It is shown that results of categorical judgement experiments cannot simultaneously conform to a single process signal detectability model and to the recently proposed two process attention band model. Since there is substantial evidence for the Thurstone case V version of the signal detectability model for categorical judgement of prothetic continua, it is argued that the attention band model cannot apply to such judgements. Methods are presented for calculating all the parameters of the attention band model from a single categorical judgement task where the physical stimuli are widely separated.