The authors compare the application of two logit models for the analysis of qualitative marketing data. A weighted least squares logit model is compared with a maximum likelihood logit model different from that mentioned by Green et ai. Empirical applications are used to compare the models. Suggestions are presented for interpreting and reporting the results of logit-type models, with special attention to interaction effects.
Get full access to this article
View all access options for this article.
References
1.
BerksonJ. (1955), “Maximum Likelihood and Minimum χ2 Estimates of the Logistic Function,” Journal of the American Statistical Association, 50 (March).
2.
BerksonJ. (1968), “Application of Minimum Logit χ2 Estimate to a Problem of Grizzle with a Notation on the Problem of ‘The Interaction,’” Biometrics, 24, 75–96.
3.
GreenPaul E., CarmoneFrank J., and WachspressDavid P. (1977), “On the Analysis of Qualitative Data in Marketing Research,” Journal of Marketing Research, 14 (February), 52–9.
4.
LeonardE. W. (1978), “Life Style Measures and Segmentation in Retail Banking: An Empirical Examination Applied to Automated Tellers,” unpublished Ph.D. dissertation, Oklahoma State University.
5.
NerloveMarc and PressJames S. (1973), Univariate and Multivariate Log-Linear and Logistic Models. Santa Monica: Rand Corporation (December).
