为了科学地量化考生在新闻编译测试中的作答表现，本研究基于新闻编译能力模型构建三分项五等级编译评分量表，包括内容转换的准确和客观、信息编辑的适度和合理、语言运用的准确与得体三大评分维度，另在文献分析的基础上编制新闻编译九级整体量表，并利用facet软件对3名评分员使用两份评分量表对同样60份编译文的评分结果进行对比分析，以验证量表的质量。数据显示，两份量表具有较好的信度和效度。相比之下，分项量表能更准确地区分考生的新闻编译能力，量表各分数段之间的区别更明显，评分员使用该量表时评分员效应更小，评分信度更高。

This study reports on the development of both an analytic and a holistic rating scale for the purpose of effectively measuring test-takers’ English-Chinese News Trans-editing Competence (NTC). Following the construct of the NTC model, the paper proposes an analytic rating scale, which consists of three dimensions, namely content transferring, information editing, and language using. Each trait contains 5 bands. The holistic rating scale is adapted on the basis of Carroll’s 9-band holistic scale. The validity of the two scales was investigated by administrating a news trans-editing performance test to 60 examinees of different levels. Many-facet Rasch Model (MFRM) was later applied to analyze the scores awarded by 3 raters who adopted both the analytic and holistic scale on the same set of 60 news trans-edited versions. Statistics showed that (1) examinees were differentiated, but the analytic scales distinguished the subjects more accurately; (2) difference in rater severity was found among raters using the holistic scale, while there was no difference for analytic rating scale whose rater effects was smaller and inter-rater reliability higher; (3) the adjacent scales of analytic scale were more distinguishable. Generally speaking, the analytic rating scale performed better in measuring the NTC of candidates.