14.2.3.1
Model A
Model A
| This tutor works in a language department. She first counts the total number of words the student has written (time can be saved by getting students to do this themselves; alternatively a quick sample count of typical line length multiplied by the total number of lines can help verify the count). Next, she underlines all the errors (including spellings, syntax, morphology, vocabulary, coherence across clauses and sentences, connectors), and totals them. She then standardizes the word length, ie if a target essay length were 400-500 words, she might take 450 as her standard length and calculate how many errors the student would have made if he/she had actually written 450 words. For example, if the total word length is 500 and the total number of errors is 45, she reduces 45 by 10% (= 50 words off 500) to give a final score of 40.5. By dividing the standard length by the standard score (450/40.5) she obtains an error quotient. This approach ensures comparability across the group of students and means that those who write longer essays, and presumably therefore make more errors, are not penalized for doing so, while those who write less are not advantaged, compared to others, through the resultant reduction in their potential error count. |
 |
|
|
 |
 |
|
|
|
|
 |
 |
|
