The trials of data collection

18 09 2008

Here is a post from deep in the trenches of data collection. As I type this I have some students working away, testing my prototype application.

I am about half way through all the classes, with one more this afternoon, and two more tomorrow. So how is it going I hear you ask? Hhhhmmmm…

Let me say I am not convinced I will have all of the last data I need. It is still early days, but it points out a major issue with data collection… getting eager participants.

Mine are coming from the student population. Since all is above board with Ethics, I ask the students if they would like to participate. As is completely fair, they all have a choice if they want to participate or not, and it is stressed that they don’t have to. It is the right thing to do, but boy does it make data collection difficult! I wish I had a dollar for each person that asked “what do I get out of it?” or “yeah, but how will it help me?”!

Anyway, after canvassing students in lectures a few weeks ago, I got about 70 willing to participate which I was pretty happy with to be honest. I felt that with all the different kinds of data I was collecting I would get some good stuff. But then we come to the actual experiment.

Well, students have an interesting work ethic god bless ’em. Unfortunately it seems this week is turning out to be one of those “I have better things to do than go to programming labs” kind of weeks. So to say my numbers are low is a bit of an understatement. But we are not through it yet, so lets see how they all pan out. But I guess I am preparing for more data collection, however if the data turns out to be quite rich, then I may have enough to write up (taking into account I have done two other stages of data collection already).

I am not upset with the students… a bit disappointed maybe… but I guess it just highlights thinking of the best ways to get keen participants. I will see how I go and what I can come up with.

So fingers crossed for the rest of the week readers! Lets see how it all pans out.

… Matt


PhD – Life – Play – Fitness Balance (or, the Otter is pretty silly)

28 08 2008

No, don’t worry… this is not going to be some philosophical posting full of lots of zen mumbo-jumbo. Just a bit of a story of my weekend and what else is happening besides my PhD.

Those that know me think I am a bit crazy. It is on a number of levels, but the main one is my ease of being swayed into physical activities. One thing that Monash University does well is their Team Monash program. This is Monash putting formal teams into major Melbourne fitness events, including the Portsea Swim Classic, the Mother’s Day Classic, Melbourne Marathon, and Around the Bay in a Day ride. Why I am crazy is that I do each of these. The Team Monash banner is addictive though… you get looked after well, get gear to run/swim/cycle in, and get be part of the big Monash group.

So, October is coming fast, which means I am running and cycling very soon. Before you think I some kind of Superman (which I am most defintely not), I am not running a whole marathon. That is just for crazy people. I was going to do a half marathon, but it is the weekend before the Around the Bay ride, which is my favourite, so I am just doing the 10km run. But it finishes in the MCG so I am pretty excited about it. My main focus is on riding around the bay.

For those of you not from Melbourne, we are situated on a bloody big bay (Port Phillip Bay). Someone, sometime decided “Hey, we could ride around this thing!”, and thus thousands of us looneys decide to do it each year. How far I hear you ask? Well, 210 butt wrenching kilometres, that’s how far. Yes, it is a long way, but it is also believe it or not it is really fun.

Below is a Google Map of what we ride. Scale is obviously hard to tell, but it is about 210-220km of riding. We get a ferry from Queenscliff to Sorrento about half way, so we literally do go around the whole bay.

This is where the title of this posting comes in… at some point I had to tell (well, force) myself to say “Hey, you need to train for this… get on your bike!”. The main problems have been:

  1. I really want to get this PhD done
  2. Our winter has been pretty cold and wet this year
  3. Training means doing rides of at least 100km (ie at least 4-5 hour sessions)

So the weekend gone I said enough is enough. Sunday, being a pretty nice day, I threw the bike in the car, drove to my favourite spot in East Malvern, and started riding. I headed into the city via the main bike trail and went through to Port Melbourne, and the start of the famous (or infamous) Beach Rd ride. And that is where it hit me.

As I headed on my way to Frankston I had one of those kind of zen moments. I thought to myself “It is a beautiful day, I am on the water, I have nowhere to be, and this is great!”. I think it was a combination of a cold wet winter, not having been on my bike in ages, work being crazy, and the fact it was a nice day with no wind. But it was just one of those blissful moments.

So I proceeded to ride close to 120km over the next 4.5 hours. Yes, it is a lot, but I needed to kick the training into gear. But if 120km can be peaceful and enjoyable, this was it.

Anyway, that is the end of the story. What is the moral? Well,

  1. Yes, I’m nuts
  2. I discovered I am pretty fit (I do a couple of gym RPM classes a week to keep my fitness and sanity)
  3. Sometimes you just need to say “Sorry PhD, today is for my body, not for my brain!”

If you are struggling with trying to fit these things in, sometimes you just have to do it. Nothing like registering to ride around a bloody big bay to get you motivated. See you on Beach Rd!

… Matt

Otters hitting their heads against brick walls

11 08 2008

It had to be too good to be true. After a few weeks of stellar progress, I have hit the wall in my development. This is because of a few things…

The first is that I have struck two things I need to do with my code that it just won’t do. I know there will be ways to do it, but one particularly is just frustrating the hell out of me. The longer I look at it, the further away a solution seems. So in part this entry is me trying to take my brain away from it for a moment.

The second problem is not really a problem, but a reality. Semester is underway and slowly but surely other commitments sneak in. Teaching, supervision, a meeting here, a meeting there. Then your week has disappeared. My load this semester is quite good, but things are starting to pile up 😦

The final influencing factor I guess is being against the clock. I really have to have this nutted out as I want to do some evaluation of it in class in a month or so. So I need to get this puppy sorted and some preliminary testing done like by yesterday. Nothing like a deadline to bring out your best!

So, here is hoping that moment of clarity comes sometime today. Please. Pretty please? Maybe it is time to change my music.

… Matt

Yet more Data Analysis… different data set though (and different issues)

2 06 2008

The Otter has been a bit quiet, but is still beavering away (pun intended). My last couple weeks have been mainly taken up by end of semester teaching commitments. I did some more data work on Friday, but not with my original data set, rather with some student exam data I am using to obtain further insight into student performance across programming curriculum.

It is proving to be quite problematic though. I am finding the data (well, the data source mostly) proves to raise more questions than answers. I first thought that by breaking the exam into specific questions relating to certain curriculum I could use that to gauge actual student performance with the programming topics. But on further discussion with Michael and Judy, it is a bit of a problem. For example:

  • Did a student perform more poorly on a question because the topic was hard or because it was just a hard question?
  • Did the student not do as well because the question was at the end of the exam rather than the start?
  • Did the student not do as well because it was a different type of question (ie multiple choice, short answer, algorithm or java coding)?

See the problem? I can’t say a topic was more difficult, because they could simply have done that question last and run out of time. I can use it though to get some supporting evidence to my ideas. I already have a big data set I am using as primary data so I am not worried… this is simply some back up data. So I am doing some basic analysis and that will be it. Just shows how thorough I am and that I can acknowledge problematic data sources.

Back to my 300 sets of exam data!

That reminds me, here is a maths/statistics joke for all you PhD nerds…

An astronomer, a physicist and a mathematician are on a train in Scotland. The astronomer looks out of the window, sees a black sheep standing in a field, and remarks, “How odd. Scottish sheep are black.” “No, no, no!” says the physicist. “Only some Scottish sheep are black.” The mathematician rolls his eyes at his companions’ muddled thinking and says, “In Scotland, there is at least one field, containing at least one sheep, at least one side of which appears black from here.”

If you don’t get it you should feel happy. If you get it… you have spent too much time working on a PhD.

… Matt

Do I work on my PhD or mark these assignments?

20 05 2008

Interesting dilemmas come up as a part-time PhD and full-time academic. As a full-time academic you have certain obligations to satisfy for your School, Faculty, Campus, and University. These span teaching, administration and also research. But how do you keep those things in check while trying to get a PhD finished? When is my PhD the most important thing, and when should other obligations take precendent? When do you have to say “screw everyone and everything else, I’m the most important thing”?

I don’t know the answer to be honest, and it has been easily the biggest struggle for me in the last 5 years. Every day, you have to make a judgement call between what is most important… this teaching/admin/other-research task, or my PhD. What your answer is speaks volumes about you I think, as every day I see different people make different decisions. But what kind of decisions do I have to make?

  • Do I work on my PhD or mark these assignments (knowing students have to get their assignments back in the next week)?
  • Do I work on my PhD or help out for a community event?
  • Do I work on my PhD or work on research unrelated to the PhD (that is important because it promotes research realtionships across the faculty or campus)?
  • Do I work on my PhD or join that committee?

And these are just the tip of the iceberg. There is no one answer to these questions, and I’m not judging people’s decisions for a second. Written on paper it seems easy… the PhD wins everytime. But in the thick of the semester, it is a bit more difficult. And then add on top, trying to do the right things for your job and your future career and it gets awfully messy. Sorry I don’t have any answers!