Opinion Leadership, Enduring Involvement and Characteristics of Opinion Leaders: a Moderating Or Mediating Relationship?

Meera P. Venkatraman, Boston University
ABSTRACT - Recently it has been recognized that enduring involvement plays an important role in opinion leadership. However, the specific nature of this relationship has largely been ignored. Previous research suggests two different relationships; moderating and mediating, between enduring involvement and opinion leadership. This paper uses the Baron and Kenny framework to determine whether the relationship is a moderating or mediating one. Based on the movie viewing behaviors of 317 undergraduate students, it finds that opinion leadership mediates the relationship between enduring involvement and activities such as information sharing.
[ to cite ]:
Meera P. Venkatraman (1990) ,"Opinion Leadership, Enduring Involvement and Characteristics of Opinion Leaders: a Moderating Or Mediating Relationship?", in NA - Advances in Consumer Research Volume 17, eds. Marvin E. Goldberg, Gerald Gorn, and Richard W. Pollay, Provo, UT : Association for Consumer Research, Pages: 60-67.

Advances in Consumer Research Volume 17, 1990      Pages 60-67

OPINION LEADERSHIP, ENDURING INVOLVEMENT AND CHARACTERISTICS OF OPINION LEADERS: A MODERATING OR MEDIATING RELATIONSHIP?

Meera P. Venkatraman, Boston University

[I would like to acknowledge the helpful comments of Terry Childers and N. Venkatraman on an earlier version of this paper. I am also very grateful to Dick Harmer for his extensive help with measure development and data collection.]

ABSTRACT -

Recently it has been recognized that enduring involvement plays an important role in opinion leadership. However, the specific nature of this relationship has largely been ignored. Previous research suggests two different relationships; moderating and mediating, between enduring involvement and opinion leadership. This paper uses the Baron and Kenny framework to determine whether the relationship is a moderating or mediating one. Based on the movie viewing behaviors of 317 undergraduate students, it finds that opinion leadership mediates the relationship between enduring involvement and activities such as information sharing.

INTRODUCTION

A significant body of research has examined the relationship between the opinion leadership scale (Childers 1986; King and Summers 1970; Rogers and Cartano 1962) which essentially measures the trait of opinion leadership, and specific manifestations of this trait such as product class knowledge, influence, information sharing and innovative behavior (Feick and Price 1987; Robertson, Zielinski and Ward 1984). Studies have found that scores on the opinion leadership scale have significant positive relationships with characteristics of opinion leaders such as knowledge and information sharing (e.g. Summers 1970) and innovative behavior (e.g. Robertson and Myers 1969). These findings suggest that the trait of opinion leadership motivates activities such as information sharing, but the question is what drives opinion leadership?

There is growing evidence that enduring involvement which is the "on-going concern with a product that transcends situational influences" (Richins and Bloch 1986; p. 280) plays a critical role in opinion leadership (Bloch and Richins 1983; Feick and Price 1987; Richins and Root-Shaffer 1988). However, the nature of enduring involvement's relationship with opinion leadership and opinion leader characteristics is not clearly understood. On one hand, based on the involvement literature, it is argued that involvement is a moderator (Gatignon and Robertson 1985; Laurent and Kapferer 1985), or it is expected that the relationship between opinion leadership and its characteristics is stronger for the people who are high on enduring involvement as compared to people who are not as enduringly involved (Refer Figure 1). On the other hand, it is argued that an opinion leader's enduring involvement with a product class may drive their desire for knowledge, information sharing and innovative behavior (Bloch and Richins 1983; Feick and Price 1987; Richins and Root-Shaffer 1988). In other words, enduring involvement may motivate opinion leadership which in turn results in information sharing and innovative behavior, which means that opinion leadership mediates the relationship between enduring involvement and_ opinion leader characteristics (Refer Figure 2).

Although moderating and mediating relationships have often been confused with each other, they are very different. According to Baron and Kenny (1986), a moderating relationship helps us understand when certain effects occur, and a mediating relationship helps us understand how and why such effects occur. If we find that the relationship between opinion leadership and its characteristics is moderated by enduring involvement and is stronger for people who are high on enduring involvement, then this finding has practical significance. It means that to effectively and efficiently influence the opinion leadership process, it is critical to identify people who are high on enduring involvement and concentrate marketing communication and information to this segment. This communication is likely to stimulate opinion leadership activities amongst those enduringly involved consumers who have opinion leader tendencies.

On the other hand, if the relationship between enduring involvement and opinion leadership is a mediating one - enduring involvement is a motivating force that drives opinion leadership activities - then we have an explanation that helps us understand why opinion leaders manifest the behaviors associated with opinion leadership. Although the practical implications of this finding are less immediate as compared to moderating relationships, it is important due to its contribution to theory building efforts in the area of opinion leadership. In summary, the moderating and mediating relationship between enduring involvement, opinion leadership and opinion leader characteristics have different implications and the objective of this paper is to determine which model, moderating or mediating, best fits the relationship.

LITERATURE REVIEW

Moderating versus Mediating Relationships

Moderating role of enduring involvement. A moderator is a variable that affects the form and/or strength of the relation between an independent or predictor variable and a dependent or criterion variable (Arnold 1982; Baron and Kenny 1986). Moderation implies that the relation between two variables changes as a function of a moderator variable. Some research suggests that in the opinion leadership context the relationship between opinion leadership and its characteristics will be stronger for consumers who are high on enduring involvement as compared to those who are not, or enduring involvement moderates the relationship between opinion leadership and its characteristics (Bloch, Sherrell, and Ridgway 1986; Richins and Bloch 1986).

FIGURE 1

MODERATING MODEL

FIGURE 2

MEDIATING MODEL

People who are high on enduring involvement for a product class seek information about this product class on an on-going basis (Richins and Bloch 1986), - their information seeking activities are not restricted to purchase related search, they are continuously seeking information from ads, magazines and friends. Due to this information seeking they gain considerable information and knowledge about the product and may also become aware of new products relatively earlier than others. Due to their ongoing concern with the product class, they are likely to buy these new products and therefore exhibit adoptive behavior (Bloch, Sherrell and Ridgway 1986). As compared to the highly enduring involved consumers, consumers who are low on enduring involvement are less likely to: (1) seek information on an on-going basis; (2) engage in product related conversations; and (3) buy new products relatively earlier than others. Therefore, it follows that for consumers who are high on enduring involvement there may be a strong positive relationship between the trait of opinion leadership and its characteristics, while for those who are low on enduring involvement the relationship may be weaker. In other words, enduring involvement may moderate the relationship between opinion leadership and its characteristics.

Mediating role of opinion leadership. A mediator is a variable that accounts for the relationship between a predictor and a criterion variable, in other words it is a mechanism through which the independent variable is able to influence the criterion or dependent variable of interest (Baron and Kenny 1986). In the opinion leadership context, it has been suggested that enduring involvement drives opinion leadership, which in turn results in opinion leader activities such as information sharing and innovative behavior (Dichter 1966; Feick and Price 1987; Richins and Root-Shaffer 1988).

Corey (1971) proposes that the unique involvement of opinion leaders with a product class distinguishes them from non-leaders. Similarly, Ditcher (1966) argues that an opinion leader's involvement with a product category motivates product related conversations, knowledge and influence. According to Ditcher, this involvement may be product related which means that people may find themselves so pleased or so disappointed by a product that they simply must talk about it. Involvement may also be other related which means that product related conversations may be expressions of friendship, neighborliness or love. In support, Summers (1971) found that involvement with womens' clothing fashions is the strongest of five variable sets that influence opinion leadership. The other variables are demographics, sociological variables, personality, media exposure and attitudes and values. Also, in a study among women opinion leaders, Myers and Robertson (1972) found that interest in 12 product categories including entertaining at home and cosmetics is related to opinion leadership. Therefore, there is some empirical evidence that a strong interest in a product is related to opinion leadership and its characteristics, although much of this research focuses on constructs that are similar to enduring involvement and not enduring involvement per se.

METHODOLOGY

Sample

The sample consisted of 317 students at a leading northeastern university. The data was collected by students enrolled in an undergraduate Marketing Research class. At the beginning of the semester, each student was instructed to ask 10 nonclass students to participate in a project which involved responding to four questionnaires over the entire semester. These questionnaires focused on opinion leadership, innovativeness and other related topics with respect to movie viewing behaviors. The objective, from the students' point of view, was to create a data base that would be useful for testing relationships using analytical techniques available in SPSSX.

Extreme care was taken to ensure data quality. The students were committed to the project both because they were interested in the results and because their grade was at stake. They impressed upon the 'panel' members their responsibility to carefully fill out the questionnaires or not do so at all. The instructor checked questionnaires for nonsense answers, systematic response patterns and any other signs of deliberate falsification. The questionnaires turned in by two students were not included due to suspect data quality.

Product Class

Movies were selected as the product class of interest for several reasons. First, undergraduates' enduring involvement with movies is expected to vary over a relatively wide range. Second, movies are inexpensive enough that the chances of income confounding the results is reduced. Third, since new movies are released every week this provides ample opportunity for the purchase of this continuous new product. Lastly, word of mouth and interpersonal influence play very important roles in influencing movie-going behavior (Wall Street Journal 1984).

Measures

Childers' (1986) version of the opinion leadership scale was used to measure this characteristic. He revised the response format of the King and Summers (1970) opinion leadership scale and used constructs such as creativity/curiosity to examine its convergent and discriminant validity. He found that the revised scale has better reliability and construct validity properties as compared to the original scale. Psychometrically valid measures of other constructs of interest in this study such as enduring involvement and opinion leader characteristics for movies were not available. Consequently, scales were developed to measure constructs such as knowledge and information sharing with respect to movies. The items selected for the scales (refer Appendix), were based on previous theoretical and empirical research. Agreement with these items was measured on a five point response format where '1' is 'Strongly disagree' and '5' is 'Strongly agree.'

The scale developed to measure enduring involvement with movies is based on research by Bloch and Richins (Bloch and Richins 1983; Richins and Bloch 1986). They argue that enduring involvement is an on-going concern with a product class that transcends situational influences. This interest is deep-rooted, strong and has a hobby-like quality that sometimes is obsessional in nature. Therefore, the enduring involvement scale was designed to capture this strong interest and concern with the product class movies. The influence scale measures the extent to which the respondents influence other peoples' opinions about movies and their choice of movies. While the knowledge scale measures subjective knowledge (Brucks 1985) or a respondent's perception of their knowledge about movies and movie stars and the information sharing scale was designed to measure the extent to which respondents talk to friends about movies, discuss and listen to other people's opinions and share their opinions with others.

As regards innovative behavior, scales vary from measuring specific behaviors, such as relative time of adoption of a new product, to capturing the trait of innovativeness (Midgley and Dowling 1978). Since, this study is interested in new movie viewing behavior and not in innate innovativeness tendencies, it was decided to develop a measure that captured behavior with respect to new movies, without asking about whether or not the respondent had seen specific movies.

TABLE 1

CONSTRUCTS, SAMPLE ITEMS AND RELIABILITY COEFFICIENTS

RESULTS

Reliability Analysis

Three reliability statistics were estimated for each of the scales used in this study. These are: (1) Cronbach's alpha; (2) Goodness of Fit Index (GFI) for the confirmatory factor model; and (3) the t-values associated with individual indicators of constructs for the confirmatory factor model (Bagozzi 1980; Bentler and Bonett 1980). As reported in Table 1, the alpha values range from .71 to .80, therefore all scales satisfy Nunnally's minimum alpha criterion of .70 (Nunnally 1978). The GFI values range from .91 to .98, therefore the scales also satisfy the Bentler and Bonett (1980) Goodness of Fit criterion of .90. Finally, with the exception of 2 items, the factor loadings range from .31 to .87, and only one item (item 5 for the opinion leadership scale), has a t-value less than 4.46. In summary, the scales used in this study are reliable.

The Moderating Effect of Involvement

Arnold (1982) differentiates between variables that 'moderate the degree' and those that 'moderate the form' of a relationship between two other variables. The 'degree' of a relationship is measured by the correlation coefficient and the 'form' is measured by coefficients in the regression equation. "Comparison of correlations answers the question does X account for as much variance in Y in group E as it does in group F?' Comparison of regression coefficients answers the question 'does a change in X make the same amount of score difference in Y in group E as it does in group F?"'(Arnold 1982; p. 146).

Accordingly, if the correlations between opinion leadership and its characteristics are different between low and high enduringly involved respondents, then enduring involvement moderates the 'degree' of the relationship (Arnold 1982; p. 151). Again, if the interaction (x*z) between OL (x) and involvement (z) is a significant predictor of an opinion leader characteristic such as influence (y), in addition to the main effects of opinion leadership and enduring involvement, then enduring involvement moderates the 'form' of the relationship.

This analysis reveals that enduring involvement does not moderate the relationship between opinion leadership and each of its characteristics. As reported in Table 2, the z-statistic (Arnold 1982) that measures differences in correlation coefficients between opinion leadership and its characteristics for low and highly involved respondents is not significantly different at the p < .05 level of significance. Therefore, enduring involvement does not moderate the 'degree' of the relationship. Again the interaction between enduring involvement and opinion leadership is not a significant predictor of any of the characteristics, therefore, enduring involvement does not moderate the 'form' of the relationship studied here.

The Mediating Role of Opinion Leadership According to Baron and Kenny (1986) to establish the mediation of opinion leadership between enduring involvement and opinion leader characteristics, such as influence, the following conditions must be met:

(a) Enduring involvement should have a significant effect on opinion leadership.

(b) Enduring involvement should have a significant effect on influence.

(c) Opinion leadership should have a significant effect on influence over and above the effect of enduring involvement. The effect of enduring involvement on influence will be diminished as compared to condition b.

TABLE 2

TESTING FOR THE MODERATING ROLE OF ENDURING INVOLVEMENT

These conditions are tested for each opinion leadership characteristic separately using multiple regression analysis. Two regression equations are estimated for each mediating relationship tested. In the first, opinion leadership is regressed on enduring involvement (condition a). In the second, influence is regressed against enduring involvement and opinion leadership, which are entered in a step wise manner (condition b and c). The results of this analysis are reported in Table 3.

Condition (a) is satisfied - enduring involvement has a significant effect on opinion leadership. Condition (b) is also satisfied for all four opinion leader characteristics, - enduring involvement has a significant effect on these characteristics (p < .01). Finally, condition (c) is satisfied for influence, knowledge and information sharing, - the relationship between opinion leadership and these characteristics is significant (p < .01) when these characteristics are regressed against both enduring involvement and opinion leadership. Therefore, opinion leadership mediates the effect of enduring involvement on knowledge, influence and information sharing. However, it does not mediate the effect of enduring involvement on innovative behavior, which means that enduring involvement has a direct effect on innovative behavior.

Table 3 also reveals that as expected the effect of enduring involvement on influence, knowledge and information sharing decreases when opinion leadership is entered in the equation, although it does not become insignificant. This indicates that while opinion leadership mediates the effect of enduring involvement on these three opinion leader characteristics, it is not the only mediator. According to Baron and Kenny (1986), "from a theoretical perspective...(this) demonstrates that a given mediator is indeed a potent, albeit not both a necessary and sufficient condition for an effect to occur" (pp. 1176).

DISCUSSION

Despite the appeal of the thesis that enduring involvement moderates the relationship between opinion leadership and its characteristics, this paper finds that the relationship between enduring involvement, opinion leadership and its characteristics is a mediating one. The analysis reveals that opinion leadership mediates the relationship between enduring involvement and opinion leader characteristics of influence, knowledge, and information sharing. This supports the model -Enduring involvement-> Opinion Leadership-> Opinion Leader Behaviors - that has only recently been explicitly tested (Richins and Root-Shaffer 1988). According to this model enduring involvement drives opinion leadership, which in turn accounts for behaviors such as exerting influence, gaining knowledge and sharing information and experience.

TABLE 3

TESTING FOR TEH MEDIATING ROLE OF OPINION LEADERSHIP

The analysis also shows that opinion leadership is not the only mediator. Other personal constructs that may mediate the relationship between enduring involvement and behaviors such as influence and knowledge are personal experience with the product, extent of information seeking activities especially ongoing information seeking behaviors, and extent of media exposure. Finally, the analysis reveals that opinion leadership does not mediate the relationship between enduring involvement and innovative behavior. This means that as suggested in the innovative behavior literature (Gatignon and Robertson 1985), enduring involvement with a product class has a strong direct effect on influencing purchase of a new product in that product class.

LIMITATIONS AND EXTENSIONS

Conclusions about the mediating role of opinion leadership must be tempered by the limitations of this study. One limitation is its use of a student sample. This study uses students due to the relatively low cost of obtaining cooperation from a relatively large sample of respondents. However, the student sample has obvious demographic differences from the general population and may also differ from it on some of the constructs used in this study such as innovative behavior. This suggests the need to replicate the study with other populations of interest.

Another limitation of the study is its use of previously unvalidated scales to measure constructs such as influence and information sharing. Consequently, this study needs to be replicated using both measures developed here and other measures in order to assess the robustness of its findings.

A third limitation is its use of a single product class to examine the moderating/mediating nature of the relationship between opinion leadership, enduring involvement and other characteristics of opinion leaders. This study should be extended to other products, both where word-of-mouth communication is important such as restaurants and where word of mouth communication is not so important such as consumer consumables.

SUMMARY AND CONCLUSIONS

This paper makes several important contributions to the research on opinion leadership. It goes beyond the direct main effect relationships that have dominated opinion leadership research, to examine more complex interactive relationships that are better models of the complicated phenomena of opinion leader activities. It does so by empirically examining the relationship of enduring involvement with opinion leadership and opinion leader activities. By exploring the role of enduring involvement it tests a relationship, which although recognized in the literature, has not been given the attention it deserves especially in empirical research. Finally, it uses the Baron and Kenny (1986) framework to distinguish between moderator and mediator relationships in the context of opinion leadership, thereby introducing a framework which will have value for testing many different consumer behavior theories and frameworks.

APPENDIX

REFERENCES

Arnold, Hugh J. (1982), "Moderator Variables: A Clarification of Conceptual, Analytical and Psychometric Issues," Organization Behavior and Human Performance, (29), 143 - 174.

Bagozzi, Richard P. (1980), Causal Models in Marketing. NY: Wiley.

Baron, Reuben M. and David A. Kenny (1986), "The Moderator-Mediator Variable Distinction in Social Psychological Research: Conceptual, Strategic, and Statistical Considerations," Journal of Personality and Social Psychology, 51(6), 1173- 1182.

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

Bentler, Peter M. and D.G. Bonett (1980), "Significance Tests in the Analysis of Covariance Structures," Psychological Bulletin, (88) 588-606.

Bloch, Peter H. and Marsha L. Richins (1983), "A Theoretical Model for the Study of Product Importance Perceptions," Journal of Marketing, 47 (Summer), 69-81.

Bloch, Peter H., Daniel L. Sherrell, and Nancy M. Ridgway (1986), "Consumer Search: An Extended Framework," Journal of Consumer Research, 13 (June), 119-126.

Brucks, Merrie (1985), "The Effects of Product Class Knowledge on Information Search Behavior," Journal of Consumer Research, 12 (June), 1 -16.

Childers, Terry L. (1986), "Assessment of the Psychometric Properties of an Opinion Leadership Scale," Journal of Marketing Research, 23 (May), 184- 187.

Corey, Lawrence G. (1971), "People who Claim to be Opinion Leaders: Identifying their Characteristics by Self Report," Journal of Marketing, 35 (October), 48-53.

Dichter, Ernest (1966), "How Word of Mouth Advertising Works," Harvard Business Review, 44 (November-December), 147-166.

Feick, F. Lawrence and Linda L. Price (1987), "The Market Maven: A Diffuser of Marketplace Information, "Journal of Marketing, 51 (January), 83-98.

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

Laurent, Gilles and Jean-Noel Kapferer (1985), "Measuring Consumer Involvement Profiles," Journal of Marketing Research, (22) 41-53.

Midgley, D.F. and G.R. Dowling (1978), "Innovativeness: The Concept and its Measurement," Journal of Consumer Research, (4) 229-242.

Montgomery, David B. and A.J. Silk (1971), "Clusters of Consumer Interests and Opinion Leaders' Spheres of Influence," Journal of Marketing Research, (8) 317-321.

Myers, James H. and Thomas S. Robertson (1972), "Dimensions of Opinion Leadership," Journal of Marketing Research, 9 (February), 4146.

Nunnally, Jum C. (1978), Psychometric Theory. New York: McGraw-Hill Book Company.

Richins, Marsha L. and Peter H. Bloch (1986), "After the New Wears Off: The Temporal Context of Product Involvement," Journal of Consumer Research, 13 (September), 280-285.

Richins, Marsha L. and Teri Root-Shaffer (1988), "The Role of Involvement and Opinion Leadership in Consumer Word-of-Mouth: An Implicit Model Made Explicit," In Advances in Consumer Research, Vol. 15, Ed. Micheal Houston, Ann Arbor, MI:Association for Consumer Research.

Robertson, Thomas S., Joan Zielinski, and Scott Ward (1984), Consumer Behavior. Glenview, IL: Scott, Foresman and Company.

Rogers, Everett and David G. Cartano (1962), "Methods of Measuring Opinion Leadership," Public Opinion Quarterly, 26 (Fall), 435-41.

Sharma, Subash, Richard M. Durand and Oded Gur-Arie (1981), "Identification and Analysis of Moderator Variables," Journal of Marketing Research, 18 (August), 291-300.

Summers, John O. (1970), "The Identity of Women's Clothing Fashion Opinion Leaders," Journal of Marketing Research, 12 (May), 178-185.

Summers, John O. (1971), "Generalized Change Agents and Innovativeness," Journal of Marketing Research, 8 (August), 313-316.

Wall Street Journal (1984), "Frank Mancuso's Marketing Savvy Paves Way for Paramount's Hits," 27th (June), 27-45.

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