Consistently inconsistent: The role of certainty, acceptability and scale in choice

Recently emerging in the stated preference literature as methods for better representing behaviour are choice certainty calibration and alternative acceptability. This paper finds that the amount of idiosyncratic error in the context of automobile choice is significant and can be explained by choice task certainty; which is a function of several respondent characteristics. However, it also finds that no theoretical framework exists for how these techniques should be applied and that econometric differences may be responsible for improvements in model fit, rather than better behavioural representation. Raising several questions about certainty indexing, researchers are advised to apply such methods cautiously.