Breathing a collective sigh of relief that they are not violating federal law, the nation’s teachers return this week to the widespread practice of letting students correct each other’s papers. On Tuesday, Feb. 19, the U.S. Supreme Court ruled on Falvo vs. Owasso School System, deciding that grading another student’s paper is legal. The court rejected the “politically correct” view that students’ grades must be secret from their classmates. The court offered the view that students can learn as much from grading each other’s tests as they do from taking them – and save their teachers time, as well. Yet, the case is not closed on the educational effectiveness of such practices.
A recent study conducted by Philip Sadler, the director of the Science Education Department at the Harvard-Smithsonian Center for Astrophysics, and his Harvard Graduate School of Education student Eddie Good examined the impact of peer-grading and self-grading. Sadler and Good wanted to determine whether the grades that students award themselves can actually replace their teacher’s grades and whether any additional learning results when students grade tests.
In four of Good’s middle-school science classrooms at the Ephraim Curtis Middle School in Sudbury, 101 students learned how to evaluate tests using written guidelines. They used time in class to correct tests and vigorously debated their judgments. “Even though the assessment included many open-ended essay questions,” said Sadler, “students quickly learned to award grades with an accuracy that rivaled their teacher. Despite the opportunity for embellishing one’s own grade, students still graded their own tests more accurately than the tests of their classmates. Only students on the lower end of the performance scale tended to inflate their own scores when grading themselves.”
The surprise came following the results of an unannounced, second administration of the same test a week later, intending to measure whether student grading results in increased learning. “This was something we really did not expect,” adds Sadler. “It put a whole new light on grading exams.” Those students who graded a classmate’s test did not improve significantly over a control group of students who did not grade any tests at all. However, students who graded their own tests improved dramatically the second time they were tested. Moreover, self-grading contributed to a rise in test scores for students at all performance levels. While peer-grading saves time for the teacher, it was self-grading that proved more educationally valuable.
Sadler and Good applaud the Supreme Court in its affirmation of student-grading as a legitimate teaching practice, but urge that these grades count only if students are taught to grade accurately. With self-grading showing a huge advantage over peer-grading in science learning, Sadler and Good hope that researchers will continue to study the impact of such popular teaching methods on the nation’s schoolchildren.