Similarities And Differences In The Use Of Cohesive Devices Between The First Drafts And The Final Drafts Of College Students’ English Summary Writings

A number of studies have shown that cohesion is an indispensable device of discourse construction and one of the possible means of discourse coherence.In recent years,a multitude of scholars at home and abroad have focused on the use of local cohesion in the polished final writings,and manual calculation of cohesive devices have been adopted by the majority of such researches.These studies have improved our abilities to use cohesive devices and our understandings of their functions in writing,but few of them focused on the changes in the use of cohesive devices in the revising of writing drafts.Furthermore,as an important academic genre and writing skill,summary writing has not been paid enough attention by researches in this field.In order to fill the gap mentioned,this study attempts to investigate:(1)Are there any changes in the use of cohesion devices between the first drafts and the final drafts of summary writings? If so,what are their differences in terms of the use of cohesive devices?(2)Can cohesive devices predict writing scores in the two drafts of the summary writing respectively? If so,are predictors different from or similar to each other between the two drafts? Computational linguistic analysis tools such as Coh-Metrix 3.0 and TAACO 2.0 were used to analyze the local,text and global cohesive devices of the collected summaries in this study.The main findings of the study are as follows: to solve the first question of the research,the results revealed that most of the local,global and text cohesion were improved from the first drafts to the final drafts of the summary writings,however,significant differences could only be identified with eleven local and text cohesive devices.With regard to the second research question,the study found that although both drafts were graded according to the same scoring rubric,they actually had different predictors of the scores.Results also demonstrate that there was no global cohesion improved significantly from the first drafts to the final drafts suggesting that further instruction might still be needed to help students to use such effective devices to further enhance their writing skills.The results also suggest that in grading,even if the same scoring rubric was used in assigning scores to the summaries in grading,the instructor actually demonstrated different behaviors.Finally,the findings of the study show that the computational tools(Coh-Metrix and TAACO)can be efficiently used in related investigations.The thesis ends with limitations of current studies together with suggestions for future studies.