What Makes a Good Answer to a Question? Testing a Psychological Model of Question Answering in the Context of Narrative Text

Two experiments tested the QUEST model of question answering (Graesser & Franklin, 1990) in the context of short stories. QUEST specifies how individuals answer why and how questions by searching through various sources of information after comprehending a text. The sources of information include the passage structure and the generic knowledge structures that are associated with the content words in the query. After these information sources are activated in working memory, search components narrow down the set of relevant answers. The present study investigated three components of the model: (1) the arc-search procedures that specify those nodes and arcs within an information source that are sampled for answers; (2) the structural distance, the number of arcs on the path that connects the queried node to the answer node; and (3) the number of information sources contributing to the answer. After reading two short passages, subjects were presented with either why or how questions and a number of answers to each question. In Experiment 1, subjects judged each answer as good (appropriate and relevant) or bad (inappropriate or irrelevant). In Experiment 2, subjects rated each question-answer item on a 4-point goodness-of-answer scale. Results supported the validity of the arc-search procedures and structural distance for both question categories, but there was only partial support for number of information sources.