The Impact of Negative and Positive Information on Source Credibility in a Field Setting

ABSTRACT - This paper examines the impact of positive and negative source information on the credibility of the advertising source and on audience perceptions of the company and advocated product. A field research study of Lee Iacocca, CEO of Chrysler Corporation indicates that the cognitive and affective dimensions of credibility respond differently to positive and negative information and that these dimensions relate differently to audience perceptions of product quality and safety.

Citation:

Joanne M. Klebba and Lynette S. Unger (1983) ,"The Impact of Negative and Positive Information on Source Credibility in a Field Setting", in NA - Advances in Consumer Research Volume 10, eds. Richard P. Bagozzi and Alice M. Tybout, Ann Abor, MI : Association for Consumer Research, Pages: 11-16.

Advances in Consumer Research Volume 10, 1983      Pages 11-16

THE IMPACT OF NEGATIVE AND POSITIVE INFORMATION ON SOURCE CREDIBILITY IN A FIELD SETTING

Joanne M. Klebba, University of Cincinnati

Lynette S. Unger, Miami University

ABSTRACT -

This paper examines the impact of positive and negative source information on the credibility of the advertising source and on audience perceptions of the company and advocated product. A field research study of Lee Iacocca, CEO of Chrysler Corporation indicates that the cognitive and affective dimensions of credibility respond differently to positive and negative information and that these dimensions relate differently to audience perceptions of product quality and safety.

INTRODUCTION

The use of the celebrity as product advocate is recognized as an effective advertising technique. Published evidence indicates that testimonials enhance readership scores (Freeman 1957, Rudolph 1947), increase awareness and induce positive attitude change toward a company and its product (Fireworker and Friedman 1977, Kamen, Azhari and Kragh 1975). In addition, source credibility investigations generally conclude that the more credible a source, the more persuasive s/he is (Sternthal, Philips and Dholakia 1978).

The widespread use of celebrity endorsers and the tees they command (sometimes in excess of $1 million annually) attest indirectly to their success as product spokespersons. An estimated 33% of all television commercials contained celebrities in 1976 (Business Week 1978). Yet advertisers who use celebrity endorsers may encounter three major problems. First, lured by the generous compensation, many celebrities opt to advocate several products, a practice which can result in overexposure and a muddied image. In a study by Mowen and Brown (1981), results indicated that subjects saw the product and advertisement more favorably and demonstrated greater purchase interest when a celebrity endorsed one rather than several products. A second problem is the question of matching the product type with the appropriate celebrity. Kanungo and Pang (1973) investigated the effects of advertising models' sex and found that "fittingness" of the model is a determinant of perceived product quality. Friedman and Friedman (1979) found that endorser type effectiveness is dependent on product type. Because of the dramatic growth in the amount of negative information in the marketplace (Weinburger, Allen and Dillon 1981), a third problem is of critical interest. Marketers must now be concerned with negative information about their product, introduced by regulatory agencies, consumer groups, competitors or rumor. Moreover, advertisers using celebrities as advocates must also be concerned with these endorsers' public reputations as formed by the press, special interest groups or public investigations.

The purpose of this study is to examine the impact or negative and positive source information on credibility and to explore the effect of such information on attitudes toward the advocated product. Negative and positive information constitute the independent variables in this study and the components of source credibility comprise the dependent variables.

BACKGROUND

The use of a company's CEO as a celebrity testimonial represents a growing variation in the celebrity endorsement technique (Fortune 1980). Mr. Lee Iacocca, Chairman and Chief Executive Officer and advertising spokesperson for Chrysler corporation is the focus of this study. Prior to Iacocca's Chrysler experience, he was also President of the Ford Motor Company. During his tenure with Ford, Iacocca was involved in the development of the Pinto and the Mustang automobiles.

In the fall of 1979; Mr. Iacocca began to appear as product spokesman in Chrysler Corporation's automobile advertising. In the spring of 1980, however, Iacocca's involvement in the development of the Ford Pinto was brought out in a liability suit based on design inadequacies in the Pinto. The suit was brought against Ford Motor Company. Although Iacocca was no longer associated with Ford or the Pinto, his involvement with the Pinto was nationally publicized during the liability trial (Advertising Age, 1981). Iacocca's involvement with the Pinto is utilized as the negative information in this study. Iacocca's earlier association with the highly successful Mustang constitutes the positive information for the study.

SUMMARY OF PREVIOUS RESEARCH

The research which deals with negative information covers two areas: the effect of negative cues about products and about persons. Researchers generally conclude that negative product information has a greater impact than positive information on attitudes toward products. Wright (1974) found that negative information affected car purchase intent more than positive cues under high time pressure conditions. Others found negative information bias in formation or attitudes toward products (Lutz 1975; Weinburger and Dillon 1980). In the area of corrective advertising and affirmative disclosure, results are mixed. Some researchers (Hunt 1977) have found that introduction of negative information in the form of corrective copy can dissipate the effects of deceptive advertising. Findings of other researchers indicate negative product information is not necessarily detrimental to product attitudes and may even enhance credibility (Dyer and Kuehl 1974; Settle and Golden 1974).

In their literature review, Weinburger, Allen and Dillon (1981) explain negative information bias using an attribution theory framework. "Negative information has the strong impact that it does because it stands out more than positive information and...would therefore have more distinctiveness. This apparently results from the fact that there are more positive cues in the individual's social environment. As a result, negative cues attract more attention and are therefore more heavily attributable to the stimulus object" (p. 398). Scott and Tybout (1981) suggested that cue valence alone does not cause differential attention to negative information and that it should be studied within the broader framework of information processing.

The psychological literature provides ample research on "person perception," the formation of impressions about people based on negative and positive information. While at least one researcher (Infante 1980) has cautioned against confusing person perception with source credibility, a review of this literature is relevant to the study of implications of negative source information. As in the case of products, researchers in the person perception area also acknowledge a negative information bias. Richey et al. (1975) found that a single report of a person's negative behavior outweighed five positive behaviors. In other studies, Richey and her colleagues introduced both positive and negative information, controlling for intensity and order effect, and found disproportionate influence of negative information in person perception formation (Richey and Dwyer 1970; Richey, McClelland and Shimkunas 1967). In reviewing the personnel literature, Weinburger, Allen and Dillon (1981) also found studies to indicate that negative information in job interviews is more salient than positive information in forming impressions of employee candidates.

Two studies are centered on the effects of negative information on particular dimensions of source credibility. Sereno and Hawkins (1967) studied the influence of speaker nonfluency, a kind of negative information, on the source credibility dimensions defined by Berlo, Lemert and Mertz (1969). They found that nonfluency adversely affected the competence and dynamism factors but not the trustworthiness factor. In another study, Richey, McClelland and Shimkunas (1967) found negative bias in perceptions of persons' moral-ethical character. While the study did not focus on source credibility, this particular area of personality might be equated with trustworthiness, an acknowledged dimension of credibility.

The vast body of research on source credibility might be divided into two areas: studies concerned with the impact of credibility on social influence and research on the underlying dimensions of credibility (Sternthal, Philips and Dholakia 1978). The literature in the first area generally concludes that the more credible the source, the more persuasive he or she is (Delozier 1976; Fireworker and Friedman 1977- Hovland and Weiss 1951; Kelman and Hovland 1953; McGuire 1969). This general conclusion has been borne out in main effects studies; however, studies of interactions indicate that low source credibility is not always detrimental nor is high credibility always an asset (Dholakia and Sternthal 1977). The emphasis here, however, is on the main effect, enhancing source credibility to increase the persuasiveness of advertising messages. An extensive review of the literature on source persuasiveness is provided by Sternthal, Philips and Dholakia (1978).

The positive relationship between source credibility and persuasiveness has often been demonstrated by manipulating the dimensions of credibility, the second area of source credibility research. The literature most often recognizes three dimensions: trustworthiness, expertise and likability (Kelman and Hovland 1953). Researchers have proposed other classifications of the components of credibility as well. Rarick (1963) found evidence for two dimensions, cognitive and affective. Included in the cognitive component are power, prestige and competence. The affective component is comprised of characteristics such as trustworthiness and likability. Berlo, Lemert and Mertz (1969) include safety (trustworthiness, friendliness), qualification (competence, experience), and dynamism (energy, aggressiveness) in their classification system. The similarity and overlap of these classification systems is clear. The trustworthy, likeable and expert components appear in some form in all of the classifications.

The important point is that source credibility depends on several traits and the receiver's perception of these traits. More detailed discussion of source credibility research is contained in Delozier (1976).

HYPOTHESES

This study uses an after-only research design to investigate the impact of positive and negative source information (the independent variables) on source credibility (the dependent variable).

The following hypotheses were tested:

H1: Subjects who possess negative information will rate the source lower on all dimensions of credibility than subjects who do not possess negative information.

H2: Subjects who possess only positive information will rate the source higher on all dimensions of credibility than subjects who possess only negative information, both positive and negative information or neither positive or negative information.

H3: All dimensions of credibility will be positively related to perceptions of the Chrysler Corporation and its products.

METHODOLOGY

Preliminary Study

A convenience sample of 90 undergraduate business students responded to a preliminary questionnaire. The purpose of this questionnaire was to confirm that respondents perceived the Mustang positively and the Pinto negatively relative to other automobiles and that they perceived the Mustang more favorably than the Pinto on certain attributes. Two measures of positive-negative information were utilized. The participants were asked to indicate their general reaction to ten automobiles produced in the United States, including the Mustang and Pinto, on a 5-point Likert scale ranging from "Very positive" to "Very negative." A list of ten automobiles was used to disguise the interest in the Pinto and the Mustang. Following this task participants were-asked to respond to seven attributes on semantic differential scales regarding the Pinto and the Mustang. A list of the attributes is contained in Table 1. Respondents also provided demographic information which was used to test for similarity between preliminary and main study respondents. There were no significant demographic differences between these groups. T-tests were conducted on the degree of positiveness/negativeness and on the semantic differential items to determine if perceptions of the Mustang characteristics were higher than those of the Pinto characteristics.

The results from the preliminary study confirmed that the Mustang was viewed more positively and favorably than the Pinto. The t-test which compared the Pinto and Mustang automobile ratings indicated a significant difference in positive/negative ratings (p=.00). On the five-point scale where 5 is "Very positive", the mean rating for the Mustang was 3.87 (s.d.=.91) and the mean for the Pinto was 1.82 (s.d.=.87). T-tests conducted on responses to the bipolar attribute scales showed that the Mustang was perceived significantly better on all seven attributes (p=.00), as shown in Table 1.

Main Study

A convenience sample of 245 undergraduate (797@) and graduate (21%) business students responded to a written questionnaire during in-class administrations over a two week period in May, 1980. All questionnaires were usable. Some 98% of respondents indicated they were drivers, and 7 9% said they planned to purchase a car within 0 years. While this sample does not represent the general auto-buying public, it represents a significant proportion of the market for the cars under study. According to Simmons (1979), 18-25 year olds bought 20% of Pintos in 1979 and 19% of Mustangs in 1979. They made up 192 of new car buyers in 1979.

TABLE 1

ATTRIBUTE DIFFERENCES ON SEMANTIC DIFFERENTIAL

The questionnaire contained ten source statements about Lee Iacocca and nine product statements about the Chrysler Corporation and its products. Respondents were to respond to these nineteen statements on a five-point Likert scale ("Strongly agree" to "Strongly disagree"). The source statements were written to include all components of source credibility suggested in the literature. The product statements incorporated attributes featured in recent advertising. The nineteen statements are shown in Table 9. The source statements were factored using principal components analysis with varimax rotation in order to isolate dimensions Or source credibility. Similarly, the product statements were factored to yield dimensions on product perceptions. The hypotheses were tested using these factor scores for the dependent (Hypotheses 1 and 2) and independent variables (Hypothesis 3).

Subjects were also asked to respond to two multiple-choice questions concerning their knowledge or Iacocca and his affiliation with Chrysler and Ford. The first question asked if they knew of his affiliation with Ford during the development of the Pinto. The second determined their knowledge or his affiliation with Ford during the development or the Mustang. Both questions contained a "don't know" option to discourage guessing.

To test the first hypothesis, subjects were classified into two groups according to their responses on the first multiple choice question. If they indicated correct knowledge or Iacocca's association with the Pinto, they were considered to possess negative information (25% of sample). Respondents selecting incorrect or "don't know" options were grouped as not having negative information (75% of sample). It is important to note that respondents were assigned to these groups based on their exposure to negative information rather than on evidence of negative beliefs. The preliminary study results, which suggest that respondents see the Pinto more negatively than the Mustang, provide only indirect evidence that knowledge of Iacocca's association with the Pinto constitutes" negative belief. T-tests for mean differences between the two groups were conducted using factor scores from the source item factor analysis as dependent variables. If the subjects who had negative information had significantly lower factor scores than chose who did not, Hypothesis 1 would be supported.

TABLE 2

FACTOR ANALYSIS RESULTS ON SOURCE AND PRODUCT ITEMS

The second hypothesis was tested using analysis or variance. Subjects were placed in four groups according to their responses on the multiple-choice questions, these questions identified four degrees of information about the source: positive information only, negative information only, both positive and negative information and neither positive nor negative information. The information criteria for all groups are summarized in Table 3. The dependent variables were again the factor scores from the source item factor analysis. The second hypothesis would be supported if the respondents having only positive information had the highest factor scores.

Multiple regression analyses were used to test the third hypothesis. The factor scores from the source item factor analysis were used as predictors, and the factor scores from the product item factor analysis were the dependent variables. Significant and positive beta coefficients in all regressions would support this hypothesis.

TABLE 3

CRITERIA FOR GROUP DESIGNATION

MAIN STUDY FINDINGS

Responses on the ten source credibility items and the nine Chrysler product items were subjected to principal components analysis with varimax rotation. On the source item analysis, two factors emerged with eigenvalues greater than one. These dimensions, shown in Table 2, explained 61% of item variation. The first factor was labeled affective, as items loading high on this factor appeared to represent trustworthiness and likability. Items loading high on the second factor were interpreted to represent power and expertise, and this dimension was labeled cognitive. These results correspond with the finds of Rarick (1963), who identified affective and cognitive dimensions of source credibility. Three factors emerged from the product item analysis, accounting for 65.5% of variance. These dimensions were labeled quality, safety and company strategic ability as shown in Table 2.

Two t-tests were conducted using the affective and cognitive source factor scores as dependent variables to test Hypothesis 1. The results, shown in Table 4, directionally support the first hypothesis for the affective component, as respondents who knew of Iacocca's association with the Pinto rated him lower on this dimension. However, this difference was not statistically significant (p > .05). On the cognitive dimension the hypothesis was not supported. Those who knew of Iacocca's association with the Pinto rated him significantly higher (p = .09) on this dimension.

TABLE 4

RESULTS HYPOTHESIS 1 TESTING

The second hypothesis was only partially supported for both the affective and cognitive source credibility dimension scores. Analysis of variance revealed significant differences among the four information groups on the affective dimension of source credibility (F=9.902, df=3/226, p=. 04). A comparison of group means on affective factor scores indicates respondents with only positive knowledge tended to rate Iacocca highest in this dimension (X = .1833). Those who were not aware of his affiliation with either car had the second highest mean factor score X = .0033). Those who possessed negative information only were next (X = -.0095), and those who possessed both negative and positive information tended to rate him the lowest on this dimension (X = -.7282). Sequential range test results indicated only the Positive and Both groupings to be significantly different (P = .05).

ANOVA results using the cognitive factor scores as the criterion also indicated significant group differences (F = 12.475, df = 31226, p = .00). Order of means, however, was different than on the affective dimension. The Positive group rated Iacocca highest on this dimension as well with a mean factor score of .4904. Those aware of both affiliations rated him next highest (.3304). The Negative group mean was -.0528 and the Neither group averaged -.2859. Results from the sequential range test indicated the Positive group rating was significantly higher than the Neither and Negative group ratings (p = .01) and the Both group rating was also greater than the Neither (p = .01) and Negative (p =.05) group means. In contrast with the results from the affective analysis, however, the Positive group was not significantly higher than the Both group.

In testing the third hypothesis, three regression analyses were conducted using the three Chrysler product dimensions as criteria. The two source credibility factor scores were predictors in all three equations. The results are summarized in Table 5. Both dimensions of source credibility were found to be positive and significant in predicting perceptions of Chrysler quality, the first product dimension. The affective component beta was significant at p = .001 and the cognitive beta at p = .05. The coefficient of determination was .11. On the second Chrysler factor, safety, the beta for the affective factor score was significant (p=.001) and positive. The cognitive score beta was positive but not significant. The R for this regression was .10. Neither factor proved significantly related to the perceived strategic prowess of Chrysler, the third product dimension. In net, the affective dimension of source credibility was positively and significantly related to perceptions of product quality and safety, while the cognitive dimension was positively and significantly related only to perceived product quality.

TABLE 5

RESULTS HYPOTHESIS 3 TESTING

DISCUSSION

The results of this study indicate that the cognitive and affective dimensions of credibility are influenced differently by negative information. The tests of Hypotheses l and 2 suggest that perceptions of a source's affective characteristics (trustworthiness and likability) may be more influenced by the introduction of negative cues than perceptions of cognitive characteristics (power and expertise). In Hypothesis 1 testing, people who knew of Iacocca's affiliation with the Pinto rated him lower on the affective dimension; however, this difference was not statistically significant. By contrast, subjects who associated Iacocca with the Pinto rated him significantly higher on the cognitive dimension than those who did not. These results conflict with the findings of Sereno and Hawkins (1967), who found that a negative source cue adversely affected ratings on competence and dynamism (equivalent to the cognitive factor in this study) but did not affect trustworthiness (equivalent to the affective factor). Theoretical advancements by Scott and Tybout (1981) may in part explain these differences. They found that situational variables (source type and information format) may intervene in the processing of positive and negative cues. Specifically, in the Sereno and Hawkins study, subjects actually experienced the speaker's nonfluency (negative cue), while in this study, respondents had received negative and/or positive information externally, from the mass media. In that study the information was presented in concrete format, while the information about Iacocca was more abstract (business press coverage of the auto industry and the technical aspects of the Pinto case) and required more cognitive work. Future research should further investigate the impact of source type, information format and other situational variables. Why the two dimensions of credibility were affected differently in both studies is unclear, and might also be explored in future research.

The mixed results of Hypothesis 2 testing also underscore the question or importance of cue valence raised by Scott and Tybout (1981). In Hypothesis 2 testing, those who possessed only positive information rated Iacocca significantly higher on the affective dimension than those who possessed both positive and negative information. While the inflated impact of negative over positive information reported in the literature could not be isolated in this field study, it might be concluded that with respect to the affective component, negative information appears to at least neutralize positive information. The testing of Hypothesis v indicated there was no statistically significant difference in ratings on the cognitive dimension between those who had only positive information about Iacocca and those who had both positive and negative information. This suggests that possession of negative information did not neutralize positive information as it did on the affective dimension. Moreover, it indirectly refutes the negative bias generally found in the literature, at least with regard to evaluating the cognitive dimension of source credibility. As Scott and Tybout (1981) have noted, it is extremely difficult to test for differential weighting effect of cue valence in a controlled lab study and virtually impossible in a field study of this nature. Consequently, these findings should be interpreted cautiously and re-examined in a more controlled environment.

In terms of strategic implications, the findings indicate that when an audience is exposed to negative source information, this information may more strongly influence perceptions of the source's trustworthiness and likability than perceptions of his/her expertise or power. Moreover, this affective dimension of source credibility was directly related to two product evaluation dimensions, as shown in Hypothesis 3 testing These findings become more critical in view of the fact that researchers have found the affective dimension to be more strongly related to overall credibility than other dimensions (Hovland, Janis and Kelley 1953, Weiss and Fine 1956). Managers faced with this situation should consequently utilize proven strategies which strengthen perceptions of the source's affective qualities. One approach to this problem would be to stress source-audience homogeneity to increase likability (Simons, Berkowitz and Moyer 1970). Another alternative for decision-makers is to deliver a message in part contrary to the source's own interest and consequently increase perceived trustworthiness in a low credibility source (Walster, Aronson and Abrahams 1966).

Although this field study represents a more practical approach' to credibility research in advertising than a laboratory study, the findings should not be overgeneralized. The limitations of the sample do not make it representative of the U.S. population, even though the 18-25 age group does make up a significant proportion of the target market for the cars under study. Additionally, the special background of Mr. Iacocca and his company make this situation unusual. Situational variables and specific media influences were not controlled in this study. In fact, at the time or the research, Chrysler's faltering financial condition was much publicized and could have confounded the findings. Future research could attempt a more stringent research design which controls for situational variables where possible. Finally, this study is an attempt to monitor only one aspect of the use of celebrity spokespersons in advertising. Advertising decision-makers routinely consider potential source credibility as well as many other source influences when choosing a celebrity.

REFERENCES

Advertising Age (1981), "Chrysler Corporation" 57, 54, 56, 58.

Berlo, D.K., Lemert, J.B., and MertZ, R.J. (1969) "Dimensions for Evaluating The Acceptability of Message Sources," Public Opinion Quarterly, 33, 562-76.

Business Week (1978), "The Big New Celebrity Boom, 77, 80.

Delozier, M. Wayne (1976), The Marketing Communications Processes New York, New York: McGraw-Hill.

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

Dyer, R.F. and Kuehl, P.G. (1974), "The Corrective Advertising Remedy Of the F.T.C.: An Experimental Approach," Journal Of Marketing, 38, 48-54.

Fireworker, R.B. and Friedman, H.H. (1977), "The Effects Of Endorsements On Product Evaluation," Decision Sciences, 8, 576-583.

Fortune (1980), "The BOSS as Pitchman," 103, 66-73.

Freeman, W. (1957), The Big Name, New York, New York: Printer's Ink.

Friedman, Hershey H. and Friedman, Linda (1979), "Endorser Effectiveness by ProduCt Type," Journal Of Advertising Research, 19; 63-71.

Hovland, Carl, Janis, Irving L. and Kelley, Harold H. (1953), Communication and Persuasion, New Haven, Connecticut: Yale University Press.

Hovland, Carl L. and Weiss, Walter (1951), "The Influence Of Source Credibility On Communication Effectiveness," Public Opinion Quarterly, 159 635-650.

Hunt, H. Keith (1979), "SOUrCe EffeCtS, MeSSage Effects and General Effects in Counter Advertising," in M. Venkatesan (Ed.) Proceedings, Third Annual Conference, Association for Consumer Research.

Infante, Dominic A. (1980), "The Construct Validity Of Semantic Differential Scales for the Measurement Of Source Credibility," Communication Quarterly, 28, 19.

Kamen, Joseph M., Ashari, Abdul C. and Kragh, Judith R. (1975). "What a Spokesman Does for a Sponsor," Journal of Advertising Research, 15, 17-24.

Kelman, HerBert and Hovland, Carl (1953), "'Reinstatement' Of the Communicator In Delayed Measurement Of Opinion," Journal Of Abnormal and Social Psychology, 48 327-335.

Lutz, Richard J. (1975), Exchanging Brand Attitudes Through Modification Of Cognitive Structure," Journal Of Consumer Research 1, 49-59.

McGuire, W. (1969), "The Nature Of Attitudes and Attitude Change," in G. Lindzey and E. Aronson (Eds.) Handbook of Social Psychology, Vol. 3 Reading Massachusetts: Addison-Wesley.

Mowen, John C. and Brown, Stephen W. (1981), "On Explaining and Predicting the Effectiveness Of Celebrity Endorsers," in Kent B. Monroe (Ed.) Advances in Consumer Research Volume VIII, Ann Arbor, Michigan: Association for Consumer Research.

Rarick, Galen R. (1963), "Effects of Two Components of Communicator Prestige," unpublished doctoral dissertation. Stanford University.

Richey, Marjorie and Dwyer, J.D. (1970), "Negative Salience in Impressions of Character: Sex Differences," Psychonomic Science, 20, 77-79.

Richey, Marjorie H., Koenigs, Robert J., Richey, Harold W. and Fortin, Richard (1975), "Negative Salience in Impressions of Character: Effects of Unequal Proportions of Positive and Negative Information," Journal of Social Psychology, 97, 933-241.

Richey, Marjorie H., McClelland, Lucille, and Shimkunas, Algimantas M. (1967), "Relative Influence of Positive and Negative Information in Impression Formation and Persistence," Journal of Personality and Social Psychology, 6, 322-397.

Rudolph, H. (1947), Attention and Interest-Factors in Advertising, New York, New York: Printer's Ink.

Scott, Carol A. and Tybout, Alice M. (1981), "Theoretical Perspectives on the impact of Negative Information: Does Valence Matter?" in Rent B. Monroe (Ed.) Advances- in Consumer Research, Volume VII', Ann Arbor, Michigan: Association for Consumer Research.

Sereno, Kenneth R. and Hawkins, Gary J. (1967), "' he Effects of Variations in Speaker's Nonfluency Upon Audience Ratings of Attitude Toward the Speech Topic and Speakers' Credibility," Speech Monographs, 34, 58-64.

Settle, Robert B. and Golden, Linda L. (1974), "Attribution Theory and Advertiser Credibility," Journal of Marketing Research, 11, 181-185.

The 1979 Study of Media and Markets: Automobiles (1979). New York: Simmons Market Research Bureau.

Simons, Herbert W., Berkowitz, Nancy N. and Moyer, R. John (1970), "Similarity, Credibility and Attitude Change," Psychological Bulletin, 73, 1-16.

Sternthal, Brian, Philips, Lynn W. and Dholakia, Ruby (1978), "The Persuasive Effect of Source Credibility: A Situational Analysis," Public Opinion Quarterly, 42, 285-314.

Walster, E., Aronson, E. and Abrahams, D. (1966), "On Increasing the Persuasiveness of a Low Prestige Communicator," Journal of Experimental Social Psychology, 9, 395-342.

Weinburger, Marc C., Allen, Chris T. and Dillon, William R. (1981), "Negative Information: Perspectives and Research Directions," in Rent E. Monroe (Ed.) Advances in Consumer Research Volume VIII, Ann Arbor. Michigan: Association for Consumer Research.

Weinburger, Marc C. and Dillon, William R. (1980), "The Effects of Unfavorable Product Information," in Jerry Olson (Ed.), Advances in Consumer Research Volume VII, Ann Arbor, Michigan: Association for Consumer Research.

Weiss, Walter and Fine, Bernard J. (1956), "The Effect of Induced Aggressiveness on Opinion Change," Journal or Abnormal and Social Psychology, 59, 119.

Wright, Peter (1974),"The Harassed Decision Maker: Time Pressures, Distractions and the Use of Evidence," Journal of Applied Psychology, 59, 555-561.

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