An Empirical Investigation of the Impact of Negative Public Publicity on Consumer Attitudes and Intentions

Mitch Griffin, Bradley University
Barry J. Babin, Louisiana State University
Jill S. Attaway, Illinois State University
ABSTRACT - Negative publicity episodes have become increasingly common over the past several years. To a marketing practitioner, such incidents represent a serious threat to a firm's well being. A better understanding of consumer reactions to negative publicity may enable marketers to more effectively deal with these threats. Limited academic research, however, has investigated the strategic impact of negative publicity. This paper reports an experimental study intended to explore the influence of various negative publicity scenarios and target firm responses on consumer attitudes and purchase intentions.
[ to cite ]:
Mitch Griffin, Barry J. Babin, and Jill S. Attaway (1991) ,"An Empirical Investigation of the Impact of Negative Public Publicity on Consumer Attitudes and Intentions", in NA - Advances in Consumer Research Volume 18, eds. Rebecca H. Holman and Michael R. Solomon, Provo, UT : Association for Consumer Research, Pages: 334-341.

Advances in Consumer Research Volume 18, 1991      Pages 334-341

AN EMPIRICAL INVESTIGATION OF THE IMPACT OF NEGATIVE PUBLIC PUBLICITY ON CONSUMER ATTITUDES AND INTENTIONS

Mitch Griffin, Bradley University

Barry J. Babin, Louisiana State University

Jill S. Attaway, Illinois State University

ABSTRACT -

Negative publicity episodes have become increasingly common over the past several years. To a marketing practitioner, such incidents represent a serious threat to a firm's well being. A better understanding of consumer reactions to negative publicity may enable marketers to more effectively deal with these threats. Limited academic research, however, has investigated the strategic impact of negative publicity. This paper reports an experimental study intended to explore the influence of various negative publicity scenarios and target firm responses on consumer attitudes and purchase intentions.

Each year billions of dollars are spent on promotional activities in the United States in an attempt to influence consumer attitudes. Favorable publicity, coming from an uncompensated source lending a high level of credibility, is a very effective element of the promotional mix. Nearly everyone, from television evangelists to political candidates, neighborhood restaurants to the largest industrial manufacturers, uses various forms of publicity to secure donations, votes, or customers. Each of these organizations, however, is also vulnerable to the potentially devastating impact of negative publicity.

The influence of negative publicity is difficult to overstate. Empirical studies have found that negative information is capable of significantly affecting consumers' consumption related beliefs and attitudes. In fact, it has been shown that a single item of negative information is capable of neutralizing five similar pieces of positive information (Richey, Koenigs, Richey and Fortin 1975). Other research has found negative information results in more strongly held attributions regarding product beliefs than does positive information (Mizerski 1982) and the effect of negative information is more enduring than positive information (Cusumano and Richey 1970; Richins 1983). Researchers have also shown that negative information more strongly influences attitudes and purchase intention than does positive information, particularly in the service sector (Weinberger and Dillon 1980). In addition to empirical findings, even casual observation of recent events (i.e. Perrier contamination, Exxon's Valdez accident, and the ongoing saga of the PTL Club) clearly illustrate the damaging impact of negative publicity.

The questions "How do negative publicity events influence consumer attitudes and purchase intentions?" and "What actions can the target firm take to minimize the impact of negative publicity?" are of relevance to both marketing academicians and practitioners. Addressing these questions requires a better understanding of the relationship among the various elements pertaining to negative publicity, and their impact on consumer attitudes. In addition, how the target's response affects consumer attitudes requires investigation. Academic research in general, however, and empirical investigations in particular, regarding consumer's perceptions of negative publicity episodes are scarce. Similarly, little attention has been focused on the strategic importance of the target firm's response to such incidents.

One notable exception is the study by Tybout, Calder and Sternthal (1981) investigating the application of information processing to the design of marketing strategies. This research provided the essential first step in the examination of negative publicity. Tybout et al. used an experimental setting to evaluate the effectiveness of different response strategies to a negative publicity rumor. The results suggest the traditional strategy of directly refuting the rumor was not effective. However, this study did not include manipulation of any situational variables, nor investigate situations where negative publicity resulted from an actual incident rather than rumor. As a future research issue, Scott and Tybout (1981) called for the investigation of "situational qualifiers" (p. 408) which influence the impact of negative information and are therefore of particular relevance to managers.

This study attempts to help fill the informational void regarding negative publicity situational qualifiers. First, the impact that three characteristics have on consumer attitudes will be measured. The three dimensions, previously posited as classification dimensions of negative publicity (Sherrell and Reidenbach 1986), are locus of responsibility, credibility of the reporting source, and the target's performance history. Second, the influence of the target's potential response strategies will be assessed for their influence on consumer attitudes. Third, respondent suggestions for response strategies will be analyzed and discussed.

BACKGROUND AND HYPOTHESES

A study was designed to provide information on the impact of the experimental manipulations on the consumer's general attitude toward the target firm and future purchase intentions. Subjects' ability to make causal attributions provides the conceptual framework for the hypothesized relationships.

Locus of Responsibility

Locus of responsibility has a theoretical foundation in attribution theory. Specifically, the "discounter principle" provides insight into the hypothesized impact of internal and external locus of responsibility. Kelley stated "The role of a given cause in producing a given effect is discounted if other plausible causes are present" (1973, p. 113). In negative publicity situations lacking plausible alternative causes, the consumer is likely to make a stronger firm-related attribution, placing the blame on the target company and lowering the consumer's attitude toward the firm. Conversely, if alternative plausible causes are present the firm attribution will be less strong and not affect the subject's attitude as severely.

Previous empirical work partially supports the theoretical propositions. Mowen, Jolly and Nickell (1981) used regression analysis to determine the situational variables influencing consumer response to recalled products. One of the variables studied was the company's responsibility for the defect. In one of four recall cases responsibility for the defect did have a significant negative influence on consumers' impression of the company. However, in the other three situations no significant relationship was discovered. In another study regarding consumer reactions to product failure, FoLkes (1984) found locus to be a significant predictor of both the consumer's future expectations and type of redress preferred. In summary, there exists intuitive, theoretical and limited empirical support for:

H1: External locus of responsibility will result in more positive attitude ratings toward the firm and greater intention to purchase.

Source Credibility

Marketers have long assumed that a more credible source is more influential (see McGuire 1968; Sternthal, Phillips and Dholakia 1978). Conceptual support for this assumption is provided by traditional attitude theory and attribution theory. Attitude theory suggests that attitude change occurs when an individual is exposed to a persuasive message which alters the individual's beliefs. Source credibility is a central variable to the formulation of such a message.

In attribution theory, the receiver of a message attempts to determine the reason for that message. The credibility of the source is one factor affecting the receiver's inference (Eagly, Wood and Chaiken 1978). A highly credible source draws the receiver into making a stronger attribution regarding the message, yielding a more significant impact on attitudes. Conversely, a source low in credibility may be "discounted" and result in little attitude change. Richins (1984) suggests that positive information is expected and has little impact, while negative information is unexpected and is more likely to be believed. Richins' position is consistent with the literature regarding the impact of negative information (Kanouse and Hanson 1972; Mizerski 1982; Richey et al. 1975; Weinberger and Dillon 1980).

Empirical research on the impact of source credibility in negative information scenarios is contradictory. As an example, Weiner and Mowen (1986) investigated the influence of source expertise and trustworthiness on consumer attitude formation. They found that in situations where the source was expected to be biased, the subjects did indeed discount the message. (For other examples see Dholakia and Sternthal 1977; Sternthal, Phillips and Dholakia 1978).

Other research has resulted in opposing conclusions. Wegner, Wenzlaff, Kerker and Beattie (1981) manipulated media source credibility to detect the influence on innuendo. The results show that regardless of the source (New York Times versus National Enquirer) the innuendo was equally effective. Similarly, Rosenbaum and Levin (1969) discovered credibility of source does not influence the impact of negative information as it does positive information. In spite of the conflicting empirical findings, the preponderance of theoretical and empirical literature supports:

H2: Less credible sources will result in more positive attitude ratings toward the firm and greater intention to purchase.

Performance History

When attributors are exposed to multiple observations, substantially more information is available to apply in an individual's causal attribution process. These circumstances give rise to Kelley's (1973) "principles of covariance". Kelley suggests that, given multiple observations of the same effect, an "effect is attributed to the one of its possible causes with which, over time it covaries" (1973, p. 108). In a negative publicity setting, a firm with multiple occurrences of a problem is more likely to be perceived as the cause of that problem. Conversely, if the episode is an isolated incident, the consumer's ability to attribute the responsibility solely to the firm is less powerful.

Empirical evidence regarding the impact of performance history has shown conflicting results. In one study the number of previous product recalls by a manufacturer was found to influence consumer perceptions of the firm and replacement products (Mowen 1979). Similarly, Folkes (1984), using Weiner's (1980) attribution categories to study product failure, found the stability dimension significantly related to consumer expectations and preferred redress. Two other studies, however, failed to reveal a significant relationship between comparable previous problems and consumer perception of the firm (Mowen 1980; Mowen, Jolly and Nickell 1981). Again, the intuitive logic and theoretical foundation, in light of inconclusive empirical evidence, led us to hypothesize:

H3: Target firms with no history of product failure will receive more positive attitude ratings and higher intention to purchase.

Response Tactic

The target responses used in the study are similar to those investigated by Tybout, Calder and Sternthal (1981). These researchers discovered, contrary to previous attitude theory and intuition, simply refuting a negative rumor was an ineffective response tactic. Based on information processing theory, an explanation for the failure of the denial strategy was offered. In brief, consumers' attitudes are affected simply by processing information they receive, even if they do not actually believe it. As plausible alternative response tactics, Tybout, Calder and Sternthal (1981, p. 74) propose utilizing a "storage strategy", whereby a second positive piece of information is interjected when the negative information is being stored, or a "retrieval strategy", through which additional stimuli are introduced as a means of diluting the association between rumor and target. Basically, these strategies involve providing additional informational cues which associate the firm with positive information or inhibit the retrieval of the negative information. Both strategies avoid repeating the rumor which is necessary in a denial strategy. Results of the Tybout et al. study supported the use of both storage and retrieval strategies.

Based on the work of Tybout, Calder and Sternthal, we believe that consumers exposed to a denial of the incident will react strongly to the scenario, reporting lower attitude ratings and purchase intentions than consumers receiving no response from the firm. Subjects receiving a redress response, a form of storage strategy, will report more positive attitudes and purchase intentions than those in the no response category. Thus, we propose:

H4: A denial response will result in the lowest attitude and purchase intention ratings, no response a neutral rating, and a redress response the most favorable attitude and purchase intention ratings.

METHODOLOGY

Subjects

One hundred and ninety-six students attending a major state university voluntarily participated in the study. The subjects were recruited from those enrolled in various undergraduate business classes and were assigned randomly to one of the treatment groups in a 2 X 2 X 2 X 3 full factorial experimental design. After editing, 188 subjects were retained for analysis.

Experimental Context

Prior to developing the scenarios and designing the questionnaires, an appropriate context had to be determined. The context had to be relevant to the experimental subjects, but yet not so personal as to limit the range of attitudes. After reviewing several possible options, the fast food industry was selected as an appropriate experimental context. Fast food franchises have experienced their share of negative publicity (i.e. McDonalds massacre; reports of ground worms in hamburger meat; kangaroo meat passed off as beef) and hold high personal relevance to the test subjects. In addition, the subjects (college-age consumers) represent an important market segment to the fast food industry. One of the recent negative publicity episodes focused on the presence of salmonella bacteria in poultry products. Applying this to the fast food industry was simply a matter of reporting that a restaurant had allegedly sold chicken that resulted in several cases of salmonella poisoning.

Scenario Development

Scenarios were constructed around the fried chicken franchise/salmonella poisoning context. The three situational manipulations were developed in all possible combinations and then matched with three response strategies. The scenarios are as similar as possible in all other respects.

Locus of responsibility was manipulated by reporting that salmonella poisoning could result from (1) improper storage and preparation of the chicken or by (2) failure to refrigerate the chicken following preparation. Under the internal (first-responsible) locus condition, it was reported that high levels of bacteria were detected in chicken tested on the premises of the target firm. The franchise was then charged with failure to properly store or prepare the product. The external locus condition reported no bacteria was found in the chicken at the store and the poisoning was attributed to the consumer's failure to refrigerate leftover chicken.

The credibility manipulation was operationalized by either (1) "city public health officials" or (2) "unconfirmed reports" as the source of the information. The high credibility construct was written in a manner suggesting that the scenario was fact, whereas the low credibility suggested that the report was an unsubstantiated rumor.

Historical performance was defined as (1) the presence or (2) an absence of similar incidents having occurred to the target firm. The manipulation was operationalized in the "isolated" scenario by reporting the poisoning was "an isolated incident"; in the "chronic" scenario by stating "similar incidents have been reported in other cities".

Target firm responses fell into three categories: (1) no response from the firm, (2) a denial response, and (3) a redress response. In the denial response tactic the target firm refused to accept any responsibility for the incident, to the point of implying that the reported sicknesses might not have even occurred. Under the redress response tactic the target firm made a concerted effort to remedy any potential problems and provided informational brochures to consumers regarding the safe handling of poultry products.

Operational Measures

Scales were developed to measure subjects' attitudes toward the chicken franchise's locations, promotional efforts and an overall evaluation. In addition, an intent to purchase measure was developed. The measures for franchise locations and promotional efforts were included as checks for demand artifacts. The overall evaluation and purchase intention measures were utilized as dependent variables in subsequent analysis.

The attitude measures consisted of seven point semantic differential scales using polar terms germane to the individual construct. The polar adjectives were adopted from previous research (Osgood, Suci, and Tannenbaum 1957) and have been found to display high levels of reliability. An effort was made to select terms relevant to the construct being measured as well as covering the three dimensions of attitude - activity, evaluation, and potency. Intent to purchase was assessed directly, but phrased in a context suited to the sample. After purification in pretesting, the scales utilized in the present study displayed coefficient alpha estimates ranging from 0.78 to 0.90.

Procedure

The experiment was administered in classrooms on the university campus. The subjects were introduced to a research assistant who provided general information regarding the task they were about to perform, but not the particular issue under investigation. The subjects were instructed to read the questions and information carefully, proceed through the test instrument in the sequence in which it was presented and to answer each question as accurately as possible at that time.

Questionnaires were constructed to gather the subject's attitude assessments and intent to purchase at two different points - before and after exposure to the experimental stimuli. The pretest measures used for analysis were embedded in a number of questions regarding fast food franchises. All subjects initially responded to identical instruments. Each subject then read a one page negative publicity/firm response scenario determined by experimental group membership. Following the suggestion of Wetzel (1977, p. 89), subjects responded to the dependents measures directly after exposure to the experimental scenarios. After completion of the attitude measures, suggestions of what the firm should have done to address the incident were solicited.

After completion of testing, the subjects were debriefed. Of primary concern was insuring that the subjects recognized the hypothetical nature of the scenarios and that the particular franchise selected was utilized solely because it was a well known franchise capable of providing a base of reference for the study. To accomplish this, both a written disclaimer and a verbal explanation were provided.

Manipulation/Confounding Checks

In experimental designs, subjects are exposed to various conditions and expected to react differently to those conditions. To insure the internal validity of this experiment, manipulations were tested and revised during pretesting (Wetzel 1977). In addition, as a strong test of manipulations, checks were included as the last questions in the main experiment (Perdue and Summers 1986, p. 320). The measures asked the respondents if: (1) the target firm was solely responsible for the incident (locus of responsibility); (2) the incident was reported by a reliable source (source credibility); and (3) the target firm had experienced similar problems (historical performance).

To examine the manipulations the subjects were classified into high and low groups for each of the manipulations and then tested for a difference in the mean response on the three check questions. With this approach, both the effectiveness of the manipulations and any potential confounding effects between manipulations can be tested (Wetzel 1977, p.88). In each case, a significant difference was found between mean scores on the appropriate measure, but not the others, suggesting that the manipulations were effective and independent. The results are presented in Table 1.

A second concern in consumer research is the impact of demand artifacts on the validity of the experiment (Sawyer 1975). In the present study, the possibility of subjects attempting to respond in a "socially desirable" manner is particularly relevant. While no fool proof method exists for detecting or eliminating demand artifacts, two specific steps were taken to minimize the occurrence in the study. First, relevant measures taken before exposure to the experimental stimuli were disguised by their inclusion with several distracter items. Second, two firm-related measures were included that were not expected to be changed by the manipulations in an attempt to detect yea saying. Analysis of these measures, reported in Table 2, suggests demand artifacts are not a major problem. The only significant effect found in this analysis pertains to the impact of response type on attitude toward a firm's promotion (p=.03). Although this could be attributed to a demand effect, it is quite plausible that this effect is due to subject perceptions of response style as part of a firm's promotional mix.

ANALYSIS AND RESULTS

Given the nature of the data, in particular that the desired subject responses were measured both before and after exposure to the experimental scenario, the general form of the model required to test the research hypotheses is given by Neter, Wasserman and Kutner (1985) as: [Interaction terms are omitted for simplicity in the equation, but were, of course, included in the analysis.]

Yijklmn = m + ai + bj + dk + pl + sn + eijklmn (1)

where 1 represents the grand mean, ai represents the ith level of locus, bj the jth level of history, dk the kth level of credibility, P1 the 1th level of response, and Sn a blocking (within-subject) factor accounting for the two responses on each subject. The results of this analysis are shown in Table 3.

TABLE 1

RESULTS OF MANIPULATION CHECKS

TABLE 2

RESULTS OF ANALYSIS OF VARIANCE FOR DEMAND ARTIFACTS

As shown in Table 3, the partial F statistics for the within subject factor are highly significant for both response measures (attitude and purchase intention). This result supports the ability of the experimental scenario to affect the dependent variables. In terms of the present study, this result, along with analysis of the group means, is interpreted as evidence of a tendency among respondents to lower their attitudes and purchase intentions toward the target firm. Furthermore, this effect holds regardless of the combination of experimental treatments they were exposed to.

In addition, Table 3 shows partial F values for each experimental treatment while controlling for both the within-subject factor and all possible two and three-way interactions. Since no interaction effects were hypothesized or found significant at a p-level of .10, they were omitted from the table. Each hypothesis was examined by analyzing these partial F values.

TABLE 3

RESULTS OF ANALYSIS OF VARIANCE FOR DEPENDENT MEASURES

Locus of Responsibility. H1 predicted that both attitudes and purchase intentions toward the firm would be relatively lower when the locus of responsibility was directed toward the store rather than the customer. The partial F values shown in Table 3 support this hypothesis for the attitude measure (p < .011) while showing only marginal support when using intentions as a response (p < .075).

Source Credibility. H2 predicted lower source credibility would be associated with relatively higher attitudes and purchase intentions. The partial F values for this effect support H2 for attitudes (p < .0001), but fail to support it using intentions as a dependent measure (p < .296).

Performance History. H3 predicted that attitudes and intentions would be significantly effected when a firm is portrayed as having chronic problems with product failures as opposed to having no such previous record. Again, the F-value supports the hypothesis with regard to lower attitudes under conditions of chronic product failures (p < .02); however, no significant effect is shown on purchase intentions (p < .524).

Response Tactic. H4 hypothesized attitudes and purchase intentions would vary according to which response level a subject was exposed (no response, redress, denial). Again, the partial F values support this hypothesis for the attitude measure (p < .027) while failing to support it for purchase intentions (p < .236). Analysis of the cell means indicates the significant effect observed on respondent attitudes was due mainly to higher mean scores on the redress response category compared to both denial and no response conditions, which displayed virtually identical scores.

In sum, each of the research hypotheses were clearly supported with respect to the impact of experimental factors on attitudes toward a firm suffering from exposure to negative public publicity. The picture is not as clear with regard to the impact of locus, credibility, history, and response on purchase intentions. While only the hypothesized effect of locus on intentions reached even a marginal level of significance, a comparison of cell means of purchase intentions showed that all observed effects were in the hypothesized direction. This is consistent with a significant reduction in overall purchase intention scores between the pre-exposure and post-exposure conditions mentioned above (within-subjects). However, despite evidence demonstrating lower purchase intentions based on exposure to negative public publicity, the individual effects of each experimental factor on intentions cannot be confidently supported based on these data.

Qualitative Data. Respondents were given an opportunity to comment on the incident reported after completing the questionnaire. The majority of subjects offered suggestions on how the target firm should have handled the negative publicity episode.

Honesty was the most common term utilized by subjects who read the denial and no response scenarios. Over half of the respondents in these conditions (56%) included some suggestion that the target deal with the situation honestly. In general, subjects wanted the target firm to face the problem rather than avoid it, and several suggested that an independent agency should investigate the incident. Offering information to consumers on proper handling of poultry products was also suggested by eight respondents. In addition, six subjects wanted punitive action taken against the firm or employee responsible (five subjects recommended publicly announcing that an employee of the firm was fired over the incident). Only two respondents stated the incident was handled properly by a denial tactic.

Subjects exposed to the redress response tactic displayed overall satisfaction with the target's actions. Many respondents (63%) stated the firm had "handled the situation well". In particular, subjects cited the target's willingness to provide informational brochures as a very positive action. No mention was made of punitive action, suggesting a more positive approach to the incident satisfies consumers' "call for blood" to rectify the problem.

DISCUSSION AND CONCLUSIONS

The results of this experiment indicate source credibility and firm responsibility, history, and response tactic are important situational factors leading to changes in consumer attitudes based on a negative publicity episode. In addition, while purchase intentions were significantly lowered by experimental stimuli, individual relationships hypothesized between situational factors and purchase intentions were not confirmed.

Although not hypothesized, a significant interaction among situational variables would not have been surprising. In particular, a locus by history interaction would correspond to the locus by stability interaction found in Folkes' (1984) study of consumer reactions to product failure. The differing results between Folkes' work and this study may be attributed to slightly different dependent measures, experimental settings and operationalization of manipulations. The main effects, however, do further corroborate findings reported by Folkes.

From a pragmatic viewpoint, it appears target firms have little to lose by taking a proactive approach toward negative publicity. One could argue denial of a rumor might have either a positive effect by persuading consumers that it is indeed untrue, or a negative impact be serving to legitimatize the charges. The results of the study, however, indicate denial produces similar results to ignoring the event entirely. On the other hand, a redress approach to negative publicity appears to help restore firm image regardless of the situational factors. An admission of responsibility may not be necessary, but taking a proactive consumer oriented approach pays dividends.

While certainly no surprise, the significance of the history manipulation reinforces the importance of maintaining a strong corporate image. Knowledge of previous problems or a poor reputation will negatively influence consumer reactions to current incidents. Therefore confronting today's negative publicity in a positive and timely manner should not only benefit accused firms immediately, but also provide insulation against whatever tomorrow might bring.

This area of academic research holds substantial future promise. While this study is somewhat limited by the wide arrange of phenomena addressed, it does provide an impetus for further, and more in-depth, investigations of each of the effects operationalized through experimental manipulations. Although further theoretical development and empirical verification are needed, the present study illustrates these manipulations are capable of significantly altering consumers' attitude ratings. The findings compliment and support the earlier work of Tybout, Calder and Sternthal (1981). Additional experimental studies with larger and more generalizable samples would also be valuable. In addition, a "vulture approach" (laying in wait for a negative publicity episode to occur) seems useful, if not efficient, approach to conducting field studies of negative publicity which may provide effects applications of relationships explored here.

REFERENCES

Cusumano, Donald R. and Marjorie H. Richey (1970)> "A Negative Salience in Impressions of Character: Effects of Extremeness of Stimulus Information," Psychometric Science, 20, 81-83.

Dholakia, Ruby and Brian Sternthal (1977), "Highly Credible Sources: Persuasive Facilitators or Persuasive Liabilities?," Journal of Consumer Research, 3, 223-232.

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.

Folkes, Valerie S. (1984), "Consumer Reactions to Product Failure: An Attributional Approach," Journal of Consumer Research, 10, 398409.

Kanouse, David E. and L. Reid Hanson Jr. (1972), "Negativity in Evaluations," in Attribution: Perceiving the Causes of Behavior, Edward E. Jones et al., editors, Morristown, NJ: General Learning Press, 47-62.

Kelley, Harold (1973), "Processes of Causal Attribution," American Psychologist, 28, 107-128.

McGuire, W. J. (1968), "The Nature of Attitudes and Attitude Change," in G. Linzey and E. Aronson, eds., Handbook of Social Psychology, Reading, MA: Addison-Wesley.

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

Mowen, John C. (1979), "Consumer Reactions to Product Recalls: An Empirical and Theoretical Examination," unpublished working paper, Department of Marketing, Oklahoma State University.

Mowen, John C. (1980), "Further Information on Consumer Perceptions of Product Recalls," Advances in Consumer Research, Jerry C. Olson, ed., Ann Arbor: Association for Consumer Research, 7, 519-523.

Mowen, John C., David Jolly and Gary S. Nickell (1981), "Factors Influencing Consumer Responses to Product Recalls: A Regression Analysis Approach," Advances in Consumer Research, Kent B. Monroe, ed., Ann Arbor: Association for Consumer Research, 8, 405-407.

Neter, John, William Wasserman and Michael H. Kutner (1985), Applied Linear Statistical Models, Homewood, IL: Richard D. Irwin, Inc.

Osgood, Charles E., George J. Suci and Percy H. Tannenbaum (1957), The Measurement of Meaning, Champaign, IL: University of Illinois Press.

Perdue, Barbara C. and John O. Summers (1986), "Checking the Success of Manipulations in Marketing Experiments," Journal of Marketing Research, 23, 317-326.

Richey, Marjorie H., Robert J. Koenigs, Harold W. Richey and Richard Fortin (1975), "Negative Salience in-Impressions of Character: Effects of Unequal Proportions of Positive and Negative Information," Journal of Social Psychology, 97, 233-241.

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

Richins, Marsha L. (1984), "Word of Mouth Communication as Negative Information," Advances in Consumer Research, Thomas C. Kinnear, ed., Provo UT: Association for Consumer Research, 11, 697702.

Rosenbaum, Milton E. and Irwin P. Levin (1969), "Impression Formation as a Function of Source Credibility and the Polarity of Information," Journal of Personality and Social Psychology, 12, 34-37.

Sawyer, Alan G. (1975), "Demand Artifacts in Laboratory Experiments in Consumer Research," Journal of Consumer Research, 1, 20-30.

Scott, Carol A. and Alice M. Tybout (1981), "Theoretical Perspectives on the Impact of Negative Information in Impression Formation: Does Valence Matter?," Advances in Consumer Research, Kent B. Monroe, ed., Ann Arbor: Association for Consumer Research, 8, 408409.

Sherrell, Daniel L. and R. Eric Reidenbach (1986), "A Consumer Response Framework for Negative Publicity: Suggestions for Response Strategies," Akron Business and Economic Review, 17, 3744.

Sternthal, Brian, Lynne Phillips and Ruby Dholakia (1978), 'The Persuasive Effects of Source Credibility: A Situational Analysis," Public Opinion Quarterly, 285-314.

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

Wegner, D. M., R. Wenzlaff, R. M. Kerker and A. E. Beattie (1981), Incrimination Through Innuendo: Can Media Questions Become Public Answers?," Journal of Personality and Social Psychology, 40, 822-832.

Wetzel, Christopher G. (1977), "Manipulation Checks: A Reply to Kidd," Representative Research in Social Psychology, 8, 88-93.

Weinberger, Marc C. and William R. Dillon (1980), "The Effects of Unfavorable Product Rating Information," in Advances in Consumer Research, Jerry C. Olson, ed., Ann Arbor: Association for Consumer Research, 7, 528-532.

Weiner, Bernard (1980), Human Motivation, New York: Holt, Rinehart & Winston.

Weiner, Joshua L. and John C. Mowen (1986), "Source Credibility: On the Independent Effects of Trust and Expertise," Advances in Consumer Research, Richard J. Lutz, ed., Provo UT: Association for Consumer Research, 13, 306-10.

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