구조방정식모델을 활용한 통역 품질 평가 항목에 관한 고찰 : 통역사 집단의 평가 항목을 중심으로
- 원문 URL
Investigating Quality Assessment Criteria in Interpreting Using SEM: Focusing on the Assessment Criteria of Professional Interpreters The purpose of this study is to investigate whether the results of interpreters' expectations survey on quality criteria match the interpreters' evaluation of a given interpreting performance. Many expectations surveys on interpreting quality criteria have shown that interpreters and users alike regard content-related criteria (sense consistency, logical cohesion, and completeness) more important than form-related criteria (correct grammar usage, use of correct terminology, appropriate style) and delivery-related criteria (fluency of delivery, native accent, pleasant voice, lively intonation). However, to see whether these criteria ratings hold true in real-world evaluations, we need to have interpreters or users evaluate actual interpreting performances and examine the effect of individual quality items on the overall performance scores. This study takes a two-pronged approach. First, 42 professional interpreters were asked in an expectation survey to rate ten linguistic items as to their significance. They were also asked to use the same ten quality criteria to evaluate seven English-to-Korean interpretations and seven Korean-to-English interpretations. Structural Equation Modeling (SEM) was then used to create interpretation evaluation models for both interpreting directions and to analyze the causal relationship between the ten quality criteria and the total performance score. Results of the expectation survey were then compared with the actual evaluation results. Results of the expectation survey are as follows. Sense consistency received the highest rating, followed by logical cohesion, completeness of interpretation, fluency of delivery, correct terminology, correct grammar usage, appropriate style, lively intonation, pleasant voice, and native accent. This result is consistent with previous studies, as content-related criteria received higher ratings than both form-related and delivery-related criteria. It was also found that for each quality item, early bilinguals had a tendency to give higher ratings than late bilinguals. Results of the interpreter's evaluations on English-to-Korean interpretations are as follows. For late bilinguals, content-related criteria, form-related criteria, and delivery-related criteria all had statistically-significant impact on the total score, with content-related criteria having the highest regression weight; however, early bilinguals regarded form-related criteria more important than content-related criteria, while delivery-related criteria had non-significant impact on the total score. It was also shown that, for female interpreters, delivery-related criteria did not have any impact on the total score; on the other hand, male interpreters' standard regression weight of delivery-related criteria was similar to that of the form-related criteria. Results of the interpreter's evaluations on Korean-to-English interpretations are as follows. For late bilinguals, content-related criteria had the highest impact on the total score followed by form-related criteria and delivery-related criteria; however, when assessing Korean-to-English interpretations, early bilinguals did not treat form-related criteria and content-related criteria as separate constructs. They adopted a rather holistic evaluation scheme where the total score would be affected by a single concept including both the content-related and the form-related criteria. For female interpreters, form-related criteria had the largest impact on the total score, while male interpreters adopted an evaluation scheme that viewed the content-related and the form related criteria as a single construct. The evaluation model of English-to-Korean interpreting and the evaluation model of Korean-to-English interpreting model differed in the categorization of the logical cohesion criterion. When evaluating English-to-Korean interpretations, interpreters regarded logical cohesion as one of the concepts that reflect students' content-related performance. Contrastingly, in evaluating Korean-to-English interpretations, interpreters viewed the logical cohesion criterion as a concept that reflects students' form-related performance. These results show that interpreters tend to focus more on the form of the output when assessing Korean-to-English interpretations.