Data Analysis and Estimation
Christensen Associates uses a variety of methods for analyzing survey response data sets. We tabulate and summarize responses to all forms of questions, with segmentations by sub-samples where applicable. For choice-based conjoint questions, we have expertise in many analysis techniques, including:
- logit
- probit
- nested logit
- mixed logit with conditioning of individual tastes
- hierarchical Bayes.
Customer segmentation can be important in some applications. For example, you may want to know which customers would purchase a product in order to target a marketing campaign towards those customers. Christensen Associates has several methods for exploring customer segmentation issues:
- segmentation of logit, probit, or nested logit models based on observable customer characteristics
- examination of individual respondent preferences estimated from mixed logit or hierarchical Bayes models according to observable customer characteristics
- formal cluster analysis on individual preferences estimated from mixed logit or hierarchical Bayes models.
For more information, contact Dan Hansen.
©2007 Laurits R. Christensen Associates