NCTE opposes the use of machine scoring in the assessment of writing. Such use devalues and eliminates the social and contextualized nature of writing. The machines’ practical shortcomings trump the benefits of expediency. Writing is a valuable human communication and must remain as such.
NCTE member Les Perelman discusses using computer to grade writing:
Computers Grade Essays Fast . . . But Not Always Well
(Morning Edition, National Public Radio)
Facing a Robo-Grader? Just Keep Obfuscating Mellifluously
(The New York Times)
Writing Assessment
(Conference on College Composition and Communication)
Teaching, Learning, and Assessing Writing in Digital Environments
(Conference on College Composition and Communication)
What Happens When Machines Read Our Students’ Writing?
Anne Herrington and Charles Moran provide a 30-year history of composition research and thinking on computer-grading of writing: “We need to say at the outset that these programs confront us with a new situation: writers writing to computers.”
Grading Student Writing: High-Stakes Testing, Computers, and the Human Touch
Crispin Sartwell has offered a sharp criticism of the template approach to teaching writing, and he fears that computer-aided grading of student writing will only increase the problems. We must be skeptical. Since computer programs and word processors are able to identify surface-feature concerns in student writing, there is a danger that computeraided writing instruction and assessment could increase the disproportionate value placed on surface features at the expense of the writing process and the content of authentic student writing. Computer writing programs that assess writing can also reduce all writing to templates. Sartwell notes that “machines are cheaper” than humans, so we must be sure that we do not allow expense to supersede quality when we are teaching writing and grading student compositions.