Research For The Effects Of Test Media On Students’ Performance In Writing Test

With constant technological improvements of language teaching in the pastdecades, there have been great changes in the media of language testing. Now,computers have been widely used in various domestic and abroad language tests. Moreimportant, besides the original objective test, the test contents on computers havealready extended into all four basic language skills: listening, speaking, reading andwriting, which brings more concern and research to the field of computer-basedlanguage testing.In line with the above trend, this study aims to comprehensively explore the effectof test media on test-takers’ writing test from the aspect of learners through theapplication of Web-based language testing into college English teaching and on thebasis of related theories and research results in this field. To be more specific, throughthe comparison between computer-based writing test and paper-based writing test, theauthor tries to answer the following research questions:1) Are there any differences instudents’ overall scores between computer-based and paper-based writing tests?2) Arethe differences of students’ levels in computer familiarity reflected in their writing testscores?3) Do test-takers with different levels of computer familiarity performdifferently in writing processes on computer-based writing test?4）If the answer toquestion3is yes, then what are the main factors affecting test-takers’ writing processesin computer-based writing test? It is hoped that the solution to the above questions canhelp students cultivate some strategies in web-based learning and testing and providesome practical values for the development of computer-based writing test.The subjects of this research were non-English major undergraduates fromChongqing University. They took part in both the paper-based and the computer-basedwriting tests, after which the questionnaire surveys were conducted and the results wereanalyzed as follows:1) Paired sample t-test results showed there was significant meandifference in the holistic scores in the two test occasions, with the scores of paper-basedwriting test being higher than those of computer-based writing test, which indicates thatthe test media can slightly influence the students’ writing scores;2) Paired sample t-testresults showed there were significant differences among students with different levels ofcomputer familiarity in their test scores, but the differences mainly came from thosewho were not so familiar with computers, showing that the test media have greatly impacted the writing scores of those unfamiliar with computers;3) One-way ANOVAresults indicated that there were significant differences in writing processes for studentswith varied computer familiarities;4) The test-takers with varied computer familiaritieswere not significantly different in the sub-processes of “goal-setting”,“generatingideas” and “reviewing”, while their “organizing ideas” and “translating” sub-processeswere greatly different, due mainly to the disadvantages of the test media itself.