For my math classes, I typically have one problem set due weekly. I struggle with prioritizing learning the material and trying all of the problems early.
Ideally, I would spread out a series of tasks on a weekly basis. I am going to try this schedule with a Beeminder set at a rate of 15 tasks per week. I will include some Buffer to avoid the way Beeminder averages the rate over individual days.
Friday – Edit solutions in AM, Watch lecture 1, 2
Saturday – Read book 1, 2 / Watch lecture 3
Sunday – Read book 3 / Write down questions based on material / Start first 66% of problems
Monday – Go to office hours / Do a second pass on first 66%
Tuesday – Try remaining 33% of problems
Wednesday – Go to office hours / Do a second pass on remaining 33% of problems
Thursday – Type solutions in AM / Edit solutions in PM
Will check in on 10/30 to report how this is going!
I made a Beeminder goal for this.
@shanaqui’s blogpost this week on Beeminding the Right Thing: A Bookcase Study prompted me to think about what actually results in me failing to spread out my tasks.
The first is a lack of urgency. Beeminder addresses this. Tonight I left a social game night after a reasonable amount of time to prevent a derail. I’m happy with this decision in hindsight, even though it was slightly painful in the moment.
The second reason I fail in spreading out my tasks in this area is not spreading out my tasks in other areas! The gaps in the graph so far have occurred when I did urgent work in other areas. I think the natural solution is to create similar goals for my work in those areas.
The third reason I fail is that I need to watch lectures before working on problems, and lectures take an hour to watch. This needs to be carefully planned out in my schedule at a time that is convenient. I might consider setting specific deadlines for this in TaskRatchet.
Going forward, I need flexibility beyond the schedule in my original post for circumstances when homework deadlines shift (like this week). I also need better fine print. I don’t necessarily want to Beemind “study time” because short bursts of work are very valuable, and I will succeed more if I revisit a problem 3 times for 10 minutes rather than for 30 minutes straight. I want to motivate myself to do those short bursts. For these reasons, I might better refine what I mean by a “topology task” (e.g. lecture watched, 25 minutes reviewing concepts, 25 minutes reading, 10 minutes on a problem, typing a problem). I am unsure what heuristic I should use when deciding whether to make “lecture watched” worth more than one task. It is certainly harder than all of the others by at least a factor of 2. It is also important for me that I watch the lectures early in the cycle of tasks.
I am also unsure whether I should change this to a daily rather than a weekly goal. What do you usually do to decide between a daily and weekly goal? I am leaning towards weekly for the flexibility, but need to avoid a gap of more than one day (is this where ratcheting could help?)
Eee, I’m excited to see that my blog post prompted some thoughts here!
I think in the end there isn’t much difference between a daily and weekly goal, except perhaps in how you think about it. If you set a daily goal of +2 and a weekly goal of +14, you get the exact same graph, because of how a straight-line graph works.
But saying a weekly goal can help you think about it in terms of how much you want to do per week, and not worry so much about what you do every day – if you think in terms of how to set it per week, you are explicitly saying to yourself that you can do more on some days and less on others. (You can actually do the same with a daily goal, in reality, but we’re talking about planning here.)
I would probably simplify things by beeminding study time, but could you say a bit more about why that wouldn’t work for you? Maybe that will help me pick apart what I’d do.