| Reading has great meaning for human beings, especially for students. There are two perspectives to judge the quality of reading activities. One is from the point of cognition and the other is from meta-cognition, the latter of which is more important in this article. According to the theory of self-regulated learning, valid reading activities must reach the following two targets:correctly understanding the meaning of articles and correctly judging the level of understanding by meta-comprehension. Only accurately judging the real level of themselves can readers correctly change their study pace and finally promote the learning outcome.Nowadays in school, traditional teaching methods are developing along with burgeoning methods, which bring speech, animation, pictures and so on into today’s classroom. And these new media are replacing the traditional writing materials. Under this tendency, reading is no longer simply activity about characters, but also having these new elements joining in. Thus, speech should be taken into account by study of reading comprehension.In the past three decades, the study of meta-comprehension has carried out in a deep-going way. Under the direction of relevant meta-comprehension theories, the factors that affect the accuracy of meta-comprehension monitoring has turned from objective stuff to subjective ones. Recent years, more and more studies related to belief and expectation have appeared in the area of meta-comprehension, however the conclusions they made have been far from consistent. So more studies should be keep on carrying out.Based on these facts, this article tried to explain whether the speech and the belief about it affected the accuracy of meta-comprehension monitoring and how these factors worked. In this article the word belief meant whether readers believed that speech had benefits to comprehension outcome. For this purpose, two experiments were designed in this study. Experiment one was two-factor completely random design. One factor was speech, which had two levels called "voice" and "silence". The other factor was belief, which had two levels called "trust" and "distrust". This experiment was designed to explore whether the speech and the belief about it affected the accuracy of meta-comprehension monitoring. Experiment two was one-factor completely random design, the only one factor had three levels, which was silence, mandarin voice and French voice. Experiment two was designed to study without the activation of belief, whether speech had any effect on comprehension and meta-comprehension.Several results emerged via two experiments above:(1) speech had no effect on the score of comprehension test;(2) without the activation of belief, speech had negative effect on meta-comprehension monitoring;(3) readers’recall test of comprehension would be promoted only if they believed the role of speech, no matter whether the speech existed or not;(4) believing the role of speech had negative effect on the accuracy of meta-comprehension monitoring, in contrast not believing the role of it had significantly higher accuracy than former one.To sum up, quiet environment had good effect on reading activity. However how to treat belief depended on different situations. If the environment was full of speech, believing the role of speech sound like a good choice. If the environment was quite peaceful, not believing the role of speech would be a better choice. Although it would decline the score of comprehension test, the accuracy of meta-comprehension monitoring would enhanced. Of course, if someone’s belief was hard to change, choosing appropriate reading surroundings would become much more important. |