The Influences Of Background Information Listening Input On The Chinese Examinees' Performance In Expositive Writing

While taking some new integrated testing forms such as reading-writing,and listening-writing tests, the examinees are given the opportunity to read apassage or to listen to a passage on the basis of which the examinees aresupposed to write a passage of a certain number of words. The reason whyexaminees are given a listening passage or a reading passage is that it canprovide examinees with some topic knowledge, which can, to some extent,solve the problem of examinees' poor performance merely because of the lackof topic knowledge, thus eliminating to a certain degree the threat to testvalidity arising therefrom.This paper deals mainly with the influence of cultural backgroundinformation input in the form of listening on the Chinese examinees' writingperformance, especially on their contrastive expositive writing. An analysis ofthe examinees' performances on this new testing forms and the resultsobtained respectively from an independent writing task (without the listeninginput) and from an integrated listening-writing task (with the listening input)reveals how this input of information, especially the cultural backgroundinformation, affects examinees' performances and their testing-takingstrategies, thus giving us some information as to its feasibility. The researchwas done mainly through Ultraedit-based text analysis in combination withmanual analysis, and analysis of the data collected from pre-test and post-testquestionnaires. The subjects were divided into three different groups, one thecontrast group, the other two experiment groups (one experiment group listensto the input only once while the other experiment group listens twice). All thedata collected are processed with SPSS 11.0. Results indicate that theperformances of subjects in the experiment groups differ relativelysignificantly from those who take the traditional independent writing test.Independent-Samples T-tests, One-way ANOVA analyses orNon-parametric—2-Independent-Samples Tests show significant differencesbetween the contrast group and the experiment groups in terms of text length,syntactic complexity (including average sentence length), and their use ofrelevant vocabulary and terms, but don't show a major positive effect in termsof their lexical complexity and grammatical accuracy. Instead the controlgroup did slightly better than the experiment groups and used more complexwords and had better grammar accuracy although the holist ratings of theexperiment groups are higher. Data from the questionnaires also support theidea that the listening input provides the subjects with the necessary help.Results from the questionnaire also indicate that some think that the listeninginput is difficult for them and therefore they find the writing task difficult orimpossible (only a few), thus rendering it impossible for them to obtain thenecessary relevant information and to produce the desired writing. From theabove, it can be seen that the new testing form can provide the necessaryinformation and thus eliminate the influence of the lack of topic knowledge onthe examinees only for the high-score group (and some mid-scoregroup)—those who at least have good listening comprehension ability,especially those who also have better ability in writing. Finally, this paperdiscusses the implications it may bring to language testing and teaching andsuggestions for further research to study other factors' influence onexaminees' performances and the applicability and feasibility of this newtesting form.