ABSTRACT
The conference participants were asked for their opinions as to the most fruitful and appropriate taxonomy for human error. In general, a successful taxonomy must be related to both the theoretical and practical purposes of the investigator. The taxonomies themselves may thus be classified according to the level at which they approach the problem of classifying errors. The number of categories in proposed taxonomies varied enormously. At one end of the continuum are simple binary taxonomies. At least two of the taxonomies proposed were in fact taxonomies of behavior. The first of these is Rasmussen’s taxonomy. The second is Moray’s modification of Altman’s taxonomy. A particularly popular binary taxonomy distinguishes between slips and mistakes as forms of error. A more problematic issue is the extent to which the taxonomy is useful only for analysis of errors after the event; while it is valuable for research and theory building, it cannot be used to predict the imminent occurrence of errors.
