Measurement of Consumer Susceptibility to Interpersonal Influence WILLIAM O. BEARDEN RICHARD G. NETEMEYER JESSE E. TEEL * The development of a scale for measuring consumer susceptibility to interpersonal influence is described. Consumer susceptibility to interpersonal influence is hypothesized as a general trait that varies across individuals and is related to other individual traits and characteristics (McGuire 1968). The construct is defined as the need to identify with or enhance one's image in the opinion of Significant others through the acquisition and use of products and brands, the willingness to conform to the expectations of others regarding purchase decisions, and / or the tendency to learn about products and services by observing others or seeking information fr~m <:>t~­ ers. A series of studies provides evidence to support the convergent and dlscnmlnant validity of a two-dimensional scale. A n important determinant of an individual's behavior is others' influence. Portrayal of products being consumed in social situations and the ~se of prominent/ attractive spokespersons endorsmg products is evidence of this belief. Further, models used to explain consumer behavior frequently include interpersonal influence. These models recognize that consumer behavior cannot be fully understood unless consideration is given to the effects of interpersonal influence on development of attitud~s, norms, values, aspirations, and purchase behavlOr (Stafford and Cocanougher 1977). Our objective is to develop a scale to assess c<;>nsumer susceptibility to interpersonal influence, WhICh is assumed to be a general trait that varies across individuals. This construct is derived from McGuire's (1968) concept of influenceability and is consistent with early research (e.g., Allen 1965; Asch 1958; Cox and Bauer 1964; Janis 1954), which demonstrated that individuals differ in their responses to social influence. Our efforts are also consistent with Kassarjian's (1971) suggestion that domain specific individual difference measures be developed to study personality issues in consumer research. INTERPERSONAL INFLUENCE THEORY AND RESEARCH An early review by McGuire (1968) summarized numerous theoretical and empirical articles dealing with various aspects of susceptibility to interpersonal influence and the relationship of susceptibility to other individual traits and characteristics. This review concluded that susceptibility to interpersonal influence is a general trait that varies across persons and that a person's relative influenceability in one situation tends to have a significant positive relationship to his or her influenceability in a range of other social situations. McGuire (1968, p. 1134) also cited evidence showing that conformity and persuasibility exist across occurrences. That is, people who conform to one source on one issue will likely conform to other sources on other issues. Further, susceptibility to influence by others is related to other personal characteristics (e.g., self-esteem, intelligence) (McGuire 1968, p. 1132; Petty and Cacioppo 1981, pp. 80-84). Cox and Bauer (1964) pointed out that people with low self-esteem comply with others' suggestions to avoid social disapproval. Cox and Bauer's research demonstrated that a relationship between self-confidence and persuasibility, previously found among men, also existed among women under some conditions. Likewise, Janis (1954) cited both clinical and correlational ·William O. Bearden and Jesse E. Teel are Professors of Marketing, College of Business Administration, Universit~ ofS,?uth Carolina Columbia, SC 29208. Richard G. Netemeyer IS Assistant Profess~r of Marketing, College of Business Administration, Lou!siana State University, Baton Rouge, LA 70803. The authors wish to thank the three anonymous reviewers for their helpful comments throughout the review process and Robert Burnkrant, Gil Churchill, and Paul Peter who commented on an initial draft of the article. 473 © JOURNAL OF CONSUMER RESEARCH. Vol. 15. March 1989 474 studies that suggested that some people consistently are amenable to social influence while others are consistently resistant. Based on his research, Janis (1954, p. 518) concluded that individuals with low self-esteem tend to be more readily influenced than others. Likewise, Berkowitz and Lundy (1957) found that persons low in interpersonal confidence are more susceptible to peer influence. Consumer Susceptibility to Interpersonal Influence Although susceptibility to interpersonal influence appears to be an important individual difference variable for the study of consumer behavior, it unfortunately has been neglected as a general trait in recent literature. However, numerous recent articles from psychological and consumer research have documented the existence of manifest interpersonal influence upon individual decision processes (e.g., Cohen and Golden 1972; Kassarjian and Robertson 1981; Moscovici 1985; Sherif 1935). In consumer research, these studies include the efforts of Ford and Ellis (1980), Moschis (1976), Stafford (1966), and Witt and Bruce (1972). However, most of these investigated the tendency of subjects to conform to group norms or to modify their judgments based upon others' evaluations and did not address the various types of interpersonal influence operative in a given situation. Only a few st.udies have addressed the dimensions of susceptibility to interpersonal influence and its effects upon decision processes. In this regard, Deutsch and Gerard (1955) posited that interpersonal influence is manifested through either normative or informational influences. Burnkrant and Cousineau (1975) defined normative influence as the tendency to conform to the expectations of others. Consumer research has separated normative influence into value expressive and utilitarian influences (Bearden and Etzel 1982; Park and Lessig 1977;Price,Feick,andHigie 1987). Value expressiveness reflects the individual's desire to enhance self-image by association with a reference group. Value expressiveness is motivated by the individual's desire to enhance or support his or her self-concept through referent identification (Kelman 1961). Value expressive influences operate through the process of identification, which occurs when an individual adopts a behavior or opinion of another because the behavior or opinion is associated with satisfying a self-defining relationship (Brinberg and Plimpton 1986; Park and Lessig 1977; Price et al. 1987). Value expressive influence was found to vary across selection decisions of products that differed in consumption conspicuousness and of services that varied regarding consumer preference heterogeneity and referent coorientation (similarity). Utilitarian THE JOURNAL OF CONSUMER RESEARCH influence, the other type of normative influence mentioned, is reflected in individuals' attempts to comply with the expectations of others to achieve rewards or avoid punishments, and it operates through the process of compliance (Burnkrant and Cousineau 1975, p. 207; see also Bearden and Etzel 1982; Park and Lessig 1977; Price et al. 1987). Compliance occurs when individuals conform to the expectations of others to gain rewards or to avoid punishments mediated by others. Deutsch and Gerard (1955) defined informational influence as the tendency to accept information from others as evidence about reality. Informational influence may occur in two ways. Individuals may either search for information from knowledgeable others or make inferences based upon the observation of the behavior of others (Park and Lessig 1977). Informational influence operates through the process of internalization, which occurs if information from others increases the individual's knowledge about some aspect of the environment. Informational influence has been found to affect consumer decision processes regarding product evaluations (Burnkrant and Cousineau 1975; Cohen and Golden 1972; Pincus and Waters 1977) and product/brand selections (Bearden and Etzel 1982; Park and Lessig 1977). In summary, recent research on manifest susceptibility to interpersonal influence has suggested that it is a multidimensional construct. Consequently, it seems reasonable to assume that the general trait of susceptibility to interpersonal influence also is multidimensional. Hence, recognizing the existence of alternative manifestations of interpersonal influence documented by recent research, consumer susceptibility to interpersonal influence is defined as the need to identify or enhance one's image with significant others through the acquisition and use of products and brands, the willingness to conform to the expectations of others regarding purchase decisions, and / or the tendency to learn about products and services by observing others and/ or seeking information from others. An initial effort to develop measures of three dimensions of manifest susceptibility to interpersonal influence was reported by Park and Lessig (1977). In their research, 14 statements were developed to assess value expressive, utilitarian, and informational influences. The statements were projective and framed for particular product and brand decisions. That is, respondents were asked on an item-by-item basis how "relevant" each interpersonal influence statement would be for an "unspecified other" for a given product/brand purchase. Although the Park and Lessig items were useful in several subsequent studies (Bearden and Etzel 1982; Brinberg and Plimpton 1986), these measures have limitations. To compare the Park and Lessig measures with general personality INTERPERSONAL INFLUENCE traits and individual characteristics, their statements have to be reworded so that the statements are not product and/ or situation specific. Further, phrasing the terms so that they are relevant to an unspecified other may add error that could cause some people to respond based either upon their own feelings or salient others' influence. Additionally, Park and Lessig (1977) did not report estimates of either internal consistency or test-retest reliability. Their only evaluation of the statements was a multitrait-multimethod analysis in which two products were used as traits. Other evidence of validity was not provided nor was the dimensionality of the measures tested. OVERVIEW OF SCALE DEVELOPMENT AND EVALUATION PROCEDURES The extant theoretical and empirical literature regarding the nature of susceptibility to interpersonal influence suggests that it is a general trait with multiple dimensions. The limitations of prior operational measures along with conflicting findings in related research suggest the need for a measure that can be applied across research settings without modification and that has been examined for reliability and validity and tested for dimensionality. Consequently, the objectives of the present research were to develop and test a general measure of consumer susceptibility to interpersonal influence. Development of the present scale began with a literature review that generated a large pool of items designed to measure the dimensions of interpersonal influence. The items were evaluated for content validity using judging procedures, resulting in a reduced set of items. A questionnaire then was administered to 220 adult consumers to assess the dimensionality and reliability of the present measures of susceptibility to interpersonal influence. Measurement items remaining after this study were reevaluated with data from a convenience sample of 141 student subjects. This study resulted in the final form and content of the present scale. Next, a series of studies was conducted to provide additional evidence regarding the validity and reliability of the scale. These studies consisted of comparisons of the present scale with individual difference measures, behavioral indices, external judge rankings, and motivation to comply. SCALE CONTENT An original pool of 166 items was generated from a review of prior research for use in the development of the present consumer susceptibility to interpersonal influence scale. Seven-place bipolar agree / disagree scales followed each of the statements (e.g., "When buying products, I generally purchase those brands 475 that I think others will approve of."). Following the item generation step, ambiguous items and statements with essentially identical meaning were eliminated, resulting in a revised pool of 135 items. Faculty and Ph.D. students were then used as judges in an evaluation ofthe content validity ofthe items. In this analysis, the five judges were exposed to the definition of each dimension plus a related explanation and an example item and were asked to allocate the statements to each dimension or to a "not applicable" category. Items that did not receive consistent classification by at least four ofthe five judges were eliminated. This initial analysis resulted in 86 statements for the three dimensions of susceptibility to interpersonal influence. Next, and similar to the procedure used by Zaichkowsky (1985), an additional four judges were given each construct's definition and asked to rate each statement as clearly representative of the construct, somewhat representative, or not representative. For the three dimensions, only items evaluated as clearly representative by three judges and somewhat representative by a fourth judge were retained. This process eliminated 24 more items. EVALUATING THE MEASURES Initial Administration Separate item analysis was performed for the 62 statements using the responses obtained from a convenience sample of 220 adult consumers. Items for each dimension of susceptibility to interpersonal influence were interspersed throughout the questionnaire used in the initial data collection. Item Reduction. The correlation of each item with the total score for each of the three susceptibility dimensions was computed (total scores were based upon a priori specification of each item's content). Items that did not have corrected item-to-total correlations above 0.50 were deleted (cf. Zaichkowsky 1985). Items that did not have statistically significant higher correlations with the dimension to which they were hypothesized to belong in comparison to item correlations with remaining dimension total scores also were deleted (Ruekert and Churchill 1984, p. 229). Principal axis factor analysis with oblique rotation (Churchill 1979) (restricting the solution to three factors) also was performed. Several additional items that failed to exhibit simple structure on factors representing the three dimensions of reference group influence were deleted. These analyses resulted in 18 items remaining in the scale. Reliability and Confirmatory Factor Analysis. The reliability and structure of the remaining items were examined using coefficient alpha and confirma- 476 tory factor analyses. First, a confirmatory factor analysis (Le., a three-factor correlated structure) revealed three items with low item reliabilities. These items then were deleted, leaving five items in each influence category: informational, utilitarian, and value expressiveness. A second confirmatory factor analysis was performed on the remaining 15 items. For a three-factor correlated structure, the overall chisquare statistic was 139.46 (df = 87, p < 0.01). Each indicator t-value exceeded 9.97 (p < 0.01). Although the overall chi-square statistic was significant, these results represent significant improvement over the chi-square values of 499.69 (df = 90) and 1717.13 (df = 105) obtained from a one-factor model and a null model. Coefficient alpha estimates were 0.86, 0.87, and 0.82 for informational, utilitarian, and value expressiveness, respectively. The estimates of construct reliability based upon the LISREL results were 0.86, 0.87, and 0.83. The more conservative variance extracted estimates were 0.55, 0.58, and 0.50. Dimensionality and Convergent/Discriminant Validity. Analysis of the relationships among the three factors revealed a very high intercorrelation (r = 0.92, sd = 0.02) between the utilitarian and value expressiveness factors. Corresponding intercorrelations between the utilitarian and informational measures and the value expressiveness and informational measures were 0.44 (standard error = 0.06) and 0.37 (standard error = 0.07), respectively. These results (similar to those reported by Brinberg and Plimpton 1986) render the dimensionality of the three-factor structure suspect. In an effort to investigate this concern, two additional tests were performed. First, to satisfy the requirements for discriminant validity, variance extracted estimates should be greater than the square of the correlation between constructs (Fornell and Larcker 1981, p. 46). This condition was satisfied for the informational-value expressiveness and the informational-utilitarian relationships, but was not met for the value expressiveness-utilitarian relation. This finding suggests that the value expressiveness and utilitarian dimensions are not distinct and that a two-factor model may be more appropriate. Second, tests were performed using the procedures recommended by Burnkrant and Page (1982, p. 557). In these tests, models, in which separate but correlated factors were allowed, were compared to models in which each pair offactors was hypothesized to have a unity correlation equivalent to a unidimensional model for the various pairs of factors. The difference between models is evaluated by a chi-square test with one degree of freedom. In each case, the correlation between factors was significant (one-tailed test, p < 0.01). For the informational-value expressiveness THE JOURNAL OF CONSUMER RESEARCH and informational-utilitarian comparisons, the confidence intervals for the correlation between factors did not contain the number one. Further, these two comparisons resulted in significant chi-square difference values with one degree offreedom of267 .57 and 328.22 for the informational-value expressive and informational-utilitarian comparisons. These results support the discriminant and convergent validity of the informational factor in comparison with the value expressiveness and utilitarian factors. The very large correlation coefficient (0.96) and the modest chisquare difference test (4.68) coupled with the variance extracted analysis recommended by Fornell and Larcker (1981) do not support the modeling of utilitarian and value expressiveness measures as distinct factors. These analyses then resulted in a two-factor structure in which utilitarian and value expressive measures were combined into a single normative factor (hereafter referred to as normative influence). The coefficient alpha and construct reliability estimates for this lO-item factor were both 0.91; the variance extracted estimate was 0.52. Second Administration The reliability and validity of the remaining 15 items again were examined in a second administration to 141 undergraduate student subjects. Again, the three-factor correlated model did not provide a better representation ofthe data in terms of structure (i.e., the number of factors and their discriminant validity) than did a two-factor correlated model. However, three items (one from each of the three original dimensions) exhibited low reliability. These items were dropped, which resulted in 12 remaining items-four informational and eight normative. For a two-factor correlated model, the chi-square goodness-of-fit statistic for 15 and 12 indicators, respectively, were 186.61 (df = 89, p < 0.05) and lO7.41 (df = 52, p < 0.05). For the two-factor, 12-indicator model, the intercorrelation between the two factors was 0.44. The overall chi-square fit statistic for a unidimensional model was 255.60 (df = 54, p < 0.01). (Reanalysis of the initial sample for the 12 items resulted in an overall chi-square statistic of 79.83 (df = 53, p < 0.01). Thus, there was a chi-square difference statistic between the 12-item, two- and one-factor models of 148.09 (df = 2, p < 0.05). This indicated that the two-factor model provided a better representation of the data than did the one-factor model. These analyses yielded a two-factor, 12-item scale comprised Offour informational and eight normative items. The mean scores, standard deviations, reliability estimates, and variance extracted estimates for both samples are summarized in Table 1. The statements along with the factor analysis item loadings for INTERPERSONAL INFLUENCE 477 TABLE 1 MEANS, STANDARD DEVIATIONS, RELIABILITY AND VARIANCE EXTRACTED ESTIMATES Reliability No. of items Range Mean SO Alpha Construct Initial administration Informational Normative 4 8 4-28 8-56 16.70 22.04 5.59 9.79 .83 .87 .83 .90 Second administration Informational Normative 4 8 4-28 8-56 19.02 27.18 4.45 9.15 .82 .88 .82 .89 Construct Retest Variance extracted .55 .53 .75 .79 .50 .54 TABLE 2 SUSCEPTIBILITY TO INTERPERSONAL INFLUENCE ITEMS Corrected item-to-total8 Item no. .61 .81 .86 .60 .45 .75 5 3 8 .75 .72 11 .69 .73 .63 .69 9 12 .70 .67 .64 .65 2 6 .78 .69 4 .75 .69 .66 .62 7 1 .74 .66 10 Factor coefficient Normative Informational Statement" I rarely purchase the latest fashion styles until I am sure my friends approve of them. It is important that others like the products and brands I buy. When buying products, I generally purchase those brands that I think others will approve of. If other people can see me using a product, I often purchase the brand they expect me to buy. I like to know what brands and products make good impressions on others. I achieve a sense of belonging by purchasing the same products and brands that others purchase. If I want to be like someone, I often try to buy the same brands that they buy. I often identify with other people by purchasing the same products and brands they purchase. To make sure I buy the right product or brand, I often observe what others are buying and using. If I have little experience with a product, I often ask my friends about the product . I often consult other people to help choose the best alternative available from a product class. I frequently gather information from friends or family about a product before I buy. • Factor corrected item-ta-total correlations. b The response format for each item is a seven-place rating scale ranging from strongly agree (7) to strongly disagree (1). the initial sample are presented in Table 2. Coefficient alpha estimates based upon the follow-up student data were 0.82 and 0.88 for the informational and normative factors, respectively. Based on the responses of 35 subjects participating in a follow-up three weeks later, test-retest reliability estimates were 0.75 and 0.79 for the informational and normative scales, respectively. All single-item and multi-item test-retest correlation estimates were significant (p < 0.01, one-tailed). Lastly, the 12-item two-factor structure was tested for stability across the initial administration (n = 220) and the second administration (n = 141) samples. The group chi-square statistics for an invariate factor structure and for the two groups separately were 186.19 (df = 118) and 173.38 = 106), respectively. This analysis then resulted in a nonsignificant chi-square difference test (df = 12). (df ADDITIONAL EVIDENCE OF VALIDITY AND RELIABILITY The scales for assessing the normative and informational dimensions of interpersonal influence were evaluated further in five separate studies. In the first study, we examined the relationships between the present measures and measures of attention-to-social-comparison-information (A TSCI) (Lennox and 478 Wolfe 1984) and self-esteem (Eagly 1967). The attention-to-social-comparison-information measure addresses the general tendency to conform and has been found to be related to fear of the evaluation of others. Lennox and Wolfe (1984) demonstrate that the measure directly addresses conformity (utilitarian) and the concern for the reaction of others (value expressive). Consequently, ATSCI was expected to be correlated more strongly with the normative factor than with the informational factor. For self-esteem, it was predicted that both dimensions of interpersonal influence would be inversely related to self-esteem. These later predictions are consistent with McGuire's (1968) and Stafford and Cocanougher's (1977) argument that susceptibility to interpersonal influence is inversely related to self-esteem. In the second study, we investigated the relationships between the alternative measures ofsusceptibility to interpersonal influence with an aggregated index of behaviors performed over multiple time periods. In the third and fourth studies, the measures were correlated with external judges' rankings of the subjects on the two dimensions of susceptibility to interpersonal influence. In the fifth study, we examined the relationship between the present measures of normative and informational influence and motivation to comply. ATSCI and Self-Esteem Forty-seven undergraduate business students responded to a questionnaire containing the 13-item ATSCI measure of Lennox and Wolfe (1984), the 20item self-esteem scale described by Eagly (1967), and the 12 interpersonal influence items included in the present scale. The coefficient alpha estimates for the ATSCI and the self-esteem scales were 0.82 and 0.88. The ATSCI measure correlated more strongly (p < 0.05 for a test of differences between dependent correlations) (Bruning and Kintz 1977) with the normative dimension of the present scale (r = 0.68, p < 0.05) than with the informational dimension (r = 0.16, p < 0.05). Both dimensions of the present scale had significant negative correlations with selfesteem. These estimates were -0.21 (p < 0.10) and -0.23 (p < 0.05) for the normative and informational measures. These results provide some evidence of construct validity. Relationships With Behavioral Indices Similar to the procedures suggested by Epstein (1979, 1980) and, more recently, by Lastovicka and Joachimsthaler (1988), the normative and informational measures of susceptibility to interpersonal influence were correlated with an aggregated index of consumer behaviors. As noted by Lastovicka and J oachimsthaler (1988), correlations between a trait and THE JOURNAL OF CONSUMER RESEARCH individual behaviors typically are weak. Additionally, correlations between traits and behaviors across single observations often are small. However, the strength of these relationships tends to increase as they are averaged over multiple observations and behaviors. First, a convenience sample of 35 undergraduate business students was used to elicit 21 normative and 17 informational behaviors reflecting the definitions of the normative and informational dimensions of susceptibility to interpersonal influence. Example normative behaviors included worrying about what others thought of the respondent's selection of clothing, copying the purchase behavior of someone the respondent admired, and having shown off something bought or owned to seek approval of others. Example informational behaviors included discussing products with friends or relatives, asking advice of others prior to making a purchase, avoiding a purchase because others said it was unsatisfactory, and having asked an "expert" about a contemplated purchase. A questionnaire containing the normative and informational items then was administered to a new sample of 43 subjects. The same 43 subjects reported at four separate intervals their performances regarding the 21 normative and 17 informational behaviors during the preceding 48 hours. Similar to the methods used by Epstein (1979), the self-reported behavioral measures were averaged over the odd (1,3) and even (2,4) administrations. The correlation between the even and odd summary averages for the normative and informational indices were 0.72 and 0.80. Based upon a 0,1 scoring and summed over observations and behaviors, the mean and standard deviation for the normative index were 21.75 and 11.49, respectively. Corresponding estimates for the informational index were 23.76 and 13.78. The normative scale was significantly correlated with the normative behavioral index (r = 0.37, p < 0.05). Further, the magnitude of this correlation compares favorably with the trait behavior correlations reported by Epstein (1979, p. 1118). Although not significant, the measure of informational interpersonal influence was correlated positively with the informational behavioral index (r = 0.15). Since only trait-behavior correlations of 0.29 or greater for single act criteria were reported by Epstein (1979), it is not possible to compute an average correlation from the summary of his research. However, an examination of Table 5 in Epstein (1979, p. 1118) revealed that 13 percent of the correlations between traits and his "objective events" were significant. In the present research, 33 percent and 12 percent, respectively, of the pairwise correlations between the average individual behaviors and the normative and informational measures of susceptibility to interpersonal influence were significant (p < 0.05). In general, these results provide stronger support for INTERPERSONAL INFLUENCE the normative scale than they do for the informational scale. External Judge Ratings The present scale was evaluated in two studies involving external judges ranking subjects on the two dimensions of susceptibility to interpersonal influence. The first external judge study was based upon data collected from 72 members of either a university fraternity (n = 39) or sorority (n = 33). Each of the subjects responded to the 12 items comprising the present scale. Two other members of each group evaluated their respective group members. Each judge was given a copy of the normative and informational dimensions of susceptibility to interpersonal influence and asked to rate each member of their group on each dimension using a seven-point rating scale. The single item measure was bounded by a rating of high in susceptibility to normative (informational) influence to a rating of low in susceptibility to normative (informational) influence. The average interjudge reliability was 0.66. The correlations between the judgment scores and the normative and informational measures were 0.40 (p < 0.05) and 0.24 (p < 0.05), respectively. In the second external judge study, a convenience sample of 43 undergraduates was given a questionnaire containing the proposed measures. They were instructed to have the questionnaire completed by someone they knew and to return it at the next class meeting. Upon receipt of the completed questionnaire, subjects were given the descriptions of the normative and informational dimensions and asked to rate the person who had completed the questionnaire. The same rating scales employed in the fraternity / sorority study were used to elicit the rankings. The correlations between judge rankings and the normative and informational measures of the present scale were 0.47 (p < 0.05) and 0.37 (p < 0.05), respectively. Motivation To Comply In a final test, it was predicted that both the normative and informational interpersonal influence scores would be related positively to individual motivations to comply with the expectations of others. Although this relationship' should be strongest for the normative dimension, as noted by Miniard and Cohen (1981), normative aspects (e.g., motivation to comply) of the theory of reasoned action also are related to informational interpersonal influence. To test this prediction, data were collected from 143 subjects. Subjects were provided with a shopping situation in which they were contemplating the purchase of a pair of jeans in a clothing store. While in the pro- 479 cess of deciding about the purchase, a close friend ( or classmate) entered the store and recommended purchase of the jeans being considered. Motivation to comply was operationalized as the sum ofthree statements developed from Ajzen and Fishbein (1980). The first item was operationalized as a seven-place scale included as part of the statement: "With respect to this purchase, I would very much like to/very much not like to do what my close friend (classmate) thinks I ought to do." The second item was operationalized as a seven-place bipolar unlikely /likely scale following the statement: "Regarding this purchase, I want to do what my close friend thinks I should do." The third item was a five-category vertical scale ranging from "not-at-all" to "very strongly" and followed the question: "How much do you want to do what your close friend thinks you should do?" The coefficient alpha estimate of this measure of motivation to comply was 0.81. Examination of the correlations between the susceptibility to interpersonal influence measures and motivation to comply supported the validity of the consumer susceptibility to interpersonal influence measures. The correlations between the motivation to comply and the informational and the normative measures were 0.39 (p < 0.05) and 0.59 (p < 0.05), respectively. These correlations also were significantly different from one another based upon a test of dependent correlations (p < 0.05). DISCUSSION We developed a two-dimensional measure of informational and normative interpersonal influence and examined its reliability and validity. A 12-item, twofactor scale emerged from the scale development and validation efforts. The scale was found to be reliable and valid across multiple samples of both student and nonstudent subjects. Correlations with other constructs demonstrated convergent and discriminant validity. Replicated confirmatory factor analysis demonstrated a stable two-factor correlated structure. This finding is consistent with other examinations of consumer interpersonal influence that limit influence scores to normative and informational dimensions (cf. Burnkrant and Cousineau 1975). Lastly, the scale was correlated as predicted with independent judge ratings, behavioral indices, and measures of motivation to comply. Generally, correlations between the normative dimension and these other measures consistently were stronger than the correlations involving the informational measure. However, the correlations between both measures of susceptibility to interpersonal influence and the related measures demonstrated a range and pattern of significance comparable to other scale development studies (e.g., Cacioppo and Petty 1982; Swap and Rubin 1983). 480 THE JOURNAL OF CONSUMER RESEARCH Certainly, additional tests ofthe scale are needed to establish its validity and final form. A caveat is also in order regarding the direction of the items. The final form of the scale reported here included only positively worded statements, leaving open the possibility for some response bias. Next, factor analysis of our data consistently revealed a two-factor correlated structure. However, our justification for multiple dimensions of consumer susceptibility to interpersonal influence largely was based upon research addressing manifest susceptibility to interpersonal influence. Further experimental research along the lines of Burnkrant and Cousineau (1975) is warranted to provide further evidence of the dimensionality of interpersonal influence. Although we tried to adhere to recommended scale development procedures, it may also be that the final set of utilitarian items contains statements that reflect response to a general normative influence rather than responses to gain rewards and punishments. However, the items are couched in terms of approval and the expectations of others, which underly the process of compliance (Burnkrant and Cousineau 1975). [Received July 1987. Revised August 1988.J REFERENCES Ajzen, Icek and Martin Fishbein (1980), Understanding Attitudes and Predicting Social Behavior, Englewood Cliffs, NJ: Prentice-Hall. Allen, Vernon L. (1965), "Situational Factors in Conformity," in Advances in Experimental Psychology, Vol. 2, ed. Leonard Berkowitz, New York: Academic Press, 133-175. Asch, S. (1958), "Effects of Group Pressure Upon the Modification and Distortion of Judgments," in Readings in Social Psychology, for the Committee on the Teaching of Social Psychology of the Society for the Psychological Studies of Social Studies, New York: Holt, Rinehart & Winston, 174-182. Bearden, William O. and Michael J. Etzel (1982), "Reference Group Influence on Product and Brand Purchase Decisions," Journal of Consumer Research, 9 (September),183-194. Berkowitz, Leonard and Richard M. Lundy (1957), "Personality Characteristics Related to Susceptibility to Influence by Peers or Authority Figures," Journal ofPersonality, 25 (I), 306-316. Brinberg, David and Linda Plimpton (1986), "Self-Monitoring and Product Conspicuousness on Reference Group Influence," in Advances in Consumer Research, Vol. 13, ed. Richard J. Lutz, Provo, UT: Association for Consumer Research, 297-300. Bruning, James L. and B.L. Kintz (1977), Computational Handbook ofStatistics, Glenview, IL: Scott, Foresman. Burnkrant, Robert E. and Alain Cousineau (1975), "Informational and Normative Social Influence in Buyer Behavior," Journal of Consumer Research, 2 (December),206-215. - - and Thomas J. Page, Jr. (1982), "An Examination of the Convergent, Discriminant, and Predictive Validity Of Fishbein's Behavioral Intention Model," Journal ofMarketing Research, 19 (November), 550561. Cacioppo, John T. and Richard E. Petty (1982), "The Need for Cognition," Journal ofPersonality and Social Psychology, 42 (1),116-131. Churchill, Gilbert A., Jr. (1979), "A Paradigm For Developing Better Measures of Marketing Constructs," Journal ofMarketing Research, 16 (February), 64-73. Cohen, Joel B. and Ellen Golden (1972), "Informational Social Influence and Product Evaluation," Journal of Applied Psychology, 56 (February), 54-59. Cox, Donald and Raymond A. Bauer (1964), "Self-Confidence and Persuasibility in Women," Public Opinion Quarterly, 28 (Fall), 453-466. Deutsch, Morton and Harold B. Gerard (1955), "A Study of Normative and Informational Influence Upon Individual Judgment," Journal of Abnormal and Social Psychology, 51 (November), 629-636. Eagly, Alice H. (1967), "Involvement As a Determinant of Responses to Favorable and Unfavorable Information," Journal of Personality and Social Psychology, 7 (November),1-15. Epstein, Seymour (1979), "The Stability of Behavior: I. On Predicting Most of the People Much of the Time," Journal of Personality and Social Psychology, 37 (July),I097-1126. - - (1980), "The Stability of Behavior: II. Implications for Psychological Research," American Psychologist, 35 (September), 790-806. Ford, Jeffry D. and Elwood Ellis (1980), "A Re-examination of Group Influence on Member Brand Preference," Journal ofMarketing Research, 17 (February), 125-132. Fornell, Claes and David Larcker (1981), "Evaluating Structural Equation Models With Unobservable Variables and Measurement Error," Journal of Marketing Research, 18 (February), 39-50. Janis, Irving L. (1954), "Personality Correlates of Persuasion," Journal ofPersonality, 22 (I), 504-518. Kassarjian, Harold H. (1971), "Personality and Consumer Behavior: A Review," Journal ofMarketing Research, 8 (November), 409-418. - - and Thomas S. Robertson, eds. (1981), "Social Processes," in Perspectives in Consumer Behavior, Glenview, IL: Scott, Foresman, 317-328. Kelman, Herbert C. (1961), "Processes of Opinion Change," Public Opinion Quarterly, 25 (Spring), 57-78. Lastovicka, John L. and Erich A. Joachimsthaler (1988), "Improving the Detection of Personality-Behavior Relationships in Consumer Behavior," Journal of Consumer Research, 14 (March), 583-587. Lennox, Richard D. and Raymond N. Wolfe (1984), "Revision of the Self-Monitoring Scale," Journal of Personality and Social Psychology, 46 (6), 1349-1364. McGuire, William J. (1968), "Personality and Susceptibility to Social Influence," in Handbook of Personality Theory and Research, eds. Edgar F. Borgatta and William W. Lambert, Chicago: Rand McNally, 11301187. INTERPERSONAL INFLUENCE Miniard, Paul W. and Joel B. Cohen (1981), "An Examination of the Fishbein-Ajzen Behavioral Intentions Model's Concepts and Measures," Journal 0/ Experimental and Social Psychology, 17 (July), 309-339. Moschis, George P. (1976), "Social Comparison and Informal Group Influence," Journal o/Marketing Research, 13 (August), 237-244. Moscovici, Serge (1985), "Social Influence and Conformity," in Handbook o/Social Psychology, Vol. 2, eds. Gardner Lindzey and Elliot Aronson, New York: Random House, 347-412. Park, C. Whan and Parker V. Lessig (1977), "Students and Housewives: Differences in Susceptibility to Reference Group Influence," Journal 0/ Consumer Research, 4 (September), 102-110. Petty, Richard E. and John T. Cacioppo (1981), Attitudes and Persuasion: Classic and Contemporary Approaches, Dubuque, IA: Wm. C. Brown. Pincus, Steven and L.K. Waters (1977), "Informational Social Influence and Product Quality Judgments," Journal 0/Applied Psychology, 62 (5),615-619. Price, Linda L., Lawrence F. Feick, and Robin H. Higie (1987), "Preference Heterogeniety and Coorientation as Determinants of Referent Influence in the Choice of 481 Service Providers," working paper, Department of Marketing, Katz Graduate School of Business, University of Pittsburgh, Pittsburgh, PA 15260. Ruekert, Robert W. and Gilbert A. Churchill, Jr. (1984), "Reliability and Validity of Alternative Measures of Channel Member Satisfaction," Journal 0/ Marketing Research, 21 (May), 226-233. Sherif, Muzafer (1935), "A Study of Some Social Factors in Perception," Archives o/Psychology, 27 (187), 60. Stafford, James E. (1966), "Effects of Group Influence on Consumer Brand Preferences," Journal 0/ Marketing Research, 3 (February), 68-75. - - and Benton A. Cocanougher (1977), "Reference Group Theory," in Selected Aspects o/Consumer Behavior, Washington, D.C.: Superintendent of Documents, U.S. Government Printing Office, 361-380. Swap, Walter C. and Jeffrey Z. Rubin (1983), "Measurement ofInterpersonal Orientation," Journalo/Personality and Social Psychology, 44 (1),208-219. Witt, Robert E. and Grady D. Bruce (1972), "Group Influence and Brand Choice Congruence," Journal 0/ Marketing Research, 9 (November), 440-443. Zaichkowsky, Judith Lynne (1985), "Measuring the Involvement Construct," Journal 0/ Consumer Research, 12 (December), 341-352.