Why Perceived Risk Failed to Achieve Middle Range Theory Status: a Retrospective Research Note

Citation:

Kjell Gronhaug and Robert N. Stone (1995) ,"Why Perceived Risk Failed to Achieve Middle Range Theory Status: a Retrospective Research Note", in E - European Advances in Consumer Research Volume 2, eds. Flemming Hansen, Provo, UT : Association for Consumer Research, Pages: 412-417.

European Advances in Consumer Research Volume 2, 1995      Pages 412-417

WHY PERCEIVED RISK FAILED TO ACHIEVE MIDDLE RANGE THEORY STATUS: A RETROSPECTIVE RESEARCH NOTE

Kjell Gronhaug, Norwegian School of Economics

Robert N. Stone, Jackson State University

ABSTRACT -

The concept of "perceived risk" was introduced in 1960 by Raymond A. Bauer. The new construct spurred an almost immediate interest, in particular within the (sub)discipline of consumer behavior, resulting in a vast number of empirical studies. Approximately twenty years ago the construct was also characterized as having the status as a "middleBrange" theory. After its peak in the early seventies, perceived risk research has declined. This paper examines critically the notion of risk as used in this (perceived risk) research, and reanalyzes Bauer's seminal statement and its inherent assumptions. Deficiencies in past research are identified, and implications for further research highlighted.

INTRODUCTION

In his seminal presidential address at the American marketing Association's Annual Conference Bauer (1960) introduced the concept of "perceived risk". The concept was almost immediately embraced by researchers in the discipline of marketing. In particular within consumer behavior, a subdiscipline to become dominant in marketing for almost two decades (see Sheth, 1993 for interesting and provocative retrospective), the new construct spurred researcher's interest. Bauer and his associates at Harvard Business School conducted several studies using the construct of perceived risk as a central tenant to explain a variety of phenomena, such as search for information and brand loyalty. For overview of output from this research program, see Cox, 1967. Also outside the Harvard Business School the new construct was acclaimed. Since its introduction the perceived risk concept has belonged to the standard inventory of any textbook in consumer behavior, and it has continued to be employed in consumer research. Gemnnden (1985), for example, in his metaBanalysis of research on perceived risk and information search traced more than 100 reported studies. And the perceived risk construct continues to be used in consumer research (see, e.g., Srinivasan and Ratchford, 1991). It (the risk construct) has also been applied in other research contexts such as industrial buying, services marketing (Murray and Schlachter, 1990) and local planning (cf. Mitchell 1994).

Research on perceived risk probably peaked in the early seventies. At that time perceived risk was characterized as having the status as a "middleBrange theory" (cf. Robertson and Ward, 1973, p. 21), indicating that it had become a crucial part of a master conceptual scheme guiding researchers to derive coherent propositions and testable hypotheses and thus enhance continuous production of knowledge (cf. Merton, 1957). This, however, did not happen. In spite of the complimentary "middleBrange" label, most marketing researchers would claim that perceived risk research has long since been "cutting edge". Upon retrospection they would probably also conclude that studies of perceived risks did not lead to a coherent research effort, but rather to a number of disjointed pools of research. The "middleBrange" label was perhaps prematurely given to risk research. Perhaps, indeed, it was simply the sheer volume of research with the construct which led to the complimentary label of "middleBrange" rather than because of the qualities associated with a valued theory (Dubin, 1969).

Why did perceived risk never reach the "middleBrange" status? There might be several reasons. Howard and Sheth (1967), for example, claimed the construct to be static. Taylor (1974) and Ross (1975) claimed risk research in consumer behavior to be without a theoretical base. Definitional and operational inconsistencies have also been argued as causes of the failing progress and status initially believed. Bettman (1975) even went so far as to remark that future risk research would be fruitless until definitional problems were adequately dealt with.

The paper is a retrospective on perceived risk research and proceeds as follows: First, we focus on the notion of perceived risk and related concepts as used in the "mother disciplines" economics, decision theory and psychology, where these concepts play a key role and are important components of wellBdeveloped theories. Next, we return to and reanalyze Bauer's (1960) seminal statement. In particular we focus on the framing of the risk concept as used in this (perceived risk) research. We also focus upon some operational issues of importance for improved understanding of past research with perceived risk. Finally, we try to assess and explain why the perceived risk perspective never did attain the middleBrange theory status.

PERCEIVED RISK AND RELATED CONSTRUCTS

In much of the literature outside marketing, risk (and uncertainty) are related to some situation of choice, where the actor may choose between two or more alternatives. The number of choice situations the individual faces is almost infinite, varying highly in importance, e.g., choice of spouse, education, purchase of house, or brand of toilet paper. Choice is also in the core of economics, as allocations of scarce resources exactly relates to how people choose to distribute their limited resources. A choice presumes two (or more) alternatives associated with specific consequences. The consequences of the alternatives are compared to some "yardstick" (e.g., preferences) and rank ordered, allowing the individual to pick the best one (cf. Simon, 1964).

Choice relates to decision making. Atkinson (1964) traces the interest in decision making by economists and mathematically oriented psychologists to the midB1700's work of Jeremy Bentham. In citing from the seminal work of Edwards (1954), Atkinson attempts to capture the excitement for decision making by mathematical psychologists and economists:

. . . given two states A and B, into either one of which an individual may put himself, the individual chooses A in preference to B (or vice versa). For instance, a child standing in front of a candy counter may be considering two states. In state A the child has $0.25 and no candy. In state B the child has $0.15 and an tenBcent candy bar. The economic theory of decision making is a theory about how to predict such decisions (p. 205).

If the consumer was completely informed, as reflected in assumption of the omniscent, "economic man", the choices will be made with certainty. For example, in its most simple form of consumer choice as depicted in neoclassical economics, the individual knows her or his preferences, and s/he knows how much different quantities yield utility as reflected in the individual stable, ordered indifference curves, allowing her (him) to maximize utility within the period's budget constraint. (For detailed discussion, see Lancaster, 1976).

The individual is, however, not omniscent. First, the future is not known with certainty, and equally important, the consequences of any decision are not certain as they are to be realized in the future. Realized ex post problems in consumer purchase decisions as reflected in the literature on consumer dissatisfaction and complaint behavior (cf. TARP, 1986) exactly reflect the uncertainty of future consequences.

Einhorn and Hogart (1985) distinguish between risk and uncertainty. Under risk, the probability of an outcome is known. Uncertainty implies that the probability distribution is unknown. The term "ambiguity" is used as well, explained by van Raaij (1991) in the following way:

. . .; one may only exclude some probability or attach some probability to probability distributions. Ambiguity is thus a "secondBorder uncertainty" (pp. 402B403).

Intuitively the notion of perceived risk as used in consumer behavior research refers more to uncertainty and ambiguity than risk.

Hogart and Einhorn (1985) propose the following components to capture the uncertainty dimension.

1. the expected product or attribute outcome level,

2. its probability of occurence,

3. the certainty/uncertainty with which these probabilities are held,

4. the unknown attribute or outcome level not anticipated, and

5.the ambiguity surrounding unanticipated levels.

The first of the above components refer to expectations, i.e., beliefs related to some outcome by an activity, e.g., purchase. The actor may be more or less confident what the outcome will be, and outcomes not expected may as well occur.

In very much of the experimental research on individual choices, the choice set is given. The basic risk paradigm is between a sure outcome and a risky outcome. However, as noted by Einhorn and Hogart (1987):

There are important psychological differences in the way people experience the uncertainty inherent in gambling devices as compared with those faced in everyday life. In gambling devices, the nature of uncertainty is explicit since there is a wellBdefined sampling space and sampling procedure. In contrast, when assessing uncertainty in real world tasks, the precision of the gambling analogy can be misleading. Specifically, beliefs about uncertain events are typically loosely held and ill defined (p. 43, emphasis added).

Consumer buying are real world decisions, where exactly as stated above, the gambling analogy can be misleading of reasons such as illBdefined sampling space (choice set) and consequences of alternatives, and vaguely held beliefs of resulting outcomes.

INHERENT ASSUMPTIONS IN THE PERCEIVED RISK PERSPECTIVE

In his initial introduction of risk to explain consumer behavior, Bauer (1960) claimed:

Consumer behavior involves risk in that sense that any action of a consumer will produce consequences which he cannot anticipate with anything approximating certainty, and some of which are likely to be unpleasant. . . . It is inconceivable that the consumer can consider more than a few of the possible consequences of his actions, and it is seldom that he can anticipate these few consequences with high degree of certainty (p. 24).

And he continues:

Consumers characteristically develop decision strategies and ways of reducing risk that enable them to act with relative confidence in situations where their information is inadequate and the consequences of their actions are in some meaningful sense incalculable (p. 25).

Closer inspection of this statement reveals some very interesting inherent assumptions:

(1) First, it (the statement) indicates the belief that the consumer tries to exhibit purposeful, i.e., intendedly rational behavior. [Viewing purchase activities as purposeful has long been the dominant perspective in consumer behavior research (cf. Sheth, 1972), as it still is (cf. Bettman et al., 1991).] The assumption of the consumer as a purposeful problemBsolver is spelled out in details in Cox's (1967) summary of the risk research program at Harvard Business School initiated by Bauer.

(2) Second, Bauer's statement directly takes into account the limited cognitive capacity of the actor, i.e., the individual (the consumer) has limited capacity to seek, store, handle, and make sense of (understand) data. To act purposefully constrained by cognitive limitation, can be conceived comparable to Simon's (1957) notion of bounded rationality, ". . ., the behaviors exhibited by the actors are intendedly rational, but only limited so" (p. xxiv).

(3) Third, also reflected in the statement, is that the belief that the actor (the consumer) must (to some extent) be confident to act, i.e., to exhibit purposeful behavior. This as such is interesting, as it implicitly recognizes that the consumer's actions take place in a specific context. Uncertainty about consequences reflects lack of declarative knowledge, i.e., lack of knowledge of the outcome of the act (purchase). The believed importance of confidence to act is explicitly reflected by including personality variables such as "general" and "specific selfBconfidence" in much of the work on perceived risk (cf. Cox, 1967, for overview). Research has also, as reflected in tests of "the confidence hypothesis", demonstrated that confidence influences consumers'evaluations (cf. Heath and Tversky, 1991).

(4) Bauer's statement also reflects a perspective on the individual (consumer) that s/he acts within her or his surrounding context as perceived (or imagined), which is in concordance with philosophy of science perspectives dominating very much of contemporary social science research. Due to the saying that the consumer only "can consider more than a few possible consequences", Bauer also (implicitly) recognized that the consumer's perspective is subjective and often biased, which he (Bauer) states in the following way: "This is because the individual can only react to and deal with risk only as he perceives it subjectively" (p. 30).

(5) A fifth point to be noted is by focusing on consequences, Bauer (1960) (implicitly) assumes that predictive knowledge is needed to act purposefully. This point was elaborated by his associate D. F. Cox (1962) in his emphasis on the predictive value of information. The importance of prediction in causal thinking, which is a prerequisite for purposeful behavior is also dealt with by Einhorn and Hogart (1982).

(6) Even though Bauer (1960) initially indicates "consequences", the emphasis is on negative consequences as reflected in the following quotes: "and some of which are likely to be unpleasant", and "Consumers . . . develop strategies . . . of reducing risk". This point is, however, somewhat ambigous, and can be interpreted in different ways, e.g. 1) through such strategies (for example search for information) the consumer may become more certain, enabling her/him to chose an alternative, e.g. to chose alternative A, or to not buy, or 2) due to perceived uncertainty the consumer overstate grossly negative consequences, which will be perceived more "realistically" through use of their risk reduction strategies, and thus enabling them to make choices.

More important is that the research program Bauer's idea initiated, solely has focused on perceived negative consequences (for overview, see Cox, 1967). The weighting of positive and negative consequences as reflected in the game (or choice) perspective discussed above is seemingly left out.

Inspection of 1)B6) also show that Bauer's perspective introduced more than three decades ago shares many of the basic assumptions underlying very much of today's research in consumer behavior!

SOME CHARACTERISTICS OF PERCEIVED RISK RESEARCH

Bauer's (1960) seminal statement of perceived risk was no completely new idea, and was probably influenced by his decisionBtheorist friend, Robert Schleifer (recognized among others for his influential book, Analysis of Decisions under Uncertainty. [Schleifer's influence is explicitly recognized in Cox's (1967) book on the Harvard Business School research program on perceived risk.]

Framing

Bauer's statement of perceived risk was also influenced by the way the concept (of risk) often is used in everyday life, i.e., by focusing on the subjectively perceived likelihood of unpleasant events or outcomes, as reflected in the following quote from

MacCrimmon and Wehrung (1986) in applying the concept of risk to various personal and product arenas:

"The ultimate risk . . . seems to be the risk of remaining single. It has been estimated that being unmarried shortens a man's life expectancy by nine and a half years.

Apparently the only way to interpret this remark is to conclude that a man can be even more certain of shortening his life if he remains single. This use of the risk concept is exactly the opposite of what is presumed in the game (choice) perspective of risk emphasized above. What here is reflected, is the increasing certainty paradigm of experiencing a loss from the phenomena that brings about the claim for risk. Research with risk in marketing has taken note of this confusion:

. . . the equivalence of these constructs is not accepted . . . if a consumer were perfectly . . . certain that a brand is totally unacceptable for purchase there would be no uncertainty or perceived risk, by definition. However, if there is no uncertainty or perceived risk in the situation, why is the brand "totally unacceptable"? (Peter and Ryan, 1976, p. 184).

As such the dilemma can be resolved, e.g. by explicitly bringing in also the benefits of the alternatives considered. Consider a "riskBneutral" consumer. If the benefits, B, are believed to present with certainty, while the negative consequences, C are associated with uncertainty), than the consumer is willing to accept an alternative, if:

E(V)=B_aBp(C_a)>0

where:

B_a=expected value that the consumer will get from alternative a, and p(C_a)=the likelyhood p that the consumer will experience the negative consequences C related to the purchase of alternative a.

Beliefs and Costs

The consumer in her or his purposeful purchases hopes for and expects positive outcomes to occur. The consumer seeks benefits, whatever they are. In the perceived risk concept, the positive consequences are implicit. Why should the consumer at all consider a purchase alternative if s/he not at all associated with some benefits? Thus the perceived risk paradigm implicitly implies that some alternative considered, e.g. a brand also is associated with perceived positive consequences. It should be noted, however, that in the recent contribution by Dowling and Staelin (1994) benefits have been taken directly into account.

It is also interesting to note that in the influential studies conducted by psychologists interested in choice, e.g., as reflected in Tversky's and Kahneman's (1981) work in "framing", the emphasis is on the (final) choice and influencing factors, and not on how the subject arrived at that choice situation, which is implicit in the perceived risk perspective.

The emphasis on how people arrive at choices is apparently related to the processB perspective underlying very much of consumer research, implying that by knowing the consumer decision processes, marketers may influence these processes and their associated outcomes (cf. Arndt, 1976).

Additional ambiguities regarding the risk construct, however, still remain. In the game paradigm, positive and negative consequences are weighted in some way or other, and the actor makes his choice. The focus on negative consequences in the perceived risk perspective indicates that evaluation of positive and negative consequences may be separated as reflected in our everyday language. Consider the familiar phrase, "The benefits outweighed the risks". How is this phrase to be interpreted? Contrary to the portrayal of risk in a game perspective, the interpretation must acknowledge a separating of an alternative into benefit and risk aspects. The benefits can then be compared to the risks to see which outweighs which. In this conceptualization risk is not an outcome of a "coupling" process wherein benefits and losses are together considered to make a risk assessment. Rather, the phenomena of risk is isolated by an individual, that is, it is psychologically separated away from the benefits, and then dealt with. The "risk framing" thus is changed as a consequence of a composite benefitBloss consideration, to one based on a consideration for only the downside, loss aspects of an alternative. It is also noteworthy that in the vast literature on consumer decision making, only modest attention has been devoted to how consumers take positive and negative consequences into account (see Bettman et al. 1991 for overview). There is ample evidence that the framing (cf. Tversky and Kahneman 1981) influences choice, but our knowledge whether or not consumers separate evaluations of positive and negative consequences, and how they are compared, is modest.

Unspecified Relationships

The perceived risk perspective also assumes that the consumer will reduce the risk to a tolerable level. How the risk reduction processes occur are only vaguely indicated in the literature on perceived risk. Is it so that when the consumer becomes more confident that a negative consequence will occur that s/he drops that purchase alternative, or through search learns that the consequence is unlikely to occur, that s/he accepts the alternative?

The perceived risk literature also states that the consumer make use of various strategies to reduce risk perceived, e.g., search for information, buy the same brand over again, or rely on an expert. In spite of that research has been conducted to examine the use of such strategies, e.g., search for information (for overview see, Gemnnden 1985) and repetitive behavior (Sheth and Venkatesan, 1968), modest emphasis has been on examining the conditions under which the various risk reduction strategies apply.

In fact, Bauer (1960) in his forceful presentation of the risk idea, started the other way around by indicating how different phenomena, e.g. adviceBseeking and brand loyalty, could be explained as strategies for handling perceived risk.

In the beginning perceived risk was used as an unmeasured, hypothetical construct (cf. Bauer, 1960). Later on perceived risk was used as an intervening variable, which in much of the early research was captured by a multiplicative index capturing consequences and degree of uncertainty (see later discussion, and Cox, 1967 for overview). As the amount of empirical perceived risk studies increased, various components of perceived risk were identified (cf. Jacoby and Kaplan, 1972) and the impact of various antecedents, in particular personality variables, were included. Also the relationships between perceived risk and a variety of risk reduction strategies were examined empirically. With few notable exceptions (see the summary chapter in Cox, 1967; Stem et al., 1977) modest attempts were made to integrate this vast research effort into a coherent theoretical structure. Also as noted above, few attempts were made to identify when the various risk handling strategies were appropriate for whom (cf. Peter and Tarpey, 1975; Roselius, 1971). For example, a vast number of studies have focused on risk and information search (see Gemnnden, 1985, for overview). Search imply "search goods" (cf. Nelson, 1970), i.e., that the consumer can search for and evaluate alternatives and consequences, and that s/he is capable and willing to do so. Overt search takes time and requires effort. The many studies reporting search in ("risky") purchase situations thus reflect that consumers find it worthwhile (cf. Shugan, 1980). In some cases, however, consumers are unable to evaluate alternatives and consequences prior to purchase, e.g., as reflected in the notion of "experience goods" (Nelson, 1970), indicating that other risk reduction strategies may be more appropriate. For example, by relying on an expert may both indicate that the consumer is unable to evaluate the consequences herB/himself, or that s/he believes this is the best information source, or when comparing benefits and costs of various strategies that this is the most appropriate one. The recent study by Folkes (1988) examining the use of available information (the availability heuristic) to cope with risk, is however, a promising effort to enhance our understanding.

Measurement Issues

An important part of the perceived risk research is empirical. Inspection reveals several anomalies. In all empirical research measurements play a crucial role. In much of the empirical research perceived risk has been captured by two dimensions, uncertainty and consequences, both measured by ordered scales, often combined into an index, in a multiplicative way, a measurement practice which is highly suspect. Another concern is whether consumers actually perform such calculations (cf. Bettman, 1975; Sjoberg, 1980).

An interesting inconsistency is also found in the way the uncertainty dimension is captured, as reflected in the following example: "Uncertainty can be described as the probability that a given (negative) event will occur" (Cunningham, 1967, pp. 266B67). Obviously it is the certainty of the (unpleasant) event that is being measured as noted above. Even more seriously, however, is that most empirical perceived risk research, with a few notable exceptions, have ignored the criterion dimension of the construct, a neglect noted long ago by Bettman (1975, p. 384). For a more detailed discussion of measurement of perceived risk, see Dowling (1986).

DISCUSSION

Research with the perceived risk construct as the central tenant never attained the status of middle range theory. Several explanations can be supplied why this did not occur.

As noted in our analysis of the assumptions underlying the perceived risk perspective as initially introduced by Bauer (1960), it is evident that the basic elements of contemporary cognitive theories of consumer decision making virtually are represented in this perspective (which also reflects Bauer's superiority as a social scientist).

In our view very much of the failure of perceived risk to attain the middleBrange status relates to conceptual as well as measurement problems. Marketing and consumer behavior have for long been recognized as "borrowing disciplines" (cf. Cox et al., 1964; Myers et al., 1979). Borrowing a concept often involves ambiguities as it is brought into a new context, and the original meaning attached to the concept mayBBin worst caseBB implicitly be changed. Zaltman et al. (1982) put it this way:

Confusion also arises because we don't recognize the difference in meaning when using old words in a new context . . . many people get an initial theory by borrowing . . . it is important to recognize that the terms that are used may not have the same meaning in the marketing area (p. 12).

In fact, the risk concept can hardly be said to ever have been properly conceptually defined, i.e., pointing at unique attributes or qualities to be subsumed under the concept (cf. FrankfortBNachmias and Nachmias, 1992, p. 31).

Moreover, one construct alone does not make a theory, as a theory as a prerequisite requires a set of interrelated concepts, or as specified by Zaltman et al. (1982),

A set of interrelated concepts, definitions, and propositions that present a systematic view of specifying relations among variables with the purpose of explaining and predicting phenomena (p. 71).

In the perceived risk research tradition there are few attempts to specify the concepts and their inherent relationships (see Taylor, 1974, and Hugstad and Taylor, 1979, for exceptions). Such conceptual and theoretical deficiencies are apparently important factors to explain why the once so promising perceived risk idea failed to initiate a coherent body of knowledge centered around a unifying perspective or middle range theory. Research with perceived risk has to a large extent been empirical. As noted above, serious measurement concerns prevail. However, in spite of measurement deficiencies, the main explanation is in our opinion the conceptual and theoretical ambiguities underlying the stream of empirical studies. A similar observation was made by Kurt Lewin (1938) more than fifty years ago:

Unfortunately . . . the emphasis on operational definitions seems to have led in some cases to a somewhat dangerous disregard of the conceptual side of constructs . . .

and continues:

I do not wish to be understood as meaning that I would like to see a slackening of this effort toward better empirical operational definitions. However, the neglect of the conceptual side of psychological constructs is nearly as dangerous . . . as is the neglect of the empirical side (p. 15).

As unpleasant events, "real" or construed play a crucial role in human life, where the behaviors of consumers are an integral part, and because human actors only can take into account what they perceive or imagine, we still believe that the idea of perceived risk is fruitful. To enhance further research on perceived risk, however, the conceptual and theoretical foundation should be specified. In addition, a thorough metaBanalysis and synthesis of past research should be done, so we really could know what we after all these years know about the perceived risk phenomenon, and from there on add to present knowledge.

REFERENCES

Arndt, J. (1976), "Reflections on Research in Consumer Behavior", in B. B. Anderson (ed.), Advances in Consumer Research, Chicago, Association for Consumer Research, Vol. III, 213B221.

Atkinson, J. W. (1964), An Introduction to Motivation, Princeton, N.J., Van Nostrand.

Bauer, R. A. (1960), "Consumer Behavior as Risk Taking", in R. S. Hancock (ed.), Dynamic Marketing for a Changing World, Proceedings of the 43rd Conference of the American Marketing Association, 389B400.

Bettman, J. R. (1975), "Information Integration in Consumer Perception: A Comparison of Two Models of Component Conceptualization", Journal of Applied Psychology, 60 (July), 381B385.

Bettman, J. R., Johnson, E. J. and Payne, G. W. (1991), "Consumer Decision Making", in T. S. Robertson and H. H. Kassarjian (eds.), Handbook of Consumer Behavior, New Jersey, PrenticeBHall, 50B84.

Cox, D. F. (1967), Risk Taking and Information Handling in Consumer Behavior, Boston, MA, Division of Research, Graduate School of Business Administration, Harvard University.

Cox, D. F. (1962), "The Measurement of Information Value: A Study of Consumer DecisionBMaking", in W. S. Decker (ed.), Proceedings of the American Marketing Association (Fall), Chicago, American Marketing Association, 250B257.

Cox, R., Alderson, W. and Shapiro, S. J. (1964), Theory in Marketing, Homewood, Ill.: Richard D. Irwin.

Cunningham, S. M. (1967), "The Major Dimensions of Perceived Risk", in Risk Taking and Information Handling in Consumer Behavior, ed. Donald F. Cox, Boston, MA, Division of Research, Graduate School of Business Administration, Harvard University, 82B108.

Dowling, G. R. (1986), "Perceived Risk: The Concept and Its Measurement", Psychology in Marketing, Vol. 3 (Fall), 193B210.

Dowling, G. R. and Staelin, R. (1994), "A Model of Perceived Risk and Intended Risk Handling", Journal of Consumer Research, Vol. 21 (June), 119B134.

Dubin, R. (1969), Theory Building, New York, The Free Press.

Edwards, W. (1954), "The Theory of Decision Making", Psychological Bulletin, Vol. 51, 380B417.

Einhorn, H. J. and Hogart, R. M. (1987), "Decision Making under Ambiguity", in R. M. Hogart and M. E. Reder (eds.), Rational Choice. The Contrast between Economics and Psychology, Chicago, The University of Chicago Press, 41B66.

Einhorn, H. J. and Hogart, R. M. (1985), "Ambiguity and Uncertainty in Probabilistic Inference", Psychological Bulletin, 92, 433B461.

Einhorn, H. J. and Hogart, R. M. (1982), "Prediction, Diagnosis and Causal Thinking in Forecasting", Journal of Forecasting, 1, 23B36.

Folkes, V. S. (1988), "The Availability Heuristic and Perceived Risk", Journal of Consumer Behavior, Vol. 15, No. 1 (June), 13B23.

FrankfortBNachmias, C. and Nachmias, D. (1992), Research Methods in the Social Sciences, London, Edward Arnold.

Gemnnden, H. G. (1985), "Perceived Risk and Information Search: A Systematic MetaB Analysis of the Empirical Evidence", International Journal of Research in Marketing, Vol. 2, 79B100.

Heath, C. and Tversky, A. (1991), "Preference and Belief: Ambiguity and Competence in Choice under Uncertainty", Journal of Risk and Uncertainty, 4, 5B28.

Howard, J. A. and Sheth, J. N. (1967), The Theory of Buyer Behavior, New York, John Wiley.

Hugstad, P. S. and Taylor, J. S. (1979), "Risk Theory and the Science of Marketing", in O. C. Ferrell, S. W. Brown, and C. W. Lamb, Jr. (eds.), Conceptual and Theoretical Developments in Marketing, Chicago, American Marketing Association Proceedings Series, 448B458.

Jacoby, J. and Kaplan, L. (1972), "The Components of Perceived Risk", in Proceedings of the 3rd Annual Conference, ed. M. Venkatesan, Champaign, IL: Association for Consumer Research, 382B393.

Lancaster, K. (1976), "Theories of Consumer Choice from Economics: A Critical Review", in NSF, Selected Aspects of Consumer Behavior, Washington, D.C., NSF/RA 77B0013, 11B33.

Lewin, K. (1938), "The Conceptual Representation and Measurement of Psychological Forces," Monograph series: Contributions to Psychological Theory, Vol. 1, No. 4, Serial No. 4, Durham, N.C., Duke University Press.

MacCrimmon, K. R. and Wehrung, D. A. (1987), Taking Risks, the Management of Uncertainty, New York: Free Press.

Merton, R. K. (1957), Social Theory and Social Structure, New York, The Free Press.

Mitchell, V. W. (1994), "Testing Assumptions in Models of Perceived Risk", in J. Bloemer, L. Lemmink & H. Kasper (eds.) (1994), Marketing: It's Dynamic Challenges, Maastrich, 23rd EMAC Conference, 653B672.

Myers, J. G., Greyser, S. A. and Massey, W. F. (1979), "The Effectiveness of Marketing's "R&D" for Marketing Management: An Assessment", Journal of Marketing 43 (January), 17B29.

Nelson, R. (1970), "Advertising as Information", Journal of Political Economy, Vol. 81 (JulyBAug.), 729B54.

Peter, J. P. and Ryan, M. J. (1976), "An Investigation of Perceived Risk at the Brand Level", Journal of Marketing Research, 13 (May), 184B188.

Peter, J. P. and Tarpey, L. (1975), "A Comparative Analysis of Three Consumer Strategies", Journal of Consumer Research, 2 (June), 29B37.

Robertson, T. and Ward, S. (1973), "Consumer Behavior Research: Promise and Prospects", in Consumer Behavior: Theoretical Sources, Englewood Cliffs: PrenticeBHall, 3B42.

Roselius, T. (1971), "Consumer Rankings of Risk Reduction Methods", Journal of Marketing, 35 (January), 56B61.

Ross, I. (1975), "Perceived Risk and Consumer Behavior: A Critical Review", in Advances in Consumer Research, Vol. II, ed. M. Schlinger, Chicago, IL: Association for Consumer Research, 1B19.

Sheth, J. N. (1993), "Acrimony in the Ivory Tower: A Retrospective on Consumer Research", Journal of the Academy of Marketing Science, 20 (Fall), 345B353.

Sheth, J. N. (1972), "The Future of Buyer Behavior Theory", Proceedings of the 3rd Annual Conference, Chicago, Association for Consumer Research, 562B575.

Sheth, J. N. and Venkatesan, M. (1968), "RiskBReduction Processes in Repetitive Consumer Behavior", Journal of Marketing Research, Vol. 5 (August), 307B310.

Shugan, S. M. (1980), "The Cost of Thinking", Journal of Consumer Research, 7 (September), 99B111.

Simon, H. A. (1964), "On the Concept of Organizational Goal", Administrative Science Quarterly, 9 (June), 1B22.

Simon, H. A. (1957), Administrative Behavior, New York, The Free Press (2nd ed.).

Sjoberg, L. (1980), "The Risks of Risk Analysis", Acta Psychologica, 45 (August), 301B 321.

Srinivasan, N. and Ratchford, B. (1991), "An Empirical Test of a Model of External Search for Automobiles", Journal of Consumer Research, Vol. 18 (Sept.), 233B242.

Stem, D. E. Jr., Lamb, C. W. Jr., and MacLachlan, D. U. (1977), "Perceived Risk: A Synthesis", European Journal of Marketing, 11 (4), 312B319.

TARP (1986), Consumer Complaint Handling in America: An Update Study, Washington, D.C., Technical Assistance Research Programs, U.S. Office of Consumer Affairs, Contract HHSB100B84B0065.

Taylor, J. W. (1974), "The Role of Risk in Consumer Behavior", Journal of Marketing, Vol. 39 (April), 54B60.

Tversky, A. and Kahneman, D. (1981), "The Framing of Decisions and the Psychology of Choice", Science, 211 (30 January), 453B458.

Van Raaij, W. F. (1991), "The Formation and Use of Expectations in Consumer Decision Making", in T. S. Robertson and H. H. Kassarjian (eds.), Handbook of Consumer Behavior, New Jersey, PrenticeBHall, 401B418.

Zaltman, G., Heffring, M. and LeMasters, K. (1982), Theory Construction in Marketing, New Jersey: John Wiley.

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