Two experiments are described in which reaction times for understanding target sentences or phrases in terms of a preceding context were measured. In Experiment 1, the target sentences followed either short or long contexts which induced either literal interpretations or metaphorical ones. Results indicated that only in the short context condition did subjects take significantly longer to understand metaphorical than literal targets. This interaction is explained in terms of the availability of appropriate schemata for interpreting the target. In Experiment 2, targets were phrases that could be given either an idiomatic or a literal interpretation. It was found

Thirty physical education students and 30 music education students read a passage that could be given either a prison break or a wrestling interpretation, and another passage that could be understood in terms of an evening of card playing or a rehearsal session of a woodwind ensemble.Scores on disambiguating multiple choice tests and theme-revealing disambiguations and intrusions in free recall showed striking relationships to the subject's background. These results indicate that high-level schemata provide the interpretative framework for comprehending discourse. The fact that most subjects gave each passage one distinct interpretation or another and reported being unaware of other perspectives while reading suggest that schemata can cause a person to see a message in a certain way, without even considering alternative interpretations.

No abstract

Readers were asked a question of a certain type after every four pages of a 48 page oceanography text. Text information relevant to questions was learned better than text information irrelevant to questions. Furthermore, reading times and probe reaction times on a secondary task were longer when subjects were processing text segments containing information of the type addressed by questions. A good account of these results is provided by a theory which asserts that readers selectively allocate a greater volume of attention to question-relevant information, and that a process supported by the additional attention causes more of the information to be learned. There were two operational measures of attention. The first was the amount of time a subject spent reading segments of the text. It was assumed that this measure reflects the extent or duration of attention. Influence of Questions on AttentionReading times have been collected in a number of previous question experi- , 1975). Times tend to be longer when questions are asked; however, in the early studies the effect was not very strong nor entirely consistent, partly because of crude measurement techniques, such as having subjects write the elapsed time on the bottom of each completed page.The second measure employed in the experiment reported in this paper was reaction time in a secondary task. Subjects were told that comprehending the text was their primary task. They were also told to depress a key as quickly as they could whenever a tone sounded. The idea is that when the mind is occupied with the primary task, there will be a slight delay in responding to the secondary task. The key assumption is that a person has a fixed amount of cognitive capacity. Ordinarily, there is spare capacity when a person is doing mental work such as reading. However, when a reader puts extra effort into processing a text element, this places peak load demands on the cognitive system. The assumption is that at this moment there is little capacity left over to process the probe and respond to it. Influence of Questions on Attention 4Thus, the reaction to the probe is delayed until capacity becomes available.Our working assumption is that probe time primarily reflects the intensity of the attention that a reader is devoting to a text element.The secondary task procedure has a considerable history in research with simple tasks. The rationale for the procedure and representative empirical results have been presented by Kahneman (1973) and Posner (1978) among others. The procedure was first used in research on text processing by Britton and his associates (cf. Britton, Westbrook, & Holdredge, 1978).They have completed one study on the effects of questions in which probe time was assessed, which we shall review shortly.Attention is a hypothetical construct that is imperfectly reflected in any operational measure.In a relatively uncharted area such as the processing of lengthy meaningful texts, the risk is high that extraneous factors will introduce bias or overshadow what ar...

Students often hold misconceptions that conflict with scientific explanations. Research has shown that refutation texts are effective for facilitating conceptual change in these cases (Guzzetti, et al., 1993). The process through which refutation texts have their effect is not clear. Our study replicated and extended previous research investigating cognitive processes involved in the refutation text effect. Undergraduates read either a refutation or expository text on seasonal change. Individual reading times were recorded. Participants' conceptions were measured at pre-, post-, and delayed-posttest. Results showed readers spent less time reading the refutation paragraph compared to the expository paragraph. The refutation text group had fewer misconceptions at posttest. These findings suggest that refutation text processing differences mirror similar findings in the attention literature which may account for their effectiveness.