Effects of Response Format in Spoken Word Recognition Tests: Speech Intelligibility Scores Obtained from Open-Set, Closed-Set, and Delayed Response Tasks1

In studies of spoken word recognition, factors such as the materials to be recognized and the degradation of the speech signal have been discussed at length in the literature. The factor that has been largely ignored, however, is that of response format. In particular, word recognition studies in the last forty years have been largely closed-set tasks, despite the fact that the original tests used by military personnel for testing radio equipment during World War II were designed to be open-set. Recent findings suggest that open-set and closed-set tests may differ in more than just their level of chance performance. Sommers, Kirk and Pisoni (1997) found that lexical competition and talker variability do not produce reliable effects on word recognition scores in closed-set tasks, but they do produce robust effects on performance in open-set tasks. The present study was designed to replicate the earlier findings reported by Sommers et al. and explore the response demands in a modified closed-set task. Specifically, one group of listeners participated in a closed-set word recognition task in which the response alternatives were not provided until one second after the offset of the signal. Their performance was compared to listeners in an open-set task and a traditional closed-set task. Results revealed effects of lexical competition only for the listeners in the open-set task. We assume that in open-set tests, the listener must access the lexicon. These findings therefore suggest that even a delay of one second is not adequate to force the listener to process the stimuli in a way that is similar to open-set processing. The major implication of these results is that even modified closed-set tasks cannot replace open-set tasks as valid measures of word recognition performance.