In dichotic listening, two different signals are played to a subject separately to the left and the right ear at the same time. The subject's reports of what he or she has understood are recorded and the efficiency of one or the other ear is expressed as a laterality quotient called "ear advantage." In this study, using consonant-vowel syllable pairs as stimuli, the extent to which each syllable in a pair gains against the other during recognition was measured. As the results show that this so-called stimulus-dominance effect is strong, a method is proposed to unbias the results. The method rests on the assumption that when the same recording is replayed to the subject once again with the headphones reversed, and the response to a particular syllable pair in both runs is in favor of syllable and not side, the response can be qualified as stimulus biased and discarded.