Score: Scrum for Research
Score is an adaptation of the Scrum agile software development methodology to the task of managing Ph.D. students in an academic research group. Our paper describes Score, conceived in October 2006, and our experience using it. We have found that Score enables us---faculty and students---to be more efficient and thereby more productive, and enhances the cohesion of our research group.
Here is the long version of
the Score paper.
If you are interested in Score, consider subscribing to the mailing
- Why meet three days per week rather than two?
It seems like meeting twice per week as a rule puts too much space
between meetings, and slows the expected pace. When we've had only two
meetings in a week, we "feel" less connected with all that's going on.
- Why meet in the late morning and not the afternoon?
We try to meet with students early enough in the day that we have time
to meet with people after the status meetings. At the same time, we
don't want to meet so early that no students would be willing to
attend in person. So late morning seems like the right compromise. A
corollary of this is that the faculty member should not book his or
her afternoons on the days of status meetings (at least not entirely).
- How do you schedule other, non-status meeting activities?
We scheduled our weekly reading group on a day that we don't have a
status meeting, to get the students in the lab on another day.
- Where is your meeting room, relative to where people sit?
It seems important for the room to not be the room that students
normally sit in. It also seems useful/important for it to not be that
far away from anyone. Finally, the layout of the room is probably
important---it's nice that we are arranged in a semi-circle and that
everyone can mostly see everyone else.