How do Respondents Handle Stated Choice Experiments? – Information processing strategies under varying information load

The popularity of stated choice (SC) experiments has spurned a large number of design strategies within which to study choice behaviour. When the amount of information provided increases, we often wonder how an individual handles such information in making a choice. Defining the amount of information (or ‘complexity’) as the product of the number of attributes and number of alternatives associated with each choice set, we investigate how this information is processed as we vary the amount of information. Four ordered heterogeneous logit and mixed logit models are developed, each for a fixed–attribute design, in which the dependent variable is the difference between the maximum (fixed) number of attributes in the design and the actual number that were maintained by the respondent in their information processing strategy (IPS). We have found that individuals adopt a range of ‘coping’ or editing strategies that are consistent with how we normally process information in real markets. Importantly, we should not argue that more information is necessarily undesirable; indeed such information may be necessary to give meaning (i.e., relevancy) to a choice context even if an individual invokes an IPS that involves excluding specific attributes and even aggregating them. That is, individuals invoke procedural strategies in the form of rules that they draw on as useful devices to process information in real or hypothetical markets. Indeed aggregating does not imply that we should provide the aggregated attribute in the design, but rather that this information is often useful (it is not ignored), and a respondent prefers to be aware of it and add it up in the processing of the SC experiment. This should not be seen necessarily as cognitive burden – indeed limited information may in itself be especially burdensome where it is an incomplete representation of the attribute space that matters to an individual. The evidence suggests that aligning ‘choice complexity’ with the amount of information to process is misleading. Relevancy is what matter