%0 Journal Article
%T 关于影响托福写作自动化阅卷工具E-Rater评分因素的实证研究
Factors That Affect the Automated Scoring Engine E-Rater in the TOEFL Writing Section: An Empirical Study
%A 李丰贤
%J Overseas English Testing: Pedagogy and Research
%P 104-118
%@ 2643-5470
%D 2021
%I Hans Publishing
%R 10.12677/OETPR.2021.33012
%X 本研究通过分析ETS和新东方合作的e-rater评分系统,通过控制变量和数据分析,探讨e-rater针对不同特征的作文评分情况,比如:1) 文章字数和文章得分的关联度;2) 文章分段和文章得分的关联度;3) 文章衔接词数量和文章得分的关联度;4) 文章错误数量和文章得分的关联度;5) 文章错误种类和文章得分的关联度;6) 文章平均词句长度和文章得分的关联度等等。
By analyzing the ETS e-rater scoring system which is now available on the New Oriental TPO website, and by analyzing the control variables and data, this study discusses the scoring performance of ETS e-rater mainly targeted at compositions with different characteristics, such as: 1) the correlation between the number of words and the essay scores; 2) the correlation between paragraph organization and the essay scores; 3) the correlation between the number of cohesive connectors and the essay scores; 4) the correlation between the number of errors and the essay scores; 5) the correla-tion between the types of error and the essay scores; 6) the correlation between the average sentence and word length and the essay scores.
%K 托福写作,自动化阅卷,电子评分,语料库,语言测试
TOEFL Writing
%K Automated Scoring
%K E-Rating
%K Corpus
%K Language Testing
%U http://www.hanspub.org/journal/PaperInformation.aspx?PaperID=44874