Conducting conjoint analysis in Python
In ratings-based conjoint analysis, standard OLS regression is used. The customers ratings of the various choices form the dependent variable of the OLS regression. Moreover, the respective level of each attribute associated with each rating is captured using dummy variables. These dummy variables form the independent variables of the OLS regressions. For each attribute included, there will be x - 1 dummy variables, where x is the total number of attribute levels.
In summary, the OLS regression model is as follows:
U(P) = α 0 + ∑ j=1 kj ∑ i=1 m β ij X ij + ϵ ij
Where:
- P is a product to be evaluated
- U(P) is the utility of product profile P
- β ij is the utility of attribute i at level j
- m is the number of attributes
- k j is the number of levels for attribute j
- X ij is a dummy variable that is equal to 1 if attribute...
