Learning from text in a foreign language: effects of text and knowledge

Abstract

Learning from a text in English by Turkish participants was examined. To observe the effect of text coherence on learning, participants read either the original or the revised version of the text. The revision was created by Britton and Gulgoz (1991) based on the Kintsch and vanDijk (1978) model. Participants who had no prior knowledge on text topic were assigned to one of two conditions. In one condition, they were provided with background information before reading the text and in the other such information was not presented Learning was measured by a multiple-choice test and a cognitive structure test. Providing background information created a difference in background knowledge but had no effect on learning from text. Multiple-choice test results were affected by text version such that readers of the revised version answered more questions correctly. Cognitive structure test showed learning in all groups but no difference between any of the groups. The results are discussed from the perspective of the model results of prior research, and problems regarding measurement of learning.