While watching my favorite TV show, The Colbert Report – some important aspects to the computer grading system of essays were presented. This immediately piqued my curiosity for further research.
It appears that with many states incorporating the Common Core Standards, computer companies are already positioning themselves toward government contracts for automatic grading of the standardized tests with their main focus on making money. But the question lies in the reliability of the computer to assess accurately.
Thus far there are two types of machines that are programed to grade tests – supervised and unsupervised. They both are programed to think faster, are cheaper to use, time efficient, and lift the burden from teachers and graduate students that are paid to grade the tests.
Real people crosscheck supervised machines. The supervised machine takes the essay and treats it as if it were a group of words. It measures written words in a sentence, the word length, quotations in text, punctuation, and length of the essay. The computer is then programed with grades previously assigned by people. Statistics reveal that the machine’s technique actually works as if teachers and graduate students were grading the tests. The unsupervised process of grading skips the people aspect altogether.
One study by Mark Shermis at the University of Akron suggested that computers and teachers were capable of recording similar scores for essays written by students. On the opposite spectrum, a group of professionals indicated that computers weren’t as good as teachers evaluating students. This study can be found at http://humanreaders.org/petition/research_findings.htm.
According to an article in the Washington Post, the National Council of Teachers of English was against the use of computers for the grading of student essays. Other professionals profess that the computer could not measure accuracy, reasoning, argument, or even organization. Writing is derived from higher level thinking skills and deserved constructive feedback.
While grading tests, a computer can possibly be fooled through the writing of inaccurate facts that are embedded into a well-structured sentence. An essay can also be padded with several nonsensical sentences and the student would still receive a high grade for the written work.
How do we want our students’ papers to be scored? These scores will eventually be compared to other states and nations. Will this process lead to having computerized teachers grading computerized students? Is the future of our educational system in the hands of computer programmers? I say – leave the scoring to the people doing the teaching – our educators.
