How Do You Grade 100,000 Papers?
I am currently taking another massive open online course. It is the "Crash Course on Creativity" course being offered by Stanford.
I am planning to teach a MOOC in 2013. I am taking notes and rethinking the way I have taught online in the past ten years. I know that I will have to change some things about the way I offer content. I will need to think about presentation and perhaps create some video lectures. And I will really have to rethink how to assess the student work.
As a recent NY Times article pointed out, assessing these massive courses is a massive problem. Some courses use automated graders, but most have had to move to alternatives. Particularly difficult are courses that require writing and analysis.
Coursera uses peer grading and five peers should grade your submission and you need to grade five assignments. If you are a teacher and have tried peer grading in your regular classes, you know that it has lots of problems. Not every student can assess accurately or fairly. Certainly we shouldn't expect that they could just step into that role.
Some course hubs are experimenting with software to grade work and to flag students who assign very inaccurate grades much like holistic norming sessions have done for years with multiple readers.
The article points to Mitchell Duneier, a Princeton professor, who is studying the peer grading on the final exam from his summer MOOC on Coursera. But considering the diversity of MOOC learners - something that is part of the mission of the courses - you have learners of all ages, experiences, training and native languages.
“We desperately need crowdsourcing,” says Cathy N. Davidson, a Duke professor of English and interdisciplinary studies. “We need a MOOCE — massive open online course evaluation.”
I am planning to teach a MOOC in 2013. I am taking notes and rethinking the way I have taught online in the past ten years. I know that I will have to change some things about the way I offer content. I will need to think about presentation and perhaps create some video lectures. And I will really have to rethink how to assess the student work.
As a recent NY Times article pointed out, assessing these massive courses is a massive problem. Some courses use automated graders, but most have had to move to alternatives. Particularly difficult are courses that require writing and analysis.
Coursera uses peer grading and five peers should grade your submission and you need to grade five assignments. If you are a teacher and have tried peer grading in your regular classes, you know that it has lots of problems. Not every student can assess accurately or fairly. Certainly we shouldn't expect that they could just step into that role.
Some course hubs are experimenting with software to grade work and to flag students who assign very inaccurate grades much like holistic norming sessions have done for years with multiple readers.
The article points to Mitchell Duneier, a Princeton professor, who is studying the peer grading on the final exam from his summer MOOC on Coursera. But considering the diversity of MOOC learners - something that is part of the mission of the courses - you have learners of all ages, experiences, training and native languages.
“We desperately need crowdsourcing,” says Cathy N. Davidson, a Duke professor of English and interdisciplinary studies. “We need a MOOCE — massive open online course evaluation.”
Comments
No comments