Comparison of rating-based and choice-based conjoint analysis models. A case study based on preferences for iced coffee in Norway

•Two approaches are compared: rating-based and choice-based conjoint experiments.•We tested the approaches on different product profiles of iced coffees with Norwegian consumers.•Product profiles varied in coffee type, production origin, calorie content and price.•Rating data were analyzed by Mixed Model ANOVA while choice data were analyzed by Mixed Logit Model.•Approaches are compared in terms of data analysis, outcomes, estimation power and practicalities.

The authors compare two conjoint analysis approaches eliciting consumer preferences among different product profiles of iced coffees in Norway: rating-based and choice-based conjoint experiments. In the conjoint experiments, stimuli were presented in the form of mock-up pictures of iced coffees varying in coffee type, production origin, calorie content and price, following an orthogonal design. One group of participants (n = 101) performed a rating task of 12 iced coffees whereas another group (n = 102) performed a choice task on 20 iced coffees presented in eight triads. Then, all participants performed self-explicated rating and ranking evaluations of the iced coffee attributes. The rating data were analyzed by a Mixed Model ANOVA while the choice data were analyzed by a Mixed Logit Model. Both models include conjoint factors, demographic variables and their interactions. Results show that the two approaches share similar main results, where consumers prefer low calorie and low price iced coffee products. However, additional effects are detected within each of the two approaches. Further, self-explicated measures indicate that coffee type is the primary attribute for consumers' selection of iced coffee. The two conjoint approaches are compared and discussed in terms of experimental designs, data analysis methodologies, outcomes, user-friendliness of the results interpretation, estimation power and practical issues.