Effect Of Rejoinder Type In Discourse Completion Test On Collecting Data

Data collection methods are of great importance in speech act studies. The results and quality of the study are very much determined by the appropriateness of data collection methods. Upon review of the frequently used data collection methods and their validation studies in pragmatics studies, especially those on different formats of Discourse Completion Test (DCT), the author puts the focus on how rejoinder type affects DCT responses on the speech act of request.The thesis compared request responses from three types of DCT items on a series of analytical measures. In analyzing the responses supplied by subjects, the coding scheme in Cross-Culture Speech Act Realization Project (CCSARP) which the analysis of data was based on was refined to adapt to the features Chinese English learners have displayed. Request responses were compared and analyzed on the measure of level of directness, internal modification, external modification, alerter, request perspective and entire length of the response. The results revealed that rejoinder type didn't have significant influence on the responses to request situations for the subjects though some minor differences were found on the measure of internal modification, external modification and the length of request response. Like previous studies of the same kind, the study again raised the issue regarding DCT construction. Firstly, it seems unnatural to provide hearer response before subjects give their responses for pre-event speech acts such as request, but one has to admit that recipient design reflects respondents'expectancies toward the possible treatment on the upcoming request. Secondly, how subjects interpret rejoinders is also very crucial: whether interpret no rejoinder as silence, no hearer or whatever; or whether subjects view the second pair part as the end of the exchange or a part of a longer conversation. Thirdly, it is possible that the intrinsic properties of a particular speech act, contextual features, rejoinder type, subject variation and other elements are interacting in a complicated way which is so far unapproachable. Fourthly, the linguistic realization of situational content also influences subjects'speech act production. Thus it should be paid more attention, especially when non-native speakers are involved because they may not well interpret the situation and in turn influences the potential performance of the required speech act.It is suggested that cross-cultural and intracultural speech act studies which rely on DCT is better to use a long DCT questionnaire where items containing different rejoinder types are included in a counterbalanced way to decrease possible instrument effect and increase comparability of different studies. To avoid possible exhaust in completing a long questionnaire and assure the quality of DCT responses, it is recommended to split the long questionnaire evenly into short ones and administrate them at a fixed time interval.