Ruminating on Rubric

Traditionally, we use a Levels of Response (LoR) approach in assigning a score to students’ Use of English in written work for English Language. Five bands are used, and each band has a marks range of 3 or 4 marks [A (‘Excellent’ band) = 16-18; B = 13-15; C= 9-12; D = 5-8; E (‘Limited’ band)= 1-4] making a possible total of 18 marks. Descriptors for each band spell out what a student must demonstrate if she were judged to have scored a mark in that band. At a benchmarking exercise, teachers typically would discuss (i) the band that a script would fall into, (ii) the specific mark in the 3- or 4-marks-range in that band that the script should earn. The first decision is arrived at by paying attention to how well a script fits the description in a band. The second decision involves a comparison between scripts. Say scripts X and Y are both judged to fall in band C; further, X is judged to have scored 10. We next ask, “While X and Y are both band C work, is Y stronger or weaker than X? How so?” Based on the evidence of work, a team at a benchmarking exercise can come to a consensus on whether Y scores 9 or 11 or 12.
The challenges of scoring using this LoR approach are:

For teacher/assessor

1. If seven bullet points make the description of a band, how do we score a script that fit 4 of 7 bullet points in one band and 3 of the 7 bullet points in a higher (or lower) band? How does the ‘best-fit’ principle apply when there is evidence that a work best fits part of the description in one band and also best fits part of the description in a neighbouring band?
2. Within a band, we are expected to discriminate further by awarding a mark out of a range of 3 or 4 marks. How do we ensure consistency in exercising such discrimination within a band?

For students

1. What information can a student reliably draw from a score? Say a student scores 10, ie. C-band Use of English, does it tell her that she has demonstrated a grasp of grammar that is adequate for communication expected of a 13-year-old? Or does it tell her that she has slipped up sometimes on spelling, punctuation and register, and that those slips can be forgiven and therefore she did not fail? Or does it tell her that she is sound in punctuation but her vocabulary range is not sufficiently nuanced to convey shades of meaning?
2. Given her current score in Use of English, does a student know what she should work on if she wishes to score higher the next time she produces written work for English Language?