Designing Surveys with a View to Accommodating Heterogeneous Information Processing Strategies
S Hess, ITS University of Leeds, UK
This paper presents the findings of a study making use of a survey designed to capture and account for heterogeneity in information processing strategies.
There is a growing body of evidence that suggests that different respondents process the information they are faced with in stated choice (SC) surveys in different ways. Indeed, ample evidence in papers by Hensher and colleagues has shown that some respondents consequently ignore certain attributes, while other respondents may in fact treat similar attributes jointly rather than separately. Not accounting for this during model estimation can have major impacts on results.
Existing work has largely looked at how to identify such heterogeneity in processing strategies in the data, and how to accommodate respondents with different strategies within existing modelling frameworks. However, there has been a relative lack of work in relation to two at face value similarly important issues, namely what causes this heterogeneity in processing strategies, and what the role of survey design is in terms of influencing processing strategies as well as possibly mitigating the effects of excessive heterogeneity in such processing strategies.
The present paper first discusses the development of a recently conducted survey aimed at retrieving individual specific information processing strategies. The first part of the paper shows the very significant extent of heterogeneity in processing strategies in the data, and discusses how part of this heterogeneity can be linked to the socio-demographic characteristics of the respondents in question. The paper then discusses the impacts of this heterogeneity on model results and shows how appropriate treatment of this heterogeneity can lead to better model performance and more reliable results, in line with previous evidence. Finally, through comparison with a standard survey conducted on the same sample, the paper shows how the specific approach used in the present survey to accommodate the heterogeneity in processing strategies leads to further insights into behaviour and more reliable results than in the base approach.
Association for European Transport