How to Select the Regression Type for Driver Analysis

There are lots of potential outcome variables (the dependent variable you are predicting) that can be used when conducting a driver analysis, from five-point rating scales through to utilities from conjoint studies. To conduct a valid driver analysis, we need to select an appropriate regression type or generalized linear model (GLM), which is consistent with our data.

This article describes how to select the Regression Type suitable for your Driver Analysis.

Requirements

A Driver Analysis output, see How to Do Driver Analysis
Any scaled of categorical variables used should be coded in ascending order - bad to good.

Method

The Regression Type required will depend on the outcome variable used in your Regression model. Hover over the variable in the Data > Linear Regression > Outcome field and click the go-to arrow to go directly to the variable in the Data Sources tree.
With the variable selected in the Data Sources tree, in Properties notice the Data > Attributes > Structure of the variable. This is a good indicator of whether it is categorical, binary, or numeric.
Click Data > Attributes > Values & Labels, Values, or Categories button (depending on the variable's structure) to ensure the appropriate categories and values are included in the analyses and to confirm the underlying value used for each.

Go to Data > Regression Type and select the appropriate model based on your Outcome variable in the table below:

Outcome variable	Example	Regression Type
Two categories	1. Yes / Selected 2. No / Not selected	Binary logit (also known as logistic regression)
Three to 11 ordered categories	1. Hate 2. Dislike 3. Acceptable 4. Like 5. Love	Ordered logit
12 or more ordered categories	How would you rate your happiness on a scale of 0 to 100	Linear regression
Net Promoter Score (NPS)	-100: Detractor 0: Passive/Neutral 100: Promoter	Linear regression
Purchase or usage quantities	Number of cans of Coke consumed per week	NBD (or, if you get a weird message, quasi-Poisson regression)
Utilities from a conjoint or MaxDiff study	That is, a variable that contains the estimated utilities for each respondent	Linear regression
Probabilities or shares from conjoint and MaxDiff studies	That is, a variable that contains the probabilities for each respondent	Use the utilities instead of the preference shares, and then use Linear regression.