On the Choice of Strategies in the Identification of Spoken Words Mixed with Noise

Abstract

When subjects are given a word set of one or two alternatives before presentation of a word stimulus from that set, the probability of correctly specifying the word in noise is consistently higher than when the word set becomes known after the presentation. in a pre-cueing situation, when the task is to identify the stimulus as being either one word which is given or any other word in the language, but not given, the subject will make a correct identification more often than when the task is to identify the stimulus as being either one of two given alternatives. The reason for a difference in this direction is posited as being the acceptance of different listening strategies. The opposite is true with post-cueing, and the assertion is that similar strategies were used.