| The present research examined answer selection within the framework of the QUEST model of question answering proposed by Graesser and Franklin. Specifically, the present study tested several hypotheses concerning the time course of answer activation during question answering. Participants in the present study read short expository passages followed by a series of prime-target pairs. After reading a prime statement (a sentence drawn from one of the passages), the participants performed a timed word-naming task to a target word drawn from a theoretically legal or illegal answer. This word-naming task was performed at two SOA intervals. In general, shorter word-naming latencies were found when the prime sentence was followed by a question stem, a finding that is consistent with the legal arc search assumed by the QUEST model. However, the results failed to conform to the expectations of any of the proposed hypotheses with regard to the time course of this activation. These results were discussed in the context of research involving question answering and the QUEST model. |
