A Study On The Effect Of Man-Machine Collaboration Feedback On College Students’ Second Language Writing Performance

Due to the advancement of artificial intelligence,the concept of man-machine collaboration is gradually being applied in various fields.Feedback is one of the essential elements of English writing.However,the teacher feedback is insufficient for practical written teaching.The previous literature mainly focused on the comparisons of various types of feedback,and there were few on the combination of online automatic feedback and teacher feedback and its impact on specific indicators of writing performance.This thesis centers on the influence of man-machine collaboration feedback on college students’ English writing performance,and explores how to achieve man-machine collaboration in writing feedback.The following research questions are explored through comparisons of score-only,teacher feedback,online automatic feedback and man-machine collaboration feedback:(1)Do students with the three types of feedback outperform the students with score-only feedback in English writing performance?(2)Which type of feedback(teacher feedback,online automatic feedback,or man-machine collaboration feedback)is the most effective in promoting students’ English writing performance in terms of accuracy,complexity and fluency after one semester’s experiment?(3)What are the students’ attitudes toward man-machine collaboration feedback?In order to answer these questions,129 students from four non-English major sophomore classes were selected as participants in a university in Jiangxi Province.After the writer tested the differences in the four classes of the pre-test,four types of feedback were respectively adopted for them.The experiment lasted for one semester,and five writing tasks were conducted.Besides,the genre of the writing task is argumentation.After the experiment,the post-test was carried out in the four classes.At the same time,questionnaires were distributed to the experimental class before and after the test.According to Distributed Cognition Theory,man-machine collaboration feedback could combine the intelligence of man and machine,and cooperate each other to make feedback more efficiently.The findings are as follows:(1)Compared with the score-only class,the students with three types of writing feedback had significant differences in terms of complexity(C/T),accuracy,and fluency.(2)Compared with teacher feedback and online automatic feedback,man-machine collaboration feedback has advantages in promoting students’ writing performance.The experimental class with man-machine feedback outperformed other groups in syntactic complexity(C/T and CP/T)and in fluency(W/M and W/S).The accuracy and lexical complexity,however,did not differ significantly from the other two classes,which supported Skehan’s Limited Capacity Hypothesis.(3)Most of the students in the experimental class showed their acceptance and recognition of man-machine collaboration feedback from cognition,emotion and behavior,and agreed that the Pigai.org could replace the teachers’ suggestions of vocabulary and grammar.According to the Distributed Cognition Theory,man-machine collaboration feedback could combine the intelligence of man and machine,and cooperate each other to make feedback more efficiently.The implications for this research are the following: instructors need give students more effective feedback besides scores.Also,the instructors could use an online automatic platform to assist their evaluation,which can lessen their workload and motivate learners to revise.During this process,they should identify and realize its functions and roles in writing assessments.Besides,the vocabulary and grammar could be corrected by Pigai.org.And the teacher could pay more attention to the content and structure of texts.