An Exploratory Investigation of Consumer Innovativeness and Interpersonal Influences

William O. Bearden, University of South Carolina
Stephen E. Calcich, University of South Carolina
Richard Netemeyer, University of South Carolina
Jesse E. Teel, University of South Carolina
ABSTRACT - Alternative weights within the theory of reasoned action were proposed for both innovators and noninnovators. The motels were tested for nine brands for both a private necessity and a public luxury product--situations in which reference group influences are expected to vary. Tests of the behavioral intention equation supported the hypothesized pattern. However, the results of the normative crossover effects analysis were inconclusive.
[ to cite ]:
William O. Bearden, Stephen E. Calcich, Richard Netemeyer, and Jesse E. Teel (1986) ,"An Exploratory Investigation of Consumer Innovativeness and Interpersonal Influences", in NA - Advances in Consumer Research Volume 13, eds. Richard J. Lutz, Provo, UT : Association for Consumer Research, Pages: 77-82.

Advances in Consumer Research Volume 13, 1986      Pages 77-82

AN EXPLORATORY INVESTIGATION OF CONSUMER INNOVATIVENESS AND INTERPERSONAL INFLUENCES

William O. Bearden, University of South Carolina

Stephen E. Calcich, University of South Carolina

Richard Netemeyer, University of South Carolina

Jesse E. Teel, University of South Carolina

ABSTRACT -

Alternative weights within the theory of reasoned action were proposed for both innovators and noninnovators. The motels were tested for nine brands for both a private necessity and a public luxury product--situations in which reference group influences are expected to vary. Tests of the behavioral intention equation supported the hypothesized pattern. However, the results of the normative crossover effects analysis were inconclusive.

INTRODUCTION

Consumer innovativeness and the diffusion of innovations are two of the most frequently researched concepts in consumer behavior. Midgley and Dowling (1978) have provided an excellent summary of the issues involved in the conceptual definition and measurement of consumer innovativeness. One outcome of their thoughtful discussion is the conclusion that "innovativeness is the degree to which an individual makes innovation decisions independently of the communicated experience of others" (Midgley and Dowling 1978, p. 235). As such, innovativeness is posited to be a generalized personality trait and adoption decisions are seen as a function of product interest, individual situations, personal characteristics, and a network of interpersonal influence. It is this later premise, i.e., the potential relationship between innovativeness and differences in interpersonal influence, that precipitated the present study.

In the exploratory effort that follows, differences between innovators and noninnovators in interpersonal influence effects within the theory of reasoned action are hypothesized and tested for multiple brands of two products assumed to differ in perceived reference group influence. Prior to describing that study and the ensuing results, a brief review of prior research and the research hypotheses are presented.

THEORETICAL PERSPECTIVES

Innovativeness and Interpersonal Influence

Much communication and diffusion research has ignored the effects of social influence (Rogers 1976, p. 298). Social influence can be categorized into informational and normative influence. Informational influence occurs when individuals accept information as evidence of reality while normative influence involves conformity to the expectations of others (Burnkrant and Cousineau 1975, pp. 206-207). These social influences are thought to operate through the processes of internalization, identification, and compliance (Kelman 1961). Information is internalized if it is perceived as enhancing the individual's knowledge of the environment. Normative social influence may also occur if the individual is motivated to realize a reward or avoid a punishment. Compliance in this situation would occur if the individual believes the behavior is visible or known to others.

It is generally accepted that word-of-mouth communications often affect the eventual acceptance and diffusion of new products (Midgley and Dowling 1978; Rogers 1983). However, empirical evidence addressing the link between the trait of innovativeness and perceived interpersonal influence is limited. Further, the evidence that is available is equivocal. For example, Berning and Jacoby (1974), in their study of information acquisition in new product purchase decisions, concluded that the decision-making process preceding the purchase of an innovation differs from the process preceding the purchase of an established alternative. The principle difference underlying that process was concluded to be that innovators first seek information from other people. In contrast, and in support of Midgley and Dowling's view, Carlson and Grossbart (1984) found a significant positive relationship between independent judgment making and inherent novelty seeking. [It should be noted that the Berning and Jacoby (1974) study focused upon actual new products while Carlson and Grossbart (1984) assessed their constructs using generalized multi-item scales.]

The replication and extension of Raju's (1980) optimum stimulation paradigm by Joachimsthaler and Lastovicka (1984) indicates that the relationship between innovativeness and information seeking is quite complex (bicausal). However, close examination of the content of the statements comprising the information seeking scale reveals that the items comprising the scale contained a mix of both personal and impersonal sources of information.

Interpersonal Influence Differences

One explanation of interpersonal influence Ln consumer behavior is the theory of reasoned action (Ajzen and Fishbein 1980). This frequently applied model provides a vehicle for exploring some of the differences in interpersonal influences between innovators and noninnovators. Interpersonal influence is assumed operative in several linkages within the motel. The basic equation proposes that behavioral intention is affected directly by subjective norms and attitudes-toward-the act of purchase. Social influences on behavioral intentions are thought to operate principally through an overall measure of subjective norms. Subjective norms, in turn, are presumed to stem from the combination of normative beliefs, i.e., beliefs that certain referents think the person should or should not perform a particular behavior, and motivations-to-comply with the referent group.

Two phenomena exist which make the isolation of perceived reference group influences in the theory of reasoned action problematic. These issues are raised nov since one provides a basis for one of the research hypotheses proposed in the next section. These two issues are: (1! multicollinearity between the Aact and SN predictors of behavioral intention and (2) crossover effects from the motel's exogenous cognitive and normative structures to the summary constructs of attitudes and subjective norms.

Multicollinearity among predictor variables in the theory of reasoned action makes the interpretation of the relative weights in the motel problematic. Since this paper attempts to examine the relative weights in the motel, caveats regarding this problem are warranted. However, in an effort to address the multicollinearity issue, both correlated and uncorrelated motels are tested and, for one set of analyses, an extra-sums-of-squares test is used involving reductions in explained variance when predictors are omitted. Miniard and Cohen (1981) have outlined the theoretical arguments for interdependence among the personal and normative components of the Fishbein and Ajzen model. In fact, it has been proposed that for those interested in distinguishing between personal and normative influences that an alternative model is more appropriate. These criticisms stem largely from the premise that the normative component does not discriminate adequately between informationally-based social influence and influences that are truly normative in nature (Miniard and Cohen 1981, 1983).

Crossover effects between normative structure and attitude and between cognitive structure and subjective norms have been addressed in recent tests of the theory of reasoned action (Ryan 1982; Shimp and Kavas 1984). Since differences for noninnovators and innovators in crossover effects for normative structure to attitudes are hypothesized and tested in the present study, some justification for these effects is warranted. Ryan (1982) proposes that inferential belief formation accounts for many of the crossover effects in the theory of reasoned action and the interdependency between attitudinal and normative variables. From Ryan's summary of earlier research, empirical studies have found that (1) normative information was more favorably evaluated if it supported the listener's views, (2) respondents more readily accepted an attitude in line with reference group majority opinion, and (3) respondents were more likely to act in accord with stated attitudes if such attitudes were consistent with the actions of others. Shimp and Kavas (1984) argue for a normative structure-attitude crossover relationship for couponing. The expectation is that because couponing is a form of behavior that has direct consequences for important others, the consumer will utilize others' opinions in addition to his/her personal beliefs when forming attitudes toward coupon usage.

RESEARCH HYPOTHESES

Innovativeness has been defined as the degree to which an individual makes innovation decisions independently of the communicated experience of others, i.e., information passed verbally between individual consumers (Midgley and Dowling 1978, p. 235). Based on the premise that noninnovators learn from others (including some feel for their expectations), the following hypothesis regarding the relative influence of subjective norms is proposed: [This hypothesis is similar to the recent proposition of Gatignon and Robertson (1985, p. 856): consumers who are highly dependent on normative influence (conformity intention) will be slower to adopt.]

H1: Subjective norms will exhibit less relative influence on behavioral intentions for innovators than for noninnovators.

Crossover effects from normative structure to attitude representing the impact of social influence on affect reflect a susceptibility or general willingness to incorporate or infer social influences on one's attitude structure. Normative information from others signals the direction of the attitudinal norm and implies that others will like the subject for holding an attitude consistent with these norms (Oliver and Bearden .985). Again, since innovators are defined as making decisions independently of others, ant, in contrast, other adopter categories seek information "from their near peers whose subjective opinion of the innovation ... is most convincing" (Rogers 1983, p. 170), a second hypothesis regarding crossover effects is proposed:

H2: Innovators will exhibit a lesser tendency than noninnovators to exhibit crossover effects from normative structure to attitude.

As alluded to above, the normative structure component of the theory of reasoned action reflects exclusively beliefs about referent expectations (Miniard and Cohen 1981, p. 315). This conceptualization of normative influence is similar to the utilitarian influences of Deutsch and Gerard (1955). However, social-influence can be based upon both informational and normative sources. The findings of Berning and Jacoby (1974) suggest that innovators seek information from others. In contrast, the Midgley and Dowling (1978) definition stresses the absence of communicated experiences for innovators (i.e., information). Given these conflicting perspectives, no formal hypothesis is offered regarding the differential role of informational versus utilitarian influences. However, the relationships between overall subjective norms and both informational and utilitarian (normative) influences will be examined for innovators and noninnovators.

METHOD

Data were collected from 139 undergraduate male and female marketing students. Responses were obtained for nine brands of two products -- toothpaste and luxury automobiles. The products were selected to represent products varying in susceptibility to reference group influence. Consistent with the typologies of Bourne (1957) and Bearden and Etzel (1982), toothpaste and luxury automobiles represent private necessities (PRN) and public luxuries (PUL), respectively. Private necessities are hypothesized to be not susceptible to reference group influence while reference group influence is hypothesized to operate for brand selection decisions among luxury automobiles.

Analysis

The sample was divided equally using a median split into innovators and noninnovators. Responses to the ten seven-place agree-disagree statements comprising the innovativeness factor identified by Raju (1980, p. 278) were used to operationalize innovativeness. (Efforts in this study to assess the reliability of the scale are described below.) Aggregate cross-sectional tests of the principle behavioral intention equation of the theory of reasoned action were conducted first for each subsample. Second, crossover effects from normative structure to attitudes were tested for both innovators and noninnovators. Third ordinary least squares regression was used to estimate the subjective norm equations using utilitarian and informational influences as predictors. For the aggregate Aact-SN and the crossover effects analyses, correlations were used as input into LISREL VI to test the alternative models (Joreskog and Sorbom 1983).

Operational Measures

The two sets of modal referent groups (Ajzen and Fishbein 1980, p. 75) used to represent normative influences were elicited in separate pretests of 15 undergraduate students. The elicitation procedure employed open-ended questions to identify referents most likely to influence toothpaste and luxury automobile purchases. Three referents were found for each product: close friends, family members, and dentist for toothpaste and close family members, friends and relatives, and salesperson for luxury automobiles. All measures were developed in the context of the respondent's next purchase.

Two measures of subjective norms (SN) were employed. One measure was the standard item: "Most people who are important to me think I (should/should not) purchase (brand)." The alternative measure was worded: "The important people in my life would ... (want/not want) me to purchase (brand)." Respondents provided their attitude toward the act of purchasing (Aact) each brand on three seven-point semantic differential scales labeled good-bat, foolish-wise, and beneficial-harmful. Behavioral intentions (BI) was assessed using a single seven-place item scale labeled likely-unlikely designed to reflect intentions to purchase. The direction of the SN and Aact items were reversed to limit acquiescence bias.

Normative beliefs were also assessed using measures similar to those typically employed in tests of the Fishbein extended model. Respondents indicated their normative beliefs for each referent by indicating "the extent to which (referent) would expect you to buy (brand)." The seven-point scales were labeled definitely would not expect me to buy (-3)/definitely would expect me to buy ("3). Motivation-to-comply was operationalized using measures such as: "With respect to the purchase of toothpaste, I want very much to (6)/I want very much not to (0) do as my (referent) expects." The use of a unipolar motivation-to-comply scale is consistent with the premise that people are unlikely to be motivated to do the opposite of what their salient referents think they should to (Ajzen and Fishbein 1980, p. 75). These measures were also converted into analogous statements reflecting the availability and willingness to obtain information (cf. Park and Lessig 1977). These latter operationalizations were used later to explore the feasibility of including both utilitarian and informational measures as predictors of subjective norms.

Reliability Estimates and-Manipulation Checks

Innovativeness. The coefficient alpha estimate of internal consistency reliability for the ten-item innovativeness scale was .77. Additional evidence of reliability is provided by its correlation with other measures collected to examine the properties of the scale. First, the innovativeness measure was significantly correlated (Pearson correlation = .63, p < .01) with a similarly measured seven-item scale (reliability = .71) of agree-disagree statements used by Needham, Harper, and Steers. Second, the RaJu scale was positively correlated (r = .44, p < .01) with a summed index reflecting the purchase (non-purchase) of 30 newly introduced or innovative products and services. This type of cross-sectional innovativeness measure has the advantage of controlling for some of the situational and communication effects associated with individual products (Midgley and Dowling 1978) and has been used in a number of earlier innovativeness studies (e.g., Darden and Reynolds 1974; Baumgarten 1975).

A twenty-item measure of dogmatism (alpha = .68) (Troldahl and Powell 1965) was also assessed to assist in examining the validity of the Raju scale. Dogmatism is one of the psychological traits proposed by Midgley and Dowling (1978, p. 236) as affecting innate innovativeness. A strong, negative correlation was anticipated between dogmatism and the innovativeness measure. However, the resulting correlation was .09 (p=.15). In retrospect, this low correlation might be attributed to one or both of two reasons. First, a closer reading of the items comprising the twenty-item dogmatism measure questions whether or not the scale addresses receptivity of new ideas (close-mindedness). Second, the results of Coney and Harmon (1975) suggest that the relationship between dogmatism and innovativeness is subject to situational influences and may not be linear as initially thought.

The dimensionality of the scale was also examined using confirmatory factor analysis (Joreskog and Sorbom 1983). Three alternative factor motels were tested: a one-factor model, a three-factor uncorrelated model, and a three-factor correlated model. The three factor structure was hypothesized based on the item category assignments used by Raju (1980, p. 278) in which different items were posited as reflecting multiple categories of exploratory purchase behavior. The one factor solution (chi-square = 86.41, p < .01, 35 df) provided the best fit to the data. The three-factor correlated model was indeterminant; the uncorrelated solution resulted in a chi-square of 192.81 (p < .01, 35 tf). In the one factor solution, all indicator estimates were significant; however, several of the individual items did possess low loadings (i.e., below .50). Nine of these same items were used by Joachimsthaler and Lastovicka (1984) to operationalize innovativeness in their reevaluation of Raju's general explanation of the role of optimum stimulation level (OSL). Their measurement model results indicated that split-half combinations of the items provided reliable indicators of innovativeness. [2Data from a two-state consumer panel study (n=218) were also available regarding the Raju innovativeness items, the Needham, Harper, and Steers measures and dogmatism scale. The coefficient alpha estimates were .81, .71, and .79, respectively. Similar validating correlations between innovativeness and the Needham scale (r = .59, p < .01) and the dogmatism measure (r = -.06, n.s.) were found.]

Subjective Norms and Attitudes. Multiple indicators for social norms and attitudes provided the opportunity to also assess their reliability. These estimates along with average brand/construct correlations are provided in Table 1. The average internal consistency reliability estimates are shown in the diagonal. These high reliability estimates are undoubtedly due to the similarity in measurement methods across indicators (e.g., bipolar semantic differentials following an attitude statement).

TABLE 1

Aact AND SN BRAND INTERCORRELATIONS

Given the large number of brands addressed and the length of the survey, the potential for methods variance was present. Some evidence regarding the discriminant validity of the SN and Aact measures and the intercorrelation among predictors is provided by examining the remaining correlations presented in Table 1. The bracketed entries suggest only moderate correlations between the SN and Aact measures within construct and product category (i.e., correlations among the SN measures for toothpaste). The low across product and construct correlations give additional evidence of the discriminant validity of the SN and Aact measures for the different brands and products. However, the underlined table entries suggest considerable intercorrelation among the two predictors in the behavioral intention equation of the theory of reasoned action.

Product Perceptions. Manipulation checks were included to verify respondent perceptions of the two types of products studied. These manipulation checks were designed to assess perceptions regarding the conspicuousness dimensions (i.e., public-private and luxury-necessity) underlying the reference group hypotheses of Bourne (1957) and Bearden and Etzel (1982). Definitions of each conspicuousness dimension were provided prior to seven-place scales for each product labeled "a public product-a private product" and "a necessity-a luxury." The resulting paired t-test values were 24.42 (p < .01) and 37.27 (p < .01) for the public-private and luxury-necessity dimensions, respectively. These significant values and the relative mean scores indicate that the two product categories were perceived as intended.

RESULTS

Analyses were first conducted across subjects for both the behavioral intention equation and for the hypothesized crossover effects. The aggregate tests of the behavioral intention equation are presented in Table 2 for both an overall analysis (i.e., across the nine brands for each product) and averaged across brands. [Ryan (1982, p. 269) reported similar weights (Aact beta = .25 and SN beta = .53) in his pooled regression results regarding behavioral intention prediction for a fictitious toothpaste brand and a .60 intercorrelation between Aact and SN.] In general, the resulting pattern among the Aact and SN weights were similar for the overall and averaged models and for the correlated and uncorrelated results. The effects of allowing the two predictors of behavioral intention to be correlated resulted in improved fit as estimated by the chi-square statistic and some reduction in the absolute value of the standardized beta coefficients:

TABLE 2

AGGREGATE ANALYSIS OF BEHAVIORAL INTENTION MODEL

The nonsignificant chi-square values for the individual brand models suggest that the correlated model provides an adequate fit to the data. This conclusion is supported by the relatively high adjusted goodness-of-fit indices.

Examination of the relative sizes of the beta coefficients supports the hypothesis of higher relative normative influence for noninnovators for the luxury automobiles. Similar (but slight) differences in weights for the private necessity product (i.e., the product for which reference group influence was not hypothesized) were also found. The differences between the Aact and SN weights for the averaged results regarding the public luxury product were .16 and .48 for the innovators and noninnovators, respectively. It is interesting to note, however, that much of this differential results from the lower Aact weight for the noninnovators.

Differences in crossover effects from normative structure to attitudes (cf., Shimp and Kavas 1984) were also posited for innovators and noninnovators. The model tested in these analyses is shown in Figure A. The corresponding results are summarized in Table 3 for both correlated and uncorrelated versions of the motel. Due to the limited number of indicators and exogenous constructs, bicausal paths between Aact and SN could not be estimated simultaneously. Consequently, an alternative model allowing the construct error terms for Aact and SN to be correlated was tested. The principal differences between the correlated and uncorrelated models were associated largely with motel fit. No significant changes in structural path coefficients were found between the correlated and uncorrelated models.

FIGURE A

CROSSOVER EFFECTS CAUSAL MODEL

TABLE 3

AGGREGATE ANALYSIS OF NORMATIVE CROSSOVER EFFECTS

As shown in Table 3, significant crossover effects were found for both groups and for both products. However, the largest crossover effects were found for the innovators in the automobile analyses. This finding is contrary to the hypothesized effects anticipated. It is interesting to note that the pattern of relative weights for attitudes and subjective norms was consistent with the results presented in Table 2 (i.e., prior to introducing the summated normative belief construct).

An additional set of aggregate brand analyses were run to test the hypotheses reflecting informational versus normative influences on subjective norms for innovators and noninnovators. These results are depicted in Table 4. Again, the estimates represent averages across nine brands with each regression using 71 and 68 subjects as data points for the innovative and noninnovative groups, respectively.

TABLE 4

AGGREGATE ANALYSIS OF INFORMATIONAL AND NORMATIVE PREDICTION OF SUBJECTIVE NORMS

The hypothesized pattern of expected greater relative normative influence for noninnovators was observed only for the toothpaste brands. For the luxury automobiles, normative expectations were more highly correlated with subjective norms (in comparison to informational influence) for both innovators and noninnovators. The interpretation of these regressions was also confirmed by examination of predictor usefulness - defined as the amount of decrease in explained variance when a variable is dropped from an equation (Draper and Smith 1966). For both groups and across brands, measures of variable usefulness were consistent with the relative beta coefficient sizes.

DISCUSSION

This exploratory effort attempted to examine the role of interpersonal influences in Fishbein and Ajzen's theory of reasoned action across two groups of respondents partitioned into innovators and noninnovators. The study was designed as an initial effort to examine the role of interpersonal influence perceptions across individuals differing along a general measure of innovativeness, and as such, represents only a limited test of Midgley and Dowling's (1978) innovativeness definition. The study also suffers from a number of limitations. Caveats are in order regarding the use of existing brands and student subjects. To maintain a reasonable sample size, a median sample split was used. However, innovators are typically described as only the first few number (i.e., 2, 5, or 10 percent) of purchasers. Further, innovativeness was measured using scale items taken from another study. While the scale did appear to be reliable, the need remains for the formal development of an innovativeness scale using sound psychometric scale development procedures. Lastly, while substantial time was provided for the respondents and careful monitoring of the actual data collection effort was insured, the task required of the subjects was lengthy.

With these limitations in mind, this exploratory study does provide some implications regarding interpersonal influences on innovative behavior and supports recent tests of the theory of reasoned action. Regarding the latter, social norms and attitudes were correlated with intentions at both the aggregate and individual levels. Some evidence was also provided that the Aact and SN weights reflected differences for the two products (i.e., the public luxury and private necessity) as predicted by the theory. However, significant multicollinearity among the predictors of intentions and crossover effects from normative structure were found. These results reiterate the conclusions of Ryan (1982, p.274) and Shimp and Kavas (1984, p. 807) that behavioral intentions are not a function of parallel, independent sets of inseparable attitudinal and normative variables, but of a complex set of interdependencies.

Crude measures of informational reference group influence were found correlated with subjective norms and did explain (at least, for one product) some unique variance beyond that accounted for by the traditional normative-expectation-based predictor of subjective norms. This very tentative finding lends credence to the arguments of Miniard and Cohen (1981, 1983) and the growing recognition that the social influence variables are under-developed.

The results of this study also provide some mixed support for Midgley and Dowling's conceptual view of innovativeness. On the positive side, the relative weights derived from aggregate tests of the behavioral intention equation of the theory of reasoned action and the prediction of subjective norms by the normative and informational measures are somewhat consistent with that predicted by Midgley and Dowling's arguments. First, for both the correlated and uncorrelated models, the relative pattern among the Aact and SN weights varied as anticipated for the public luxury and private necessity products. (It was noted, however, that the subjective norm coefficients were high for both innovators and noninnovators.) Second, highest normative weights (compared to the informational weights) were found for the product hypothesized to involve greater reference group influence. In contrast, the results of the crossover effects were not directly supportive of the hypothesized reference group paths. For both products (and particularly the luxury automobiles), the crossover effects were unexpectedly highest for the innovators.

REFERENCES

Ajzen, Icek and Martin Fishbein (1980), Understanding Attitudes and Predicting Social Behavior, Englewood Cliffs, NJ: Prentice-Hall, Inc.

Baumgarten, Steven A. (1975), "The Innovative Communicator in the Diffusion Process," Journal of Marketing Research, 12 (February), 12-18.

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.

Berning, Carol A. Kohn and Jacob Jacoby (1974), "Patterns of Information Acquisition in New Product Purchases," Journal of Consumer Research, 1 (September), 18-22.

Bourne, Francis S. (1957), "Group Influence in Marketing and Public Relations," in Some Applications of Behavioral Research, R. Likert and S.P. Hayes (eds.), Basle, Switzerland: UNESCO, 207-225.

Burnkrant, Robert E. and Alain Cousineau (1975), "Informational and Normative Social Influence in Buyer Behavior," Journal of Consumer Research, 2 (December), 206-215.

Carlson, Les and Sanford L Grossbart (1984), 'Toward a Better Understanding of Inherent Innovativeness," in 1984 AMA Educators' Proceedings, Belk et al. (eds.), Chicago: American Marketing Association, 88-91.

Coney, Kenneth A. and Robert R. Harmon (1975), "Dogmatism and Innovation: A Situation Perspective," in Advances in Consumer Research, Vol. 6, W.L. Wilkie (et.), Miami: Association for Consumer Research, 118-121.

Darden, William R. and Fret D. Reynolds (1974), "Backward Profiling of Male Innovators," Journal of Marketing Research, 11 (February), 79-85.

Deutsch, M. and Harold B. Gerard (1955), "A Study of Normative and Social Influences Upon Individual Judgment," Journal of Abnormal and Social Psychology, 51, 624-636.

Draper, N. and H. Smith (1966), Applied Regression Analysis, New York: John Wiley, Inc.

Gatignon, Hubert and Thomas S. Robertson (1985), "A Propositional Inventory for New Diffusion Research," Journal of Consumer Research, 11 (March), 849-867.

Joachimsthaler, Erich A and John L. Lastovicka (1984), "Optimum Stimulation Level-Exploratory Behavior Models," Journal _of Consumer Research, 11 (December), 830-835.

Joreskog, K.G. and D. Sorbom (1983), LISREL: Analysis of Linear Structural Relationships by the Method of Maximum Likelihood, Version VI, Chicago: National Educational Resources, Inc.

Kelman, H.C. (1961), "Process of Opinion Change," Public Opinion Quarterly, 25, 57-78.

Midgley, David F. and Grahame R. Dowling (1978), "Innovativeness: The Concept and Its Measurement," Journal of Consumer Research, 4 (March), 229-242.

Miniard, Paul W. and Joel B. Cohen (1981), "An Examination of the Fishbein-Ajzen Behavioral Intentions Model's Concepts and Measures," Journal of Experimental Social Psychology, 17, 309-339.

Miniard, Paul W. and Joel B. Cohen (1983), "Modeling Personal and Normative Influences on Behavior," Journal of Consumer Research, 10 (September), 169-180.

Oliver, Richard L. and William 0. Bearden (1985), "Crossover Effects in the Theory of Reasoned Action: A Moderating Influence Attempt," Journal of Consumer Research, forthcoming.

Raju, P.S. (1980), "Optimum Stimulation Level: Its Relationship to Personality Demographics, and Exploratory Behavior," Journal of Consumer Research, 7 (December), 272-282.

Rogers, Everett M. (1976), "New Product Adoption and Diffusion," Journal of Consumer Research, 2 (March), 290-301

Rogers, Everett M. (1983), Diffusion of Innovation, 3rd edition, New York: The Free Press.

Rosenberg, M. (1965), Society and the Adolescent Self- Image, Princeton, NJ: Princeton University Press.

Ryan, Michael J. (1982), "Behavioral Intention Formation: The Interdependency of Attitudinal and Social Influence Variables," Journal of Consumer Research, 9 (December), 263-278

Shimp, Terence A. and Alican Kavas (1984), "The Theory of Reasoned Action Applied to Coupon Usage," Journal of Consumer Research, 11 (December), 795-809.

Troldahl, V.C. and F.A. Powell (1965), "A Short-Form Dogmatism Scale for Use in Field Studies," Social Forces, 44, 211-214.

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