Interferencias e importancia de la informaciónefectos de las implicaciones y de los conocimientos previos

Abstract

In this study, three hypotheses are tested to account for the inferential superiority of main sentences as compared with secondary sentences in a text. The probability of inference construction depends on (a) the number of implications in the text, (b) the presence of implications in the reader�s macrostructure on text representation, and (c) the reader�s prior knowledge. The results of the experiments indicate, first, that main sentences had a higher false rate than secondary sentences in both short and long-term retention tests, but response times were equivalent. Secondly, that text implications and prior knowledge were more likely for main sentences than for secondary ones. And, thirdly, the effect of sentence importance on false alarms was removed when differences in text implications between main and secondary sentences were controlled. These results indicate that sentences with important information are more likely to be inferred than secondary sentences because the former are implied more frequently by other information stated in the text than are the latter. The additional finding that response times for hits and false alarms were equivalent suggests that inferences occurred during reading or encoding.