Tag Archives: ilaqh

Posterior samples

Everything I need to know about Bayesian Statistics, I learned in eight schools. At first I thought this meant not really understanding it until having worked in many places with different people but it’s actually a reference to a particular example of hierarchical modelling.

Hadley Wickham talking about dplyr. Very fancy.

I’m finally attempting to install RStan. My computer is misbehaving.

And it’s ARC Discovery Project writing time.


Web presence

I first met the friend who’s going to try learning R over the summer by way of a blog that she and a mutual friend of ours ran way back in the day. She recently posted a link to a Guardian article about cringing at the blogs we (as a society) had ten years ago and mused a bit on her old group blog. Ten years ago I was being introduced to Livejournal by a fellow engineering student at QUT but as far back as 2000 I was dabbling in blogging and writing webpages.

Fed up with some of the attitudes at our high school, a friend and I had joined a webring, where we wrote terrible poetry, passable HTML/CSS and banal updates about the coding of our respective webpages. I’m definitely not going to post a link and I’m glad I never used my name on that site.

Ten years ago (almost to the day, according to the source) the first version of the ILAQH website went live. If you’re reading this beyond Monday 16 you may not be able to see what I can see now. Below is a screen grab of what I see.


The website seems to have not had a design refresh in the last ten years, on account of various factors to do with QUT’s IT policies, not having a postdoc who’s also a professional web designer, etc. ILAQH has been around in one form or another since 1995, which means that as much as I cringe looking at the layout of our old site it could have been much worse.

We now have a new website that thematically fits with the rest of QUT’s web presence and is much more appropriate for a university research group who publish good quality work in the top journals in their field, have organised an international conference recently, and have a string of ARC Discovery grant successes.

Clearly a good web presence is important for research groups but it’s increasingly becoming more important for individual researchers. It’s a way to connect with potential employers, students, colleagues, etc. and a high quality site can really show what sets you apart from others. I’m trying to convince some of the other people at my sort of academic level to get their own websites so that they’re easily searchable online and have a page that’s separate from any university affiliation but also not tied to any one social network (not even the professional ones like ResearchGate or LinkedIn).

Out of all the blogs and web presences I’ve had, I’m definitely proudest of this one and doubt that in ten years time I’ll be looking back at it and cringing. I’m not a professional web developer, just someone who’s got a WordPress account and a bit of experience mucking around with software. Any researcher has the skills required to come up with a website for themselves and their research.


I’m not sure if I’ve discussed this before but there’s a huge difference in culture between the two research groups I’m involved with. ILAQH (physics) meets once every four to six months at the moment and BRAG (statistics) meets fortnightly. At the fortnightly meetings we spend a few minutes having a quick check-in around the room where we discuss what we’ve been up to for the last two weeks (any papers published, talks given, new research plans) and then someone will give a talk on a topic that is within a semester-long theme. The theme last semester was visualisation, now we’re moving on to the analysis of big data.

ILAQH, by contrast, meets at a frequency that doesn’t really facilitate a catch-up. Instead, any big developments in the big projects are discussed and visitors are welcomed and/or farewelled appropriately. Both meetings serve different purposes in different ways.

One of the biggest benefits of the BRAG fortnightly meetings is that PhD students get a lot of practice giving talks. You’re almost expected to give a talk once a semester, and students are encouraged to give practice talks for the milestones of their PhD candidature (confirmation at 12 months, final seminar when completed). Given that it’s typically two and a half years between confirmation and the final seminar, being expected to give a talk every few months means that you get practice at talking and that the practice is somewhat regular. This will be especially important for ILAQH students as an overwhelming majority of our students are not native English speakers. Getting regular practice at academic speaking will help when it comes to final seminars.

I’ve spoken to Lidia Morawska (Director of ILAQH) about setting up such a system and she thinks it’s a good idea. Now I’ve just got to check that the rest of the group sees it as valuable and then put the legwork in to get it off the ground (organise a time, book a room, etc.).

Becoming a grown up (at least a grown-up academic)

There are only four months left in my current postdoctoral appointment and I’m discussing plans for next year with my supervisor. There’s still a lot of unfinished UPTECH work that needs to happen, including helping a handful of PhD students with the stats in their papers for their thesis, dealing with the clinical data and putting together some research plans for what to do next. The plan at the moment is to look for some funding both internally and externally to provide a research appointment. I’m also interested in continuing lecturing next year, whether in SEB113 or another mathematics/statistics unit.

Most of ILAQH is away as of today or tomorrow, as they travel to Prague for the 2013 European Aerosol Conference. The work that I’ve been doing with some colleagues from ILAQH and Italy, on personal sampling, will be featured on a poster. The paper has been submitted to ES&T but hasn’t been accepted yet, so unfortunately I can’t put a preprint up to show off the cool statistics that I had to learn to do the modelling in the paper. As a result of everyone being away, I’ll be one of two academic staff members left here. It’s going to be quiet, with most of the staff and a few PhD students gone.

Barring the Finnish paper that I’m still revising, this personal sampling paper has been the paper which has required the most creative and original programming as there have been many different steps along the way. I am particularly proud of this paper and the work that went into it. When I was first brought on board there didn’t appear to be much clarity regarding what we wanted to investigate; we had a lot of personal sampling data but didn’t quite know what to do with it. I think the paper we’ve developed does service to the amount of work that was put into collecting the data and is aligned with what the UPTECH project was set up to do. I’m grateful to all co-authors on the paper (and everyone who was out there in the schools) for the work that they put into bringing this to fruition.

I’m still finishing the final corrections for my thesis, due in a few weeks time. After that’s handed in I’ll be taking another step in becoming a grown-up academic: supervising a PhD student. I’ll be the replacement associate supervisor for a student whose original associate supervisor has moved from QUT to another university. QUT requires an internal primary and associate supervisor and I’m the one most familiar with the modelling that this student is doing as part of their thesis. We’ve already set a meeting schedule for the time when his primary supervisor is overseas and have discussed what sort of things I’ll expect to see. It’s a strange responsibility to have for someone who’s only just finishing up their thesis. I wonder how long it will be until I’m the primary supervisor for a student. Two years? Five? Ten?

Worrying about funding, writing grant applications, supervising students, lecturing (writing assessment!). It’s a strange place to find oneself.

Posterior samples – July wrap

I had some Posterior Samples to share before going on leave but didn’t get around to posting them. Here’s what’s been on my mind this month:

Maths and science units are popular with (Kentucky) students until they realise that they’re hard. While not directly relevant to the Australian university education model it’s probably an important thing for the Science and Engineering Faculty to keep in mind.

I’m looking at ggplot2 more these days, so the idea of a “grammar of graphics” is beginning to resonate with me. Here’s a talk about building one for clojure (which I don’t use).

Something for me to keep in mind when delivering SEB113 slides this semester is what your maths slides don’t need. Probably also good pointers for any PhD students graduating soon.

An interesting article in the New York Times about air pollution from cooking. This is something that ILAQH has a research interest in and our nanotracer paper contains a bit of analysis of inhaled surface area dose from particles that originate from cooking.

Another NYT article, this time with a delicious visualisation of the geographical trends in income disparity and social mobility.

Andrew Gelman writes at Slate about some of the problems with scientific publishing and the publication of spurious findings (which isn’t always willingly dishonest).

A special “Big Bayes Stories” issue of “Statistical Science” will be published soon, focussing on the real world application of Bayesian statistics where other methods were inapplicable. Christian Robert has written the preface; the issue is being edited by Robert, Kerrie Mengersen (one of my PhD supervisors) and Sharon McGrayne, author of “The Theory That Would Not Die“.

Also I went to Questacon and it was awesome.

On leave next week

I’m on leave next week so am attempting to spend today and tomorrow getting ready some things finished before I disappear. This means finishing my contribution to two papers by one of the UPTECH PhD students, finding a new home for our personal sampling paper (Environmental Health Perspectives didn’t want it), getting SEB113 material ready for our 67 students next semester, contacting collaborators interstate about data, getting some Bayesian Network elicitation typed up and helping various PhD students with their data anlysis. It has been quite full on despite my supervisor being overseas.

Ex Post Docto

I am now at the half-way point in my year-long postdoctoral fellowship. Since the start of the year the number of papers submitted with my name on them has approximately doubled, I’ve got my name on some grant applications (some successful, some not) and am co-writing a proposal for a PhD student position at ILAQH that I will likely end up co-supervising. It’s been quite the experience so far and it’s really only just getting started. I’ve been applying for lecturing positions at QUT and will continue to look further afield to see what’s out there that inspires me.

Today the staff and post-docs in my group had a meeting to discuss the handling of lab business and how we maintain our space. For the longest time I’ve felt like a bit of an outsider, particularly with regards to lab stuff. I couldn’t tell you how to maintain a CPC, what the difference between a P-trak and Q-trak is or how to compare SMPS to NSAM data. I was not a member of the measurement team for UPTECH and my involvement in the research is mainly through data analysis and statistical consultation. Being given responsibilities within the lab is still a bit strange to me but I’m very happy to be helping out, as I am part of a team and I rely on those around me for my work. Statistics doesn’t happen in a vacuum (unless you’re a probabilist).

A friend of a friend is finishing up their chemistry PhD and looking for work for next semester and beyond. They’re applying for a more technical job and we spent some time this evening with our mutual friend (who has recently started a postdoc) discussing how to rearrange the CV in order to best highlight their experience. I showed my own CV to explain what I wanted to highlight, as all three of us had different opinions on style, format, the flow of the text and whether to include a photograph or not. Obviously I’m pitching myself at academic, rather than technical, positions and I said that I believe professional experience in a lab is far more important to show off than academic awards from undergrad. An interesting moment in the conversation was the disbelief from the friend of a friend that I would make my CV so public. Why wouldn’t I want to show the world who I am, what I’m working on and what I’m interested in?

It appears that developing multiple versions of a CV is necessary in order to have something to send to different bodies. University faculties are looking for a very different set of attributes than the Australian Research Council or other funding groups. I haven’t yet managed to whittle my CV down to two pages but I suspect it would include removing much of the tutoring and undergraduate experience I’ve had, my radio and TV interviews and conference organisation background, focussing instead on my top publications, professional experience and track record with grants. I will continue to need to tweak my CV as I continue to apply for jobs and this means having a look around for good resources from those who have gone before me. Some examples of advice I’ve come across are:

All of those blogs are worth following anyway.

The next six months will be a challenge, as I attempt to juggle the remaining time in this postdoc with other commitments and the need to find ongoing employment.