Communication à un séminaire :

In this paper we utilise paradata relating to the response latency as a measure of the cognitive effort invested by respondents in self-administered online stated preference surveys. While the effects of response latency have been previously explored, this paper proposes a different approach. Specifically, we attempt to disentangle preference, variance and processing heterogeneity and explore whether response latency helps to explain these three types of heterogeneity. To test our methodology we use stated choice data collected via an online survey to establish consumers’ preferences for various food attributes. Results from our analysis reinforce that response latency has a bearing on the estimates of error variance and the utility coefficients. Our findings raise concerns about the appropriateness of assuming deterministic choice sets, and we also show a link between response latency and the consideration sets that were actually used by respondents. We further observe that the manner in which response latency is accommodated has implications for willingness to pay estimation. Importantly, results in this paper draw attention for the need to better explain the variations that exist among respondents, in terms of preferences, error variance and processing strategies, and that only focusing on one (or two) of these is likely to lead to erroneous inferences. Key words: choice experiments, willingness to pay, food choice, online surveys, paradata, response latency, scale-adjusted latent class, independent availability logit.