“Our overall conclusion is that, within the confines of our study, both male and female students can and do grade their homework honestly” (p. 57). That’s not an expected conclusion, or one most faculty would hold about their students. If the homework counts in final grade calculation, then there’s a significant incentive for students to generously grade their work, if not outright cheat. Moreover, there’s the question of expertise. Do students have the qualifications needed to grade their work accurately?
Authors Simkin and Stiver recognize these problems. Self-graded homework problems are not an option in every class. But they did work in seven sections (266 students) of an introductory information systems course where, in each class session, students were required to create as many as five small programming applications in Visual Basic. Each session began with students reviewing a suggested solution for each problem and a table of maximum points that could be awarded for each exercise. Students could give themselves partial credit. These self-graded homework assignments counted for either 20 or 25 percent of the course grade. “The instructor told students they were on the honor system and also noted he would not challenge their self-assigned grades” (p. 53).
And what evidence is reported that supports the opening conclusion? “One way to examine the concern for overly generous homework marks is to look at minimum scores. If students are charitable in their homework grades, then one might expect to see high minimum scores” (p. 54). A table in the article confirms the presence of low homework scores in all seven of the sections studied. Beyond that, these faculty researchers compared homework scores from one of the sections with professor-graded homework from another course. That comparison revealed that “the homework scores in the IS classes were more dispersed than those from the benchmark statistics classes. In the IS classes, the minimum score was lower, the sample standard deviation was larger, and the coefficient of the variation of 29.2 percent was almost three times larger than the statistics class value of 10.3 percent” (p. 55). Finally, the research team looked at correlations between homework scores and scores on the final exam. “If students had accurately and honestly graded themselves, we would expect to see a significant relationship between homework performance and final exam performance.” And they found that significant positive relationship. “The lead author’s experience was that counting self-graded homework in the final course grade was not as risky as he originally feared” (p. 56).
It is a surprising result, perhaps partly explained by the median age of these students, which was twenty-six. But these results confirm the positive experiences with self-grading reported in some other research. Moreover, “comments on end-of-semester class evaluations from students suggested that most liked the process, appreciated the vote of confidence in their integrity, and believed they were honest graders” (p. 56). The authors also note that grading the homework at the beginning of the class sessions got students arriving for class on time with their homework ready for grading. In some cases, students had alternative solutions. The authors write that some of these were “novel” and “creative” approaches. “In our opinion, the ensuing class discussions about the viability of such alternative tactics were one of the most educational portions of the course—especially for the instructor.” (p. 56)
A self-grading scheme might be most viable in courses where the homework involves unambiguous right answers. However, with the advent of rubrics and other mechanisms that require students to assess specific aspects of answers, it may be that students could do self-grading of other kinds of homework as well. This article lists an impressive array of benefits accrued when students self-assess, not the least being the immediacy of the feedback. Self-grading may also make it easier for students to see the difference between a right and wrong answer or see what makes one answer better than another. References are included to studies that suggest that “self-grading improves class attendance, makes the classroom experience a friendlier, more productive cooperative environment, and provides a shared sense of ownership for the learning process” (p. 52).
No, self-grading isn’t an option in every class, but perhaps it’s an approach that shouldn’t automatically be ruled out.
Reference:
Simkin, M., and Stiver, D., (2016). Self-graded homework: Some empirical tests of efficacy. Journal of Education for Business, 91 (1), 52–58.
