Consumer preferences for internet services: a choice-based conjoint study

Using choice-based conjoint analysis, the researcher investigated consumers preferences for Internet services. Price, speed of connection, brand name, software applications, and support were the Internet service attributes in this study. Speed of connection was found to be the most important attribute affecting consumers choices. The regression approach and the nested approach of using covariates in latent class analysis were found to be equivalent in model fit and prediction accuracy measures. The latent class model and the mixed logit model were also found to be equivalent in model fit and prediction accuracy.