[SatvikBeri]

Interesting stats. I would love to see some stats on classes using different deadline models. For example, the Machine Learning course gave you 80% of the grade if you were late, and the NLP course gives 50%, as opposed to losing everything. I wonder how that affects the drop-out rate.

[robrenaud]

This isn't about coursera, but there was some interesting research in presented in Predictably Irrational about deadlines and student performance. Basically, deadlines help a lot, and having many small deadlines to ensure continuous work helps prevent procrastination.

http://bookoutlines.pbworks.com/w/page/14422685/Predictably%...

> Group 3 (imposed deadlines) got the best grades. Group 2 (no deadlines) got the worst grades, and Group 1 (self-selected deadlines) finished in the middle.

[SatvikBeri]

This is wonderfully circular-the study I posted in my original post was by Dan Ariely :)

[robrenaud]

Argh, sorry. One of the grand parent posts was too long, I didn't read the parents above it.