Sub Ubi

Last term, the first of my research data rolled in.

By the end of the week, I’ll have a draft of a paper presenting a partial analysis of that data completed.

Tonight, I didn’t sleep much, as (today) I’ll be presenting that analysis to my research group, and anyone else who tags along.

Memo to self: 60-second breakpoint, common errors :: common words.

Is it a good idea to go to bed at 6:15 AM, when you have to present at 10:00 AM? Or, is that just asking for trouble?

I’m going to find out. Someone call me at 9:15 AM…

I just caught Brooklyn Joe’s pointer to the Reagan Dime Proposal, put forward by Rep. Mark Souder, Indiana.

I seem to remember something about a record US deficit and a Vietnam-esque “police action” in Iraq (a sovereign nation we invaded without provocation), while US representatives are bitching at each-other over dimes, because of a television show that paints Reagan in an unflattering light? (full script, 8MB PDF)

Please. Someone. Take. The. Rudder.

May 2005.

In May 2005, I’ll be done with my Ph.D. It’s the target, and I only intend to miss with reasonable error (+/- 3 months).

Having a few weeks of uninterrupted quiet at home was wonderful; coming back to the UK was great, too, but the challenge now is to carry that productivity all the way through to the end. I’ve noticed that I tend to rely on an ad-hoc approach to scheduling my time, and always focus intensely on one project or task when working. My goal at this point is to work on improving my time-management skills, with a particular focus on handling shifts between projects more gracefully, more often.

Unless someone has a better strategy, the approach I’m going to take is to begin structuring my time more rigidly. I already have a few things that are locked in for the remainder of the term (Tuesday/Thursday nights are the IT certificate course and CSCS, respectively), and I need to keep a weekly squash date in as well. While that leaves a lot of time, I think I’ll get more done if I explicitly block out time for the work I need to do, be that reading, writing, programming, analysis, planning, etc. Flexibility is good, but I think it’s time to experiment with less flexibility and more focus on distributing the time more evenly among the various things that must be done to complete my research and get papers and a dissertation written, as well as individual projects and things I do for personal enjoyment.

We’ll see how it goes. There can’t be much harm in trying to get more organized, anyway.

(Ooh! And a little countdown timer on the left! Clearly, that’s what I’ve been missing… )

Powerless is a very strong word. I doubt I’ve ever personally experienced powerless in my lifetime. Right now I may be hard pressed, frustrated, and a bunch of other things, but I’m not powerless. That’s important.

I’ve tracked onto some good ideas for eliminating sources of frustration in my life right now, which helps. I also feel good about my return to the UK, where I can keep working on my research and some projects that are important to me. Hell, the real truth is that I will be returning to England armed with O’Reilly’s 2nd edition of “Learning the Bash Shell.” How can life not be on the up-and-up?

I’ve learned something once again about being a grad student; the word I’ll use is powerless.

As a graduate student, I’ve been remarkably lucky: my general enthusiasm and willingness to do has often been welcomed, and I’ve been able to accomplish a great deal. I have not necessarily traveled in straight lines, but I’ve had a great deal of control over my own destiny. This suits me; I enjoy being able to set goals and achieve them (or not; but I can deal with my own failures well enough).

The last few weeks have been a struggle, as I have encountered things over which I have no control. While all my department’s systems were down for a week, I focused intensely on analyzing my research data; I eventually reached a point where I needed input from an expert. While I felt I was out of my depth, I was empowered to find people. I called universities in the area, and wrote faculty I had (tenuous) ties to, in hopes I might learn what I needed to know to keep me moving forward, and to achieve my goal of having a summary research paper written by the end of the month.

When systems came back on-line, I focused on finishing setup of software to collect research data from multiple sites this semester: two univeristies in England, and if I’m lucky, one or two from Australia and/or New Zealand. In November, I began this work, and before I left for the States I had achieved 90% setup on redundant data stores, both in the USA and in England–two sides of the trans-Atlantic cable.

The last ten percent in the USA took no time at all. The last ten percent at my home institution took a week. I may not yet be out of the woods—I may yet find that I won’t have software in place to collect data come Monday, despite my spending 12+ hour days dialed in, trying desperately to affect my own destiny from 5000 miles away. Rarely do I loose sleep; it is now well past midnight, and while my ability to focus critically is gone, my teeth are still locked tight.

I have failed to deliver fully on some responsibilities over the last two months that I regret, and I am not prone to regret things. I agreed to collaborate with a colleague on some Scheme scripting work; I’ve left him high-and-dry at times as I’ve banged my head on my own problems. I have a colleague at another university in England who is waiting for software and documentation I can’t deliver, because of problems out of my control, despite all of my best efforts and intentions. I’ve fallen behind on arranging talks for the CSCS group this next term; I should have delegated that responsibility before Christmas. And as my wife drifts off to sleep in the room next to me, I can’t help but feel a bit bitter, as I should be there in her arms, drifting off to sleep as well, and not here, venting my spleen. Given that we live apart for most of our time in England, this time is precious, and I’m ever aware of that fact.

Why the regret? Because the people I like collaborating with most are people who I can always count on to do their best, people who won’t let me down. This feeling of powerlessness is like a knot that only gets tighter with struggling, and as I’ve struggled these last two months, I’ve been bound tighter and tighter, leaving me in a position where I’ve not done for others the way I’d hope they would do for me.

Mostly, I’m dissapponted in myself, and my lack of forsight. As a result of this bind, I’ve learned something about myself. And as a result of that learning, I now have ideas for how I can eliminate potential problems like this in the future. But for some reason, right now, I don’t feel a whole hell of a lot better right now for having learned that lesson.

Tomorrow, perhaps, will be a better day. I’ve been saying that for well over a week now, but perhaps it will yet come true. I’m taking Chinese food over to my grandmother’s tomorrow for lunch, and will try and put all of this bullshit out of my head for that time. Then, I’ll try yet again to affect the world around me in a way that is conducive to completing my PhD.

Lesson for future PhD students? Set your goals. Chart your paths. Identify those aspects of your planned paths that are out of your control, and have contingency plans for them. Rinse and repeat.

I’ve taken up residence at the Strongsville Public Library: a 37,000 ft. state-of-the-art facility with WiFi access everywhere. Comfy, well lit, and connected.

Anyway. I had written one or two faculty in the Cleveland area to see if I could chat with them about statistics (and my lack of knowledge thereof). I had the good fortune of hearing back from Prof. Richard Little at Baldwin-Wallace College, who was willing to take a look at what I had and see if he could help. Not completely a random pick–my father had him when he was doing his MBA a few year back.

It was an excellent 2¹/₂ hours; conversation ranged from questions about how I ended up doing what I do to why I do it where I do, and what I hope to do next after I’m done doing what I’m doing now. He shared with me some of his experiences in doing what he has done, and over all… well, a lot of chat about people doing things.

Truth told, the conversation was particularly good because I didn’t have anything I presented to him completely blown out of the water. If anything, I just needed to present what my efforts and explorations to a colleague who knew as much or more than I about the analysis I’m carrying out (methodologically speaking) to make sure I hadn’t navigated my way out into left field. I don’t think I could have stumbled upon a better sounding board.

I feel a lot more confident about my intuitions at this point; I still don’t know a standard normal distribution from a molehill (well, ok… only p = .5 moles live in a standard normal distribution, whereas you can’t have .5 moles in the real world), but I do know that the reading I’ve been doing hasn’t been completely lost on me. And while most of my focus at Indiana was on qualitative research methods, what I picked up regarding quantitative research methodology has served me well. I suspect I’ll have a draft of a PPIG together fairly soon. Given how extensive some of my notes have been as I have dug through the material, I think writing will go smoothly.

I have some New Year’s resolutions to make, and I have a comment or two from previous posts to reply to. But, right now, I need someone who knows their stats.

In particular, I need someone I can ask some questions of who will respond in a timely way, and is able to point me at resources, be they articles, books, or software. I’m up to my neck in data from last term, and am in way over my head in terms of analysis.

In collecting data about novice programmers at Kent, I’m able to make statements about

1. time between compilation events,
2. volume of code change between events,
3. result of compilation event (syntax error or error free),
4. type of error (in the case of error),

and several other interesting things. Even taking this one variable at a time (say, time between compiles), I am running into a bit of a wall. I’m attempting to characterize my univariate distributions (eg. I believe my ‘time between compiles’ data is distributed according to a Weibull distribution), but my intuitions about statistics break down very quickly; characterizing unknown distributions is currently not my forte, nor do I even know what tools (analytical or in terms of software) would be of the most aid to me at this point.

It is also possible that I’ve managed to do a good job with my analyses so far, have actually developed some decent intuitions about my data, have applied some appropriate teechniques in the appropriate places, and am just confusing/worrying myself unneccessarily.

Regardless, if you have a friend who knows their statistics, ask them if they’d mind a phone call or email from a computer science PhD student doing some exploratory data analysis on a novel, unique set of data. If they’re in the greater Cleveland, OH. area, I’ll buy them lunch as part of the deal.