Most online courses that incorporate grading have tended to rely on robo-graders. These can work well for fields such as math and programming—for which one can devise tests with easy-to-verify objective answers—but don’t extend easily outside the “hard” sciences. Coursera is working around this by using peer grading (students grading other students) for some of their courses, such as Introduction to Sociology and Human Computer Interaction.

The obvious question that pops up at this point is: “How well does this work?”

In an interesting article on her online learning blog, Debbie Morrison relates her recent experience with peer grading in the introductory sociology class on Coursera. The following is a summary of her findings, but I encourage you to read her complete article along with the comments from her readers.

1. Overall, peer grading works reasonably well and is reasonably accurate.
2. Grading the work of their peers’ enhances students’ own learning.
3. In order for peer grading to work well, it needs detailed scoring standards and clearly explained instructions.

And so this is how you scale a non-technical class to hundreds of thousands of students at a time. Ladies and gents, the future is now.