Quality level descriptors and score range It is also necessary for teachers to consider both the wording and score range in rubric design. The use of such words as excellent, good or poor is likely to evoke subjective responses from students and interfere with their self-judgements of performance. Those words maybe replaced by more neutral equivalents like emerging, developing and proficient. It is also worth considering broadening the score scope (a point one for instance) to provide more room for students to fine-tune the judgement of their writing performance. Further research would be needed to explore whether the enlarged score range would result in increased accuracy of self-ratings and improved self-feedback information. Rubric users domain knowledge Regarding the role of students domain knowledge in rubric use, it is worth exploring the rubric’s function with reference to Hattie and Timperley’s ( 2007 ) three questions on feedback (1) Whereto go (2) Whats up (3) What next As noted by the students, although the rubric has well fulfilled its function of clarifying the goals that they should reach (Question 1), they still held idiosyncratic conceptions of writing derived from their previous writing experience and might therefore assess their writing on that basis. Additionally, while the students realised their present levels of performance (Question 2) using the rubric, their limited English proficiency and insufficient knowledge about assigned essay topics might inhibit them from improving their writing performance (Question 3). In this sense, the rubric can be regarded as doing well in answering Question 1&2 while lacking instructional power regarding Question 3. Such a result demonstrates the secondary position of assessment (rubrics included) to learning. The learning potential of assessment seems to be only tapped after students have acquired sufficient domain knowledge. Without a sufficient knowledge base, assessment has but limited instructional power.