2010 July 8
[P]lans are useless, but planning is indispensible. — Dwight Eisenhower
My qualifying exam is coming up, and I’ve started thinking about how I’m going to prepare for it. This post will develop a rough timeline for my preparations up until the exam on September 8th.
There are two components to the examination:
- A talk presenting a plan for my thesis project
- Questions on the talk and the background science
The talk has to do a only few things, in a very schematic sense:
- Summarize the work that I’ve done
- Explain what I plan to do in the future
- Justify the scientific merit of the overall project
- Justify the feasibility of the future plans
- Present a rough, plausible schedule for the work I’ll do until graduation
So, what preparation will I need to do? Obviously, there will be some brushing up on the background physics for the question phase. The talk is the more interesting challenge. Now, of course, Geoff and I already have a pretty good idea of what work I’ll do for my thesis. But, to be honest, that “pretty good idea” doesn’t rest on the firmest foundation: I still haven’t sat down and thought through the justifications (regarding both merit and feasibility) for my project. It seems to me that the first order of business should be to fix that. (Of course part of the intention of the qual exam is kicking students into doing exactly these exercises.) Once I’ve collected my thoughts there, I’ll hopefully be able to work out the plan of attack in a bit more detail than I have before.
Based on this line of thinking, here’s an ordered list of steps for the talk preparation, with time estimates. I’ve tried to give (very approximate) 1-sigma error bars on the estimates, assuming that it’s likely that something will take much longer than complete rather than much shorter. I doubt the simple linear ordering will hold in practice.
- Research and think through scientific merit case. Product: write up a summarizing blog post and put together some references. I think I have a pretty good handle on this, so I think it should take 1 +/- 0.5 solid days of work.
- Research specifics regarding feasibility. Given that the plans ought to be malleable, I’m trying to think of this as research to help me answer the question “What can I reasonably expect to achieve in my projects?” Products: summarizing blog post, references, key numbers / equations. This will take some time: 3 +3/-1 days.
- Digest my feasibility results, reassess plans. Product: summarizing blog post. This will also take some time: 3 +3/-1 days.
- Review previous work, think about how it fits in. Shouldn’t be hard. Product: blog post with references. Time: 0.5 +/- 0.5 days.
- Work out a timeline. Should also be straightforward. Product: timeline. Time: 0.5 +/- 0.5 days.
Then there’s studying and various logistical things to do. Here’s a first-cut schedule of milestones, taking into account various non-qual obligations in my calendar, some expected delays, sleep, etc:
- Today, now – Publish post outlining prep timeline, working out main tasks.
- July 9 (Friday!) – Publish post summarizing scientific merit work.
- July 16 – Publish post summarizing feasibility background work.
- July 27 – Publish post summarizing deeper feasibility thoughts, their effect on plans.
- July 29 – Publish post summarizing what I’ll discuss about my previous work in the talk, relevant issues to bone up on.
- August 2 – Publish post with example graduation timeline.
- August 6 – Publish post with rough plan for studying.
- August 20 – Finish first-draft slides and script for talk.
- August 23 – Have assembled materials for pre-exam meetings with committee members
- August 27 – Give practice talk
- September 8 – Take exam
Obviously, I’m under no illusions that this schedule will be rigidly met. But I will try to stick to it. If my time estimates happen to all be perfect, I have 27 days of work to do, about about 60 days in which to do it, which seems like the right ratio to me, so hopefully the sequencing and spacing are reasonable. If I start slipping the schedule early, I think that will be genuine cause for concern. (I suspect that I could do a good-enough qual with less prep effort, but as a matter of pride and personal development I’ll aim for a very good qual.)
Also obviously, I don’t need to demonstrate my preparation by writing about it on this blog, but I find that writing entries such as this one helps me clarify my thoughts (as is indeed the case now), and I think I’ll be motivated to avoid blowing deadlines on the posts, arbitrary and insignificant though they may be.
Well, I think I’ve dealt with everything I wanted to do for the first milestone. I feel like my plan is a good one, and I feel good knowing that I have it in hand, even if it will surely evolve. It’s easy to go overboard with meta-procedural stuff, and it did take me several solid hours to work out this stuff today, but I feel pretty confident that this, and future posts on the topic, will have been worthwhile.