Currently I am writing my thesis about consumers' preferences towards a store when buying health and beauty care products. I have included 6 attributes: distance, price level, promotions, assortment, knowledge of personnel and store layout.
I have also added some demographic questions like age and gender. Furthermore, I have one moderating variable in my model, which is whether the consumer is a small or large basket shopper (classified based on a certain amount that they spend per store trip).
So I also want to investigate whether the attribute importances and preferred levels differ for small versus large basket shoppers, but how can I do this in Sawtooth?
One thing I was thinking about was creating a filter, so I can select only the small basket shoppers or only the large basket shoppers when doing an analysis. However, is this sufficient? Because it doesn't indicate whether the moderating variable is significant at all, so where could I check this? And which analysis is better for this: Latent Class or Hierarchical Bayes?
Hope my question is clear!