Furthermore, the use of a multivariate approach can provide the same information yielded by a one-way ANOVA, in addition making possible direct study of measures of the relative strength of the independent variables' effects. For these reasons a series of multiple regressions were performed on the 30 subjects who received CI.

The first regression results are presented in Table 2. The criterion variable was confidence in the decision for each product choice situation. The three independent variables were entered hierarchically, in this order: awareness, perceived comprehension, and importance of the CI. The results for bulbs show that all three independent variables together explained only about 84,' of the variance in confidence, and that almost all of that was contributed by importance/use. The F-test for the regression was nonsignificant, indicating that the predictive power of the independent variable set was not significantly greater than that due to chance

For the high-risk, high-involvement cassette decks, however, results were very different. While awareness of the frequency response CI could explain less than 1% or the variance in confidence, addition of perceived comprehension increased this to about 31% (R=.55). The further addition of importance of frequency response information to the decision added a nonsignificant 1% to the total.





To further test the hypotheses, another regression as measured continuously within a product was treated as an independent variable. The dependent variable again was confidence in the decision; the independent variables were perceived risk and comprehension (entered hierarchically in that order) and an interaction term constructed by multiplicatively combining the risk and comprehension measures in such a way as to preserve the direction of relationship of each to confidence. If a significant addition of explanatory power were provided by such a multiplicative term, this would indicate an interaction between the two components. For decks, the addition of the interaction term did add a significant amount of explanatory power to the equation (p<.01). Use of within-bulb perceived risk yielded no such significant interaction effect, and as in the previous analyses very little (5%) confidence in the bulb decision was accounted for at all.


The results generally supported the hypotheses of the study. Some of the findings were not specifically anticipated, but can be accounted for in the light of previous research in consumer and other information processing.

The post hoc ANOVA comparisons were consistent with the hypotheses concerning CI effects on confidence for the high-risk/involvement product. There appeared to be a perceived comprehension-dependent effect of CI awareness on confidence in the decision for the high-risk, high-involvement decision only. For the cassette deck decision, awareness of the CI significantly increased decision confidence only if the subject felt that he or she understood the meaning of the CI relatively highly. The nonsignificant differences among the light bulb decision groups means that provision and awareness of the lumens information had no effect on decision confidence, regardless of the level of perceived comprehension.

The regression analysis supported the ANOVA results, confirming that for the high-risk, high-involvement choice, perceived understanding of the CI was much more related to confidence than awareness alone. Once awareness and comprehension were entered, though, indicated use of the CI had little explanatory power to add. This suggests that awareness and understanding of the frequency response information was enough to increase confidence, even without actual use of it in the decision. This may reflect the overall higher level of knowledge and confidence in the product area, which resulted in those persons not feeling that they needed the numbers (possibly they felt that they knew enough to go by brand only, or by appeal to past experience alone).

The use of cassette deck intraproduct perceived risk as an independent explanatory variable provided a test of the effects of perceived comprehension and risk not possible with the perhaps too low-involving light bulbs. Those second regression results suggest that comprehension was associated with confidence more for those who perceived the product as riskier.

The consistent lack of significant effect of awareness and comprehension in the bulb decisions reinforces some past notions of low-involvement information-processing and decision making. Not only was there less overall awareness or the lumens information, (which may have been due to the information format used), but awareness, when present, seemed to make no difference in confidence. Similarly, understanding was not significantly related to decision confidence. Indicated importanCe (interpreted as use) or the lumens was the independent variable most related to decision confidence. This brings to mind suggested information processing hierarchies where behavior may precede other states (DeBruicker 1979).


The results support the second hypothesis: that generally, consumer information must be perceived as understandable before it can make one feel more confident in the purchase decision. More importantly, although there was no direct demonstration of confidence gain in the absence or comprehension, which previous research suggests, the analyses do support the notion that confidence gain is less dependent on information comprehension when risk and involvement are low.

The major implication for consumer information provision, if the effects summarized in the hypotheses continue to be borne out by further research, is that careful attention must be paid to the correspondence between intended and actual outcomes. Care should be taken that comparative, evaluative information which is presented be not only understandable, but that it also be the most crucial and relevant information for informed consumer decision making. Presumably policy makers would like to believe that disclosures programs, even for low-cost products, have end results that transcend confidence in the "decision process, the brand, the product, and the seller."



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

Brody, R. P., and Cunningham, S. M. (1968), "Personality Variables and the Consumer Decision Process," Journal of Marketing Research, 5, 50-57.

Chaffee, Steven H., and McLeod, Jack M. (1973), "Consumer Decisions and Information Use," in Consumer Behavior: Theoretical Sources, eds. Scott Ward and Thomas S. Robertson, Englewood Cliffs, NJ: Prentice-Hall.

Cox, Donald F. (1967), Risk Taking and Information Handling in Consumer Behavior, Cambridge, MA: Harvard School.

Daly, P. (1976), "The Response of Consumers to Nutrition Labeling," Journal of Consumer Affairs, 10(Winter), 170-178.

Day, George S. (1970), Buyer Attitudes and Brand Choice Behavior, New York: The Free Press.

Day, George S. (1974), "On Confidence and Consumerism," in Advertising and the Public Interest, ed. S. T. Divita, Chicago: The American Marketing Association.

Day, George S. (1975), "Full Disclosure of Comparative Performance Information to Consumers," Journal of Contemporary Business, 4, 53-68.

Day, George S. (1976), "Assessing the Effects of Information Disclosure Requirements," Journal of Marketing, 40, 42-52.

DeBruicker, F. Stewart (1979), "An Appraisal of Low-Involvement Consumer Information Processing," in Attitude Research Plays for High Stakes, eds. John C. Maloney and Bernard Silverman, Chicago: American Marketing Association.

Farley, John U., Howard, John A., and Ring, L. Winston L. (1974), Consumer Behavior Theory and Application, Boston: Allyn & Bacon.

Harrell, Gilbert D. (1979) "Industrial Product Class Involvement, Confidence in Beliefs, and Attitude-Intent Relationships," in Attitude Research Plays For High Stakes, eds. John C. Maloney and Bernard Silverman, Chicago: American Marketing Association

Howard, John A., and Sheth, Jagdish, N. (1969), The Theory of Buyer Behavior, New York: John Wiley & Sons.

Jacoby, Jacob, Speller, Donald E., and Kohn, Carol A. (1974), "Brand Choice Behavior as a Function of Information Load," Journal of Marketing Research, 11, 63-69.

Jacoby, Jacob, Speller, Donald E., and Kohn Berning, Carol A. (1974), "Brand Choice Behavior as a Function of Information Load: Replication and Extension,' Journal of Consumer Research, 1, 33-42.

Lastovicka, John L., and Gardner, David M. (1979), "Components of Involvement," in Attitude Research Plays for High Stakes, eds. John C. Maloney and Bernard Silverman, Chicago: American Marketing Association.

Lenahan, R. J., et al. (1972), "Consumer Reaction to Nutrition Information on Food Product Labels," Search Agriculture, 2, 15.

McNeill, D. L., and Wilkie, William L. (1979), "Public Policy and Consumer Information: Impact of the New Energy Labels," Journal of Consumer Research, 6, 1-11.

Mayer, Robert N., and Nicosia, Franceso M. (1976) "Consumer Information: Sources, Audiences, and Social Effects," in Protecting the Consumer Interest, ed. R. N. Katz, Cambridge, MA: Ballinger Publishing Co.

Miller, John A. (1978) Labeling Research--The State of the Art, Cambridge, MA: Marketing Science Institute.

Ray, Michael L. (1973), "Marketing Communications and the Hierarchy-of-Effects," in New Models For Mass Communication Research, ed. Peter Clarke, Beverly Hills: Sage.

Robertson, Thomas S. (1976), "Low-Commitment Consumer Behavior," Journal of Advertising Research, 16, 19-24.

Sheth, Jagdish N., and Venkatesan, M. (1968), "Risk Prediction Processes in Repetitive Consumer Behavior," Journal of Marketing Research, 5, 307-310.

Thorelli, Hans B., Becker, Helmut, and Engledow, Jack (1975), The Information Seekers, Cambridge, MA: Ballinger Publishing Co.

Wilkie, William L. (1975), How Consumers Use Product Information, Washington, D.C.: National Science Foundation.

Wilkie William L., and Gardner, David M. (1974), "The Role of Marketing Research in Public Policy Decision Making," Journal of Marketing, 38, 38-47.

Wright, Peter L. (1974), "Analyzing Media Effects on Advertising Responses," Public Opinion Quarterly, 38 (Summer). 192-205.