<p>I’ve graded standardized tests. Please believe me – for the vast, vast majority of students, simply getting frequent feedback on basic structural problems, spelling errors, and grammatical errors would be a tremendous benefit, especially if the program was smart enough to display the problems and the associated explanation in a way that makes sense to students, not just as a series of changes or simply as a grade. </p>
<p>This doesn’t get away from the teacher needing to help the students develop arguments and voice, but personally I’d rather work with students on the quality of their arguments and supporting data before the arguments have been written into an essay. In my experience, most students’ willingness to radically change their arguments and find better supporting evidence is directly proportional to the amount of time they’ve already invested in the essay. Once they’ve done a full first draft it can be very hard to budge them. </p>
<p>And on the accuracy front? Did you know that many states adopt grading rubrics for statewide writing tests that prohibit graders from incorporating the accuracy of any assertions into the score? Or that the rules may prohibit dinging a student more than once for a certain kind of grammar/spelling error – so a kid who writes an essay with 1 subject-verb disagreement and 1 spelling error gets dinged just like the student with eight subject-verb disagreements and 25 spelling errors. Yes, they’re scored by humans, but the rules may not be what you think they are.</p>