文章以上海对外经贸大学会计学院2014-2015学年第一学期所开设课程的教学班、教师及参评学生作为样本,分别测算课程内一致性、学生间可靠性和项目间可靠性指标。实证结果显示,由于"随意型"参评学生的"打分惯性"行为导致"课程内一致性"和"项目间可靠性"指标发生向上扭曲,而"学生间可靠性"指标发生向下扭曲。此外,评教问卷不对评教子项目加以适当归类,导致"项目间可靠性"指标进一步膨胀。由于学生评教初始结果受到不成熟参评学生及问卷设计的影响,所含的信息噪音较大。
This study estimates the within- group inter- rater agreement,the inter- rater reliability and the internal consistency reliability of the student ratings sample of instructors and students from classes of the Accounting School of the A University in Shanghai in the first semester of the academic year 2014- 2015. We find that the 'momentum'rating conducted by the 'nonsense'students would lead to an upward bias in the within- group inter- rater agreement and the internal consistency reliability,and also a downward bias in the inter- rater reliability. Besides,the pooling of ungrouped items would additionally inflate the internal consistency reliability. The result of the student ratings could be assessed as it contains too much noise due to the immature raters and questionnaires.