In the chronicle yesterday, David Youngberg wrote a commentary titled 'Why Online Education Won't Replace College - Yet'. While he makes one good point, there are many other problems I have with his position. Some are just different ways of thinking about education and assessment, while others are perspective.
Youngberg's first criticism is in cheating, namely that if a MOOC (his central concern in online education) were for credit, cheating would be rampant. Well, cheating is already rampant in higher education, and it is sometimes very difficult to catch. My favorite was the student who took pictures of the test, sent them to other students waiting outside, who then came in "late" for the exam (the first person showed up early). Yes, they got caught.
For me the issue of cheating revolves around the goal of the assignment. Youngberg does point out that in the Udacity course he enrolled in did allow collaboration on discussions, but then goes on with the idea that there would be cheating if the course was for credit. I state again, your concept of cheating is ultimately tied up in what you view as the GOAL of the assignment.
In my classes, I use a number of online quizzes as formative assessments. Students have unlimited attempts, and the database of questions that randomize between each quiz is large. I encourage the students to cooperate in answering the questions. The goal is to get them thinking, to reference their textbooks, and online sources. I want them to work on finding the answers. Colleagues have accused me of allowing students to cheat, because a quiz "should reflect what the student knows." How do you argue with someone whose perception of assessment is so myopic? Even online exams (which are not worth a killer amount of points) are not proctored. Could the students be sitting next to each other? Could they be comparing? Could they be looking thing up? Yes, and I expect that they are doing that, but they are still learning.
Let me state again, these are not punitive exams point wise. These exams are not worth 1/3rd of their grade, like most people post exams. If they do poorly, it doesn't stop them. Instead, it shows them where their still struggling.
This past semester, my students only had one in class exam: the final. Even then, they had a second chance to take it if they did poorly (new question set though). I had fewer suspicious students during that final exam than I've ever had before. Fewer students with stray glances, furtive looks at what could be a crib sheet, or even the tell-tale bulge of a cellphone. I took the pressure off the exams, and the cheating went down.
When I asked students how much help they got when taking the exams, they said a little. When I asked them to explain, they said they sat near someone, but that the other person wasn't that helpful (they had 50 minutes to do a 50 question multiple choice). Some admitted that they looked up an answer they couldn't figure out. Strangely, it didn't both me. I still saw it as learning. That, and I ended out with the highest average on the final exam (first round) than I've ever had.
So, to sum up, I think cheating is predicated on the goals you have for your assessments.