ABSTRACTSimple ranking and rating procedures for hedonic testing were developed from which signal detection analysis gave simple probability values indicating preference. Data were pooled over judges to give preference scores for specific groups of consumers. Rating was preferable when forced choice was to be avoided; ranking was preferable when response bias was a consideration. The rating procedure was less prone to instructional bias than the 9‐point hedonic scale. Ranking methods were extended to measure‘likelihood to buy', when price was included in the decision, as well as degree of prefere