论文部分内容阅读
本文报道一项应用Rasch模型分析英语写作评分行为的研究。研究旨在通过应用Rasch模型的计算机软件FACETS分析4位评分员的评分行为。分析显示,评分员之间总体宽严度把握有差异,而且差距较大。评分员自身评分的一致性很难完全达到,并且都出现个案的偏差。结果表明,为了给教学提供正确的反馈信息,提高评分的公正性,缩小偏差,应制定明确、清晰的评分标准,对评分员进行培训以保证正确理解和应用评分标准,并有必要在评分前组织试评。
This article reports a study of the use of Rasch model analysis of English writing grading behavior. The study aims to analyze the scoring behavior of four raters using FACETS, a computer software that uses the Rasch model. The analysis shows that there is a difference between the overall width and severity of the graders, and the gap is large. The consistency of the scorers’ own ratings can hardly be fully achieved, and there are individual case deviations. The results show that in order to provide correct feedback to teaching, improve the impartiality of scoring and reduce the deviation, a clear and clear scoring standard should be established, the scoring system should be trained to ensure the correct understanding and application of the scoring standard. Organizational evaluation.