Using Log Linear Models to Examine the Relationship Between Purchase Influencer and Influence

ABSTRACT - In this paper we illustrate the use of log linear models for analyzing the relationship between opinion leaders and seekers. -We show how these models can be used to test three general hypotheses about this relationship. We also illustrate how combinations of these hypotheses can be tested and show how to test product-specific effects on the relationship. To illustrate the use of these models we analyze a classic data set due to Katz and Lazarsfeld (1955). Analyses suggest that in the purchase of common household products, consumers seek advice from individuals of the same social status. For fashion products, more than one model adequately describes the relationship between opinion leader and seeker. While the small dimensions of the table limit our ability to analyze product-specific effects, the models discussed provide a richer interpretation than simple chi-square tests.

Citation:

Lawrence F. Feick and Jo Ann Novak (1986) ,"Using Log Linear Models to Examine the Relationship Between Purchase Influencer and Influence", in NA - Advances in Consumer Research Volume 13, eds. Richard J. Lutz, Provo, UT : Association for Consumer Research, Pages: 664.

Advances in Consumer Research Volume 13, 1986      Page 664

USING LOG LINEAR MODELS TO EXAMINE THE RELATIONSHIP BETWEEN PURCHASE INFLUENCER AND INFLUENCE

Lawrence F. Feick, University of Pittsburgh

Jo Ann Novak, University of Pittsburgh

ABSTRACT -

In this paper we illustrate the use of log linear models for analyzing the relationship between opinion leaders and seekers. -We show how these models can be used to test three general hypotheses about this relationship. We also illustrate how combinations of these hypotheses can be tested and show how to test product-specific effects on the relationship. To illustrate the use of these models we analyze a classic data set due to Katz and Lazarsfeld (1955). Analyses suggest that in the purchase of common household products, consumers seek advice from individuals of the same social status. For fashion products, more than one model adequately describes the relationship between opinion leader and seeker. While the small dimensions of the table limit our ability to analyze product-specific effects, the models discussed provide a richer interpretation than simple chi-square tests.

For further information, write to:

Lawrence F. Feick / Graduate School of Business / University of Pittsburgh / Pittsburgh, PA 15260

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