Test Of Rater Accuracy In Peer Assessment Of English-Chinese Chemical Translation Project In The Process Of Translation Quality Assessment

Assessment is an important part of translation pedagogy,including both process assessment of translation competence acquisition and final translation quality assessment.At present,as an essential process evaluation tool for translation teaching,peer assessment has been widely used in practice.However,from the perspective of research content,although it has a great impact on translation teaching,its effectiveness has not received much empirical support.In addition,translation quality in authentic translation projects is mostly assessed by teachers,while experts and clients as raters are seldom considered.And in terms of the research methods,the factors of research tools and data analysis have yet to be further supplemented and improved with the development of interdisciplinary research.Therefore,this thesis combines the evaluator and translation quality assessment to study the effectiveness of peer assessment in authentic English-Chinese translation projects,and explores rating scales and models so as to provide guidance and support for students throughout their learning process.This thesis has conducted an experiment focusing on evaluating the translation quality of 38 students majoring in chemical engineering by applying the translation quality assessment model in the context of authentic translation project including peer assessment(PA),expert assessment(EA)and client assessment(CA).After qualitative and quantitative analysis through the Multi-facet Rasch model,questionnaire and semistructure interview,this thesis intends to answer the following three questions:(1)How did peer raters differ from experts and clients in terms of rating severity and consistency?(2)Are the rating scales in this thesis accurate for testing translation quality?(3)What is the model of the translation quality assessment in authentic translation project? What are the attitudes of students,experts and clients towards peer assessment and the model?Results show that,firstly,peer assessment appears to be severer than the other two methods,and expert assessment is more stringent than client assessment.Tests by the Rasch model prove that the three scoring methods have no significant differences.Secondly,among the rating scales designed in this thesis,accuracy is more difficult to be evaluated than information completeness and fluency.Peer assessment,expert assessment and client assessment show great consistency under the guidance of rating scales.Thirdly,more than half of the student translators have a general attitude toward peer assessment,and their enthusiasm for expert assessment and client assessment far exceeds that for peer assessment.Moreover,experts and clients actively believe that peer assessment can be added to the translation quality assessment model as an auxiliary assessment method.