Word of Mouth Communication As Negative Information

Marsha L. Richins, Louisiana State University
ABSTRACT - Negative word-of-mouth (NWOM) is a consumer response to dissatisfaction. This paper reviews existing literature on NWOM and related topics, especially examining the circumstances under which it occurs and the factors which influence the impact of this form of communication on recipients.
[ to cite ]:
Marsha L. Richins (1984) ,"Word of Mouth Communication As Negative Information", in NA - Advances in Consumer Research Volume 11, eds. Thomas C. Kinnear, Provo, UT : Association for Consumer Research, Pages: 697-702.

Advances in Consumer Research Volume 11, 1984      Pages 697-702


Marsha L. Richins, Louisiana State University


Negative word-of-mouth (NWOM) is a consumer response to dissatisfaction. This paper reviews existing literature on NWOM and related topics, especially examining the circumstances under which it occurs and the factors which influence the impact of this form of communication on recipients.


Because so few studies have systematically investigated negative word-of-mouth, no definition of the construct appears in published literature. By synthesizing and slightly modifying definitions of word-of-mouth advertising (Arndt 1967a, p. 190) and negative information (Weinberger, Allen, and Dillon 1981b, p. 398) negative word-of-mouth ma be defined as:

interpersonal communication among consumers concerning a marketing organization or product which denigrates the object of the communication.

Two characteristics of this definition are noteworthy. First, it says nothing about the direction of the communication. While traditional communications theory has emphasized a linear communications model from source to receiver, more recent communications research has emphasized the two-way nature of most interpersonal communications. This definition preserves this dynamic relationship among communicators.

Second, no statement about the veracity of the word-of-mouth communication is made in this definition. Researchers have not investigated how much NWOM is true relative to that which is false, perhaps because of the difficulty of making such measurements. Because a large amount of NWOM concerning products and companies is based on communicators' personal experiences, most of it is probably true, though subject perhaps to some exaggeration and to usual distortions in message communication. From a manager's point of view, the veracity of NWOM is probably unimportant: As long as the recipient of NWOM believes the information to be true it has the potential to affect his or her attitudes.

One may distinguish between random and systematic NWOM. As the name implies, there is no particular pattern to the occurrence or content of random NWOM. Most NWOM is of this type and frequently arises from the source's own negative experiences with the company or product involved. Since this type of communication is in fact random, it tends to be offset by positive word-of-mouth from other sources. Random NWOM tends to reflect occasional lapses in product quality control or customer relations .

Systematic NWOM occurs when large numbers of customers experience the same problem with a product or company. Such NWOM will normally not be offset by positive word-of-mouth. Systematic NWOM indicates potentially serious problems for a marketing organization. It may reflect problems in product design or serious quality control inadequacies or stem from marketing factors such as misleading advertising. If the magnitude of systematic NWOM is sufficiently great, the problem causing NWOM may be reported by mass media and reach the attention of various government monitoring agencies. rn such cases current customers may defect to competitors and potential new customers may avoid the firm. Clearly, systematic NWOM can have a serious impact on a marketing organization.

NWOM involves dynamic, two-way communication. However, one party normally initiates the topic or stimulates the communication of negative information. Studies conducted in the 1960's rather consistently reported that about 50% of word-of-mouth communications are initiated by the source of the information (see Arndt 1967a, b). Unfortunately, these studies failed to separate positive word-of-mouth from negative in determining these percentages and examinedly conversations concerning only relatively low-cost, low-risk products. Percentages ma,v differ for high importance items such as appLiances or automobiles.


The importance of negative word-of-mouth in the marketplace can be gauged by the amount or incidence of NWOM and by its impact on NWOM recipients. Each of these factors is discussed below.

Incidence of NWOM

Several studies have investigated the incidence of NWOM. Diener and Greyser (1978) found that 34% of those consumers dissatisfied with a personal care product told others about their dissatisfaction. Richins (1983a) reported that 57% of consumers dissatisfied with either a clothing item or an appliance told individuals outside their immediate family about the dissatisfaction (word-of-mouth rates did not differ for the two product classes) and in another study (Richins 1983b) found that those experiencing a dissatisfaction with a clothing item told an average of 5 other persons about the dissatisfaction. On the average, 3 of these individuals were household members and 2 were coworkers or other acquaintances. These studies show that the incidence of NWOM, once a dissatisfaction is experienced, is quite high and can be extensive, since most consumers tell several others about their dissatisfaction.

While the extent of NWOM makes it a factor of concern for marketing organizations, its impact may be blunted if it is offset b,v positive word-of-mouth. Thus it is of interest to know whether consumers are more likely to relate their negative experiences or their positive experiences to other consumers. Three studies have examined this question.

Holmes and Lett (1977) studied consumers with favorable and those with unfavorable reactions to a product sample of instant coffee they received through the mail. Those with favorable brand attitudes told significantly more others about their reaction than those with unfavorable attitudes. Engel, Kegerreis, and Blackwell (1969) examined the amount of word-of-mouth b,v customers satisfied and dissatisfied with an innovative automobile diagnostic center, finding no difference in extent of word-of-mouth by these two groups. Finally, a study by the Technical Assistance Research Program (1981) compared the extent of word-of-mouth by those satisfied and those dissatisfied with the corporation's response to their complaint. Results, contrary to the preceding studies, showed negative word-of-mouth more likely to occur than positive word-of-mouth.

Clearly the studies examining this phenomenon are contradictory. Their differences may be reconciled, however, by examining their differing contexts. The study involving the least effort or commitment on the part of the consumer (Holmes and Lett 1977) resulted in the lowest ratio of negative to positive word-of-mouth; the study where the situation involved a rather high degree of effort or commitment (making a complaint to the corporate customer relations department) showed the nighest ratio of negative to positive word-of-mouth. While the conclusion is tentative because of its post-hoc nature, evidence does suggest that the greater the consumer commitment to the product or situation causing dissatisfaction, the higher the likelihood the consumer will engage in NWOM when dissatisfaction occurs.

Impact of NWOM

The impact of NWOM on recipients has been noted many times. Rogers (1962), in his study of the diffusion of innovations, documented the role of "active rejectors," opinion leaders who effectively inhibit the diffusion of an innovation through negative comments about it. Arndt (1967b) studied word-of-mouth communications among wives living in married student housing and determined that those wives receiving NWOM were less likely to purchase a new coffee brand than those receiving positive word-of-mouth or those receiving no word-of-mouth communications about the product. More recently Tybout, Calder, and Sternthal (1981) studied NWOM in a laboratory experiment. Those subjects who heard from a stooge that McDonald's hamburgers were rumored to contain worm meat had a less favorable evaluation of eating at McDonald's, even when they reported they did not believe the rumor.

Negative word-of-mouth information seems to be especially potent in affecting recipients' attitudes for a number of reasons. Borgida and Nisbett (1977) compared the effectiveness of face-to-face word-of-mouth from communicators with summary data based on the experiences of a large group of individuals. This latter type of data is similar to that offered by some consumer publications such as Consumer Reports summaries of automobile repair experiences of subscribers. The researchers found that the face-to-face communications had a substantially greater effect on recipients than the written summary data and hypothesized that the greater vividness of face-to-face communication partially accounts for its superior impact.

Word-of-mouth communication is also especially impactful because it comes from a non-marketing dominated source. The superiority of such sources has been documented repeatedly.

The negativity of NWOM partially accounts for its substantial impact. Research on negative information has repeatedly demonstrated that negative information has more impact on consumers than positive information. Weinberger, Allen, and Dillon (1981a) further demonstrated that the impact of negative information persists even when it has been refuted. While several explanations for the stronger impact of negative information have been offered, they all in some way hypothesize that the rareness of unexpectedness of negative information leads to greater attention to and weighting of that information (see Mizerski 1982).


Because of NWOM's potential impact upon recipients, it would be valuable to understand consumers' motivations for engaging in it. Research in consumer behavior and on rumors is reviewed.

Word-of-Mouth Literature

Two authors have investigated motivations for engaging in word-of-mouth. Arndt (1967a) reviewed studies investigating word-of-mouth and suggested six possible motivations: altruistic. instrumental (the desire to appear well-informed), ego-defensive, ego-involvement, to establish cognitive clarity in an ambiguous situation, and to reduce cognitive dissonance.

Dichter (1966) used an involvement framework and identified four motivations for word-of mouth.

1. Product involvement, where "experience with the product (or service) produces a tension which .. must be channeled by way of talk..." (p. 148). Indirect evidence for the existence of this motivation comes from work by Holmes and Lett (1977), who determined that heavy users of a product engage in more word-of-mouth than light users, and from Engel et al. (1969), who found that innovators (assumed to be higher in product involvement than non-innovators) engage in more word-of-mouth than other consumers

2. Self-involvement, where word-of-mouth is used to meet the consumer's need to reassure himself in front of others. Several studies have shown that individuals high in anxiety engage in more word-of-mouth and are more likely to transmit rumors than those low in anxiety (e.g., Biberian, Anthony, and Rosnow 1975), pointing to the validity of self-involvement explanations for some occasions of word-of-mouth.

3. Other involvement, where consumers engage in word-of-mouth with the intent to help others.

4. Message involvement, where word-of-mouth is "mainly stimulated by the way the product is presented through advertisements ... but is not necessarily based on the speaker's experience with the product proper" (p. 148).

Both Dichter and Arndt were concerned primarily with positive word-of-mouth. For negative communications, it is useful to examine the body of research concerning rumors, a type of word-of-mouth activity frequently negative in content.

Rumor Literature

Knapp (1944) in his classic work identified three types of rumors: (1) Pipe dream or wish rumors are usually positive and express the hopes of the communicator. Knapp gave as an example the unconfirmed tales of Allied victories in World War II. (2) Bogies are rumors which mirror fear and anxiety and are negative or frightening. A rumor that table salt is carcinogenic may fall into this category. (3) Wedge-driving or aggressive rumors are negative in content and serve to divide groups or create scapegoats. Many rumors concerning racial groups fall into this category Knapp's work suggests that rumors mirror and provide outlets for emotions. According to Dichter, word-of-mouth can serve the same function. However, even with emotional arousal rumors do not always occur.

Allport and Postman (1947) and Rosnow and Fine (1976) all agree that for rumors to occur, evidence pertaining to an important topic is ambiguous or lacking. Rosnow and Fine further stipulate that emotional arousal in the form of anxiety must also be present for rumor to occur.

What, then can we conclude are the motivations for engaging in unsolicited NWOM? It appears that under most circumstances the impetus for NWOM arises from a first-hand negative experience with a product or company, though in some cases the dissatisfaction may have been experienced by an acquaintance or family member of the communicator. Once this dissatisfaction occurs, synthesis of the rumor and word-of-mouth literatures suggests three major motivations: (1) catharsis and anxiety reduction, basically an attempt to reduce emotional arousal by sharing verbally the experience with others, (2) altruism, an attempt to prevent others from experiencing a similar fate, and (3) vengeance, an aggressive motivation similar to that documented by Knapp (1944) in which the communicator attempts to "get back at" the offending marketing institution by turning other consumers against it. The importance of the product involved or the strength of the dissatisfaction may increase the likelihood of NWOM occurring by strengthening the motivations, e.g., by heightening emotional arousal and anger and/or increasing the importance of warning others. Richins (1983a) in fact found that the more severe the problem experienced with a Product, the greater the likelihood of engaging in NWOM.

It should be noted that the desire to engage in NWOM is frequently at odds with self-presentation needs. Individuals normally desire to present themselves in as positive a fashion as possible (Goffman 1959). Telling others that a product one purchased was unsatisfactory is, in essence, admitting failure as a consumer. Apparently, when NWOM does occur the motivations listed above are stronger than self-presentation needs, at least momentarily. Attributions of blame and individual self-esteem may also affect whether NWOM occurs in the face of this conflict.

Situational Factors

The research described above primarily refers to communicator characteristics inciting NWOM. There is some evidence that situational or environmental factors may also give rise to such conversations. Belk (1971) reported that food-related cues (shopping for food, drinking coffee) were present in at least 75% of the reported incidents of word-of-mouth activity about a new brand of coffee. Consistent with this, Richins (1983b) found that the most frequent situational context in which NWOM about an unsatisfactory clothing item occurred was while dressing or wearing the item. These findings suggest that use of the product itself or other products in the product class stimulate NWOM, perhaps by reminding consumers of their dissatisfaction and again arousing the emotions experienced earlier.


Communication theory has documented the distortions which occur in communication, and word-of-mouth is no exception. Early studies of rumor (see Allport and Postman 1947) found at least three possible distortions. Leveling is the elimination of some of the message details, sharpening is selective attention to or heightening of particular message details, and assimilation is message bias which occurs presumably to build a better overall il formation structure or to fit new information into existing structure. More recently Pace and Boren (1973) reviewed research on transmission biases and listed nine message transformations which may occur during serial transmission, most of which simplify the message to ease transmission.

While the existence of severe and even wild distortion in rumor transmission has been frequently noted, numerous studies show that word-of-mouth communications are usually quite accurate (see Davis 1977). Buckner (1965) suggests that distortions are less likely to occur when the communicator and receiver are motivated to separate true from false information. This type of screening probably can be assumed to operate in many word-of-mouth communications concerning products. Certainly, receiver relying on word-of-mouth information to make a purchase decision will be motivated to question unreasonable statements.

It is important to note that distortions carl occur in the initial transmission of a message is well as in retelling. When a consumer relates a negative experience to another, a number of biases may occur. First, the individual may not have possessed all the facts of the situation initially. Thus, the communicator may incorrectly believe himself to be communicating a true picture of the situation. Second, even if the individual knew the relevant facts initially, at the time of transmission some facts may not be communicated because of memory and cognitive processing limitations. .A third important consideration involves attributions of blame. Where NWOM is concerned both the communicator and receiver usually are interested in who is responsible for the dissatisfaction's occurrence. Attribution research has repeatedly shown that participants in a situation are more likely to attribute failure (i.e., purchasing a unsatisfactory product) to "circumstances" while they tend to attribute success to personal factors such as effort or skill. While several factors account for this finding, one explanation is that such attributional biases partially reflect attempts to maintain self-esteem. If a consumer feels his or her self-esteem or image is damaged by a dissatisfaction experience, some biasing of the NWOM communication may occur. This can be as minimal as a change in the emphasis or exclusion of particular details of the message; or the communicator may knowingly pass along untrue information to preserve the desired image.

The communicator's motivation for engaging in NWOM may also result in selective emphasis or other distortions. When the motivation is primarily altruistic, the communicator may emphasize the seriousness of the dissatisfaction or the marketing organization's role in causing it to strengthen the warning to recipients. If the motivation is primarily that of revenge, exaggeration of the organization's misdeeds may occur.

Finally, audience factors may cause distortion. Communicators are known to adjust their messages to recipient characteristics (Crawford 1974). The desire to impress the recipient or to make the message appear more interesting may result in conscious leveling and sharpening, as will the possible desire to say what the recipient wishes to hear.


The study of NWOM is important for its potential impact upon recipients. While NWOM impact can be considered a form of attitude change, the attitude change literature will not be reviewed exhaustively but only as it specifically relates to NWOM.

Several theories have addressed the conditions under which attitude change will occur and the persistence of this effect. Petty and Cacioppo (1980) note that there are two routes through which attitude change from a message is likely to occur. Persuasion through the central route occurs when the message recipient is motivated and able to think about the issue. Persuasion through the peripheral route occurs when either motivation or ability is low. In these cases attitude change occurs due to the reward for advocating a particular viewpoint, because of source attractiveness, or for other reasons unrelated to cognitive processing of the message itself. .Attitude changes via the central route are relatively enduring; those through the peripheral route are not likely to endure useless the change is subsequently bolstered by cognitive processing.

Kelman's (1958) identification of three sources of attitude change can be examined in terms of Petty and Cacioppo's framework. Compliance occurs when one believes that agreement with the communicator will bring rewards or allow the avoidance of punishment. Attitude change based on identification occurs when one likes or is attracted to the source of the attitude change message. Both compliance and identification are peripheral routes to attitude change. Internalization, the third source of attitude change, results when message recipients believe message acceptance will help them to realize important, central values; it involves at least a moderate degree of cognitive processing as in the central route described by Petty and Cacioppo.

The above approaches and numerous studies of attitude change suggest that the impact of NWOM depends at least in part on the amount of cognitive processing the NWOM message stimulates. Factors expected to influence degree of processing are discussed below and are classified into source message, and receiver characteristics. Channel factors are presumed to be relatively constant, consisting of face-to-face interactions among consumers.

Source Factors

Attribution theorists have studied source credibility and suggest that the recipient of a message attempts to infer reasons for the communication (e.g., Eagly, Wood, and Chaiken 1978). When a recipient hears positive information concerning a product recently purchased by the communicator. less credibility is placed on that information because the recipient may infer that the source is attempting to justify his or her purchase by saving positive things about it. The positive information is "expected" from the source and therefore discounted. This discounting is less likely to occur if the source makes negative, and therefore unexpected, statements about the newly-purchased product. In such situations the message is more likely to be believed.

Likability and physical attractiveness are source characteristics which have been shown to result in attitude change (Chaiken 1979). Such characteristics may lead to identification. As noted previously, however, such influences on attitude are likely to be short-lived unless the recipient has an ongoing relationship with the source.

If the source of the NWOM is a reference group member, the likelihood of attitude change is enhanced. There are two avenues through which attitude change may be effected by reference groups. First, reference group members are normally well-liked, and attitude change through identification may occur. Second, if group norms are especially strong there may be pressure to conform with resulting attitude change through compliance. The extent and persistence of this attitude change will probably depend on whether reference group members can continue to observe the receiver's expression of attitude (see Cook and Flay 1978) and by the degree of consensus among reference group members on the topic in question.

Message Factors

Shibutani (1966) proposed that for a rumor to be accepted by recipients it must fit with that person's beliefs and inclinations and be plausible. Certainly, such stipulations are consistent with social judgment theory (Sherif and Sherif 1967), which suggests that the greatest amount of attitude change occurs when a message is slightly discrepant from receivers' beliefs rather than highly discrepant.

Repetition of NWOM is likely to influence the impact of NWOM. Numerous studies (see Cialdini et al., 1981) have shown a relationship between repetition and attitude change, especially if the message is complex in which case repetition will facilitate cognitive processing.

In NWOM situations repetition may act in other ways. If a consumer hears the same NWOM message from a variety of sources, patterns of distinctiveness, consensus, and consistency (see Kelley 1973) will emerge and increase the likelihood that attributions of blame for the dissatisfaction should be placed on the product or marketing organization rather than on the product user (i.e., communicator). This would then lead to a negative shift in attitude.

When NWOM does influence recipients' attitudes, it normally has a negative influence. However, several consistency approaches to attitude change (e.g., dissonance and self-perception theories) suggest that under some circumstances NWOM would have an opposite influence on attitude, causing a positive shift (see Cook and Flay 1978). For example, if a consumer hears NWOM about a well-liked and highly valued product he or she owns, cognitive imbalance and a threat to self-esteem may result. This may be handled by derogation of the source of the information or by questioning message veracity Another avenue to restore balance would be to increase the emphasis on positive aspects of the product.

Receiver Characteristics

Involvement with the product class is probably the receiver variable most likely to affect the impact of NWOM. Social judgment theory (Sherif and Sherif 1967) suggests that individuals with high involvement with the topic of a message are least likely to be affected by that message, presumably because they have a more firmly fixed attitude than those lower in involvement. Petty and Cacioppo (1979), however, found that those with high issue involvement were more likely to be persuaded than others if the message contains good arguments, apparently because such individuals are motivated to process the information in the message.

Issue involvement has been said to bear a relationship to product involvement, and here it is necessary to distinguish between enduring and situational involvement with a product class (Bloch and Richins 1983; Rothschild 1979). Enduring involvement consists of a long term interest in and enthusiasm for a product class such as that exhibited by audiophiles and wine connoisseurs. Situational involvement refers to short term interest in the product class due to situational factors, such as an anticipated purchase of an expensive item.

Enduring involvement can affect the impact of NWOM in two, sometimes competing, ways. First, receivers possessing high enduring involvement will be motivated to process the information contained in the NWOM message and therefore will experience a change in attitude if the evidence is compelling. Second, however, they are likely to possess a high degree of knowledge about the topic and will thus be unconvinced by information short of compelling. Research on McGuire's (1964) inoculation theory has shown that if receivers have information or experience with which to refute an attitude change message, persuasion is less likely to occur.

Situational involvement occurs when a purchase is near and risk perceptions are high (Bloch and Richins 1983). Two factors are normally accepted to determine the level of perceived risk--level of uncertainty about choice outcomes and the amount at stake in the purchase. Uncertainty will be high if product knowledge is low and/or product complexity is high (among other factors); thus susceptibility to NWOM will be enhanced. Amount at stake in a purchase is high if the product is expensive, will be used in an important situation, etc. Evidence from several studies (e.g., Midgely 1983) suggests that the impact of NWOM will be positively related to the degree of perceived risk. However, the only systematic study of perceived risk and NWOM, while supporting this relationship, had a sample size too small for results to be conclusive (Arndt 1967b).

Shibutani (1966) identified the factors which make individuals susceptible to rumors as anxiety, tension, frustration, and lack of trusted information sources. Many of these conditions are present when perceived risk is very high and may increase susceptibility to NWOM.

Research in information processing also supports the link between perceived risk and the impact of NWOM. Studies have found a greater emphasis on negative information when personal relevance of the decision is high (Einhorn 1971) or when the decision involves greater finality of commitment for the decision-maker (Slovic 1969). Many products for which perceived risk is typically high (e.g., automobiles, home purchases) are characterized by a high level of commitment.


As this review clearly indicates, only a few studies have directly investigated NWOM. Most of our understanding about this phenomenon comes from studies of positive word-of-mouth or attitude change. Because of the widespread occurrence and importance of NWOM, at least three areas especially need further investigation. First, more knowledge about the impact of NWOM is needed. From research in attitude change, one can infer that NWOM frequently influences attitudes. However, the extent of this impact on attitudes and on subsequent behaviors has not been adequately examined. For instance, it is now known if attitude change from NWOM persists longer than changes from other sources or whether the impact on behavior is stronger for NWOM than for other types of information.

Second, the communication network for NWOM has not been clearly identified. While a few studies have investigated the number of individuals to whom a dissatisfied consumer relates a dissatisfaction, research has not investigated whether this information is retransmitted by the receiver and in what form.

Third, reporting biases in NWOM have not been investigated. It is not known to what extent intentional and unintentional distortions occur and whether such distortions are influenced by communicator characteristics such as blame attributions, self-esteem, emotional arousal, and so forth. Given the importance of NWOM in the marketing environment, greater understanding of the phenomenon is clearly indicated.


Allport, Gordon W. and Leo J. Postman (1947), The Psychology of Rumor, New York: Henry Holt.

Arndt, Johan (1967a), "Word of Mouth Advertising and Informal Communication," in Risk Taking and Information Handling in Consumer Behavior, ed. D. F. Cox, Boston: Harvard University, 188-239.

Arndt, Johan (1967b), "Perceived Risk, Sociometric Integration, and Word of Mouth in the Adoption of a New Food Product," in Risk Taking and Information Handling in Consumer Behavior. ed. D. F. Cox, Boston: Harvard University, 289-316.

Belk, Russell W. (1971), "Occurrence of Word-of-Mouth Buyer Behavior as a Function of Situation and Advertising Stimuli," paper presented at the American Marketing Association Educators Conference.

Biberian, M. J., Susan .Anthony and Ralph L. Rosnow(1975), "Some Determining Factors in the Transmission of a Rumor," cited in Rumor and Gossip, R. L. Rosnow and G. A. Fine, New York: Elsevier.

Bloch, Peter H. and Marsha L. Richins (1983), "A Theoretical Model for the Study of Product Importance Perceptions," Journal of Marketing, 47 (Summer).

Borgida, Eugene and Richard E. Nisbett (1977), "The Differential Impact of Abstract vs. Concrete Information on Decisions," Journal of Applied Social Psychology, 7(3), 258-271.

Buckner, H. Taylor (1965), "A Theory of Rumor Transmission," Public Opinion Quarterly, 29, 54-70.

Chaiken, S. (1979), "Communicator Physical Attractiveness and Persuasion," Journal of Personality and Social Psychology, 37, 1387-1397.

Cialdini, Robert B., Richard L. Petty and John T. Cacioppo (1981), "Attitude and Attitude Change," Annual Review of Psychology, 32, 357-404.

Cook, Thomas D. and Brian R. Flay (1978), "The Persistence of Experimentally Induced Attitude Change," Advances in Experimental Social Psychology, Vol. 11 New York: Academic Press. 1-57.

Crawford, T. J. (1974), "Sermons on Racial Tolerance and the Parish Neighborhood Context," Journal of Applied Social Psychology, 4, 1-23.

Davis, Keith (1977), Human Behavior at Work, New York: McGraw-Hill.

Dichter, Ernest (1966), "How Word-of-Mouth Advertising Works," Harvard Business Review, 44 (November-December), 147-157.

Diener, Betty J. and Stephen A. Greyser (1978), "Consumer Views of Redress Needs," Journal of Marketing, 42 (October), 21-27.

Eagly, Alice H., W. Wood and S. Chaiken (1978), "Causal Inferences About Communicators and Their Effect on Opinion Change," Journal of Personality and Social Psychology, 36, 424-435.

Einhorn, Hillel J. (1971), "Use of Nonlinear, Noncompensatory Models as a Function of Task and Amount of Information," Organizational Behavior and Human Performance, 6, 1-97.

Engel, James F., Robert J. Kegerreis and Roger D. Blackwell (1969), "Word-of-.Mouth Communication by the Innovator," Journal of Marketing, 33 (July), 15-19.

Goffman, Erving (1959), The Presentation of Self in Everyday Life, Garden City, N. Y.: Doubleday Anchor.

Holmes, John D. and John D. Lett, Jr. (1977), "Product Sampling and Word of Mouth," Journal of Advertising Research, 17 (October), 35-39.

Kelley, Harold H. (1973), "The Process of Causal Attribution," American Psychologist, 28 (February), 107-128.

Kelman, Herbert C. (1958), "Compliance, Identification, and Internalization: Three Processes of Opinion Change," Journal of Conflict Resolution, 2, 51-60.

Knapp, Robert H. (1944), "A Psychology of Rumor," Public Opinion Quarterly, 8, 23-37.

Midgely, David F. (1983), "Patterns of Interpersonal Information Seeking for the Purchase of a Symbolic Product," Journal of Marketing Research, 20 (February), 74-83.

McGuire, William J. (1964), "Inducing Resistance to Persuasion: Some Contemporary Approaches," Advances in Experimental Social Psychology, 1, 192-229.

Mizerski, Richard W. (1982), "An Attribution Explanation of the Disproportionate Influence of Unfavorable Information," Journal of Consumer Research, 9 (December). 301-310.

Pace, R. Wayne and Robert Boren (1973), The Human Transaction, Glenview, IL: Scott, Foresman.

Petty, Richard E. and John T. Cacioppo (1979), "Issue-Involvement Can Increase or Decrease Persuasion," Journal of Personality and Social Psychology, 37, 1915-1926.

Petty, Richard E. (1980), Attitudes and Persuasion: Classic and Contemporary Approaches, Dubuque: Brown.

Richins, Marsha L. (1983a), "Negative Word-of-Mouth by Dissatisfied Consumers: A Pilot Study," Journal of Marketing, 47 (Winter), 68-78.

Richins, Marsha L. (1983b), "Word-of-Mouth as an Expression of Product Dissatisfaction," in International Fare in Consumer Satisfaction and Complaining Behavior, eds R. L. Day and H. K. Hunt, Bloomington, IN: Indiana University, 100-104.

Rogers, Everett M. (1962), Diffusion of Innovations. New York: Free Press.

Rosnow, Ralph L. and Gary Alan Fine (1976), Rumor and Gossip: The Social Psychology of Hearsay, New York: Elsevier.

Rothschild, Michael L. (1979), "Advertising Strategies for High and Lo; Involvement Situations," in Advertising Research Plays for High Stakes, eds. J. Maloney and B. Silverman, Chicago: American Marketing Association, 74-93.

Sherif, Carolyn W. and Muzafer Sherif (1967), Attitude, Ego-Involvement and Change, Ne York: Wiley.

Shibutani, Tamotsu (1966), Improvised News: A Sociological Study of Rumor, New York: Bobbs-Merrill.

Slovic, Paul (1969), "Differential Effects of Real vs. Hypothetical Payoffs on Choice Gambles," Journal of Experimental Psychology, 80(3), 434-437.

Technical Assistance Research Programs (1981), Measuring the Grapevine--Consumer Response and Word-of-Mouth, Washington, D. C.

Tybout, Alice M., Bobby J. Calder and Brian Sternthal (1981), "Using Information Processing Theory to Design Marketing Strategies," Journal of Marketing Research, 18 (February), 73-79.

Weinberger, Marc G., Chris T. Allen and William R. Dillon (1981a), "The Impact of Negative Marketing Communications: The Consumers Union/Chrysler Controversy," Journal of Advertising, 10(4), 20-28.

Weinberger, Marc G. (1981b), "Negative Information: Perspectives and Research Directions," in Advances in Consumer Research, Vol. 8, ed. K. B. Monroe, Ann Arbor: Association for Consumer Research, 398-404.