Abstract
In this comment on Bronner's article "Decision style in transport mode choice" two problems are discussed. The first deals with the quality of prediction. The second concerns the invalidity of the model tests. As an alternative method of attitude assignment the use of a matching constant is suggested.