Eventyr

WIP-PPIG

(Belated in organisation, but now done.) 

The WIP-PPIG is a workshop for students and established researchers to present ongoing work in the psychology of programming and will be at the University of Sussex on 21-22 February 2008. The call for papers for the PPIG Work-in-Progress Workshop 2008 is now online.  Submission of extended abstracts (1 to 2 pages) is due by 14th January 2008.

"A feature of the PPIG workshops has been their openness to a wide spectrum of concerns related to programming and software engineering, from the design of programming languages to communication issues in software teams, and from computing education to high-performance professional practice. Similarly, PPIG entertains a broad spectrum of research approaches, from theoretical perspectives drawing on psychological theory to empirical perspectives grounded in real-world experience."

More information available from the workshop website.

HCT Postgraduate Workshop

Thursday 6th December was the Human-Centred Technology Postgraduate Workshop that we* organise at the University of Sussex.

This year things were done a bit differently.  Rather than traditional research paper presentations the day was re-focussed on research methods (by year) appropriate to everyone's level and interests.  The idea was to take full advantage of the functionality of the Sussex Creativity Zone (CETL) and provide the chance for all attendees to critically reflect on the research approaches they are taking.

Unexpectedly I found out the week before that I was to take part as well.  I am glad I did - even if my digital poster did suffer somewhat from being generated at the last moment - as it was enjoyable and useful.

So, on the day, I finished class early so that we could all get to the workshop.  We arrived during Ben's talk on 'Doing a PhD: Managing Your Supervisor', and I quickly had to jump up again to make refreshments for the first tea-break.  The day was then split into alternating sessions; digital poster presentations by year, and practical sessions targeted to level. 

In my case:

• Activity Session 1: Speed dating for the 3rd year doctoral students, elevator pitches for presenting your research.  In my case I repeated to each person I sat with: "I have no idea what I am doing!"  We then analysed the challenges and the shared positives. 
• Activity Session 2: Drawing pie-charts of our work life balance, looking at the work we completed in the previous week.  Mine was depressing, but I expected no less having started to deliberately reflect on this anyway.  (Nothing like having to teach time-management to make you notice just how badly you are doing.)  We then drew out the drivers and some general tips for the group. 
• Activity Session 3: Pairing up to sketch out our ideal thesis: type, contribution, thesis and structure.  I found this useful, even though it is a bit premature for me.  The shape could be anything at present.  Afterwards my supervisor happened by and commented that at least I knew I needed an abstract, introduction, literature review, discussion and conclusion!  Ho hum.

All in all a really useful day for me, and probably a good motivator for helping me to clarify just what I need to be doing at this stage in my game.

Challenges:

• Thinking time / tangible output
• Constraining the focus / area of study
• Maintaining a focus on the positive / not beating yourself up
• Barriers/difficulties (Technology, tools, pragmatics, financial)
• Pathological patterns - hard to break
• Theory
• OOPB / Too much to do
• Temporal and technical constraints
• Transition between disciplines to make contributions

Positives:

• Novelty
• Freedom to explore
• Found method to tell the story despite lack of clarity in literature
• Knowing where you are going and what you are doing
• Lots of potential

Drivers:

• Self-pressure / last minute
• Deadlines (internal/external)
• Anxiety
• Other people's agendas, requests and expectations
• Structure
• Priorities
• Problem/Barrier
• Boundaries
• Urgent vs. important

Tips:

• Explicit thinking time
• Explicit downtime
• Calming activities - i.e. list making
• Assure yourself that you can do it
• Talk about your problems
• Breaking up big tasks

My photos of the day are here.

----

* Thankfully this year I was allowed to take a backseat organisational role and was only responsible for registrations, payments and receipts.

Research Methods in CSAI

Second time round and the Research Methods in CSAI course I lead for incoming DPhil students in Informatics is now over.  As ever the students in this class were a pleasure to work with - I don't think I shall ever miss working with those behavioural-problem teenagers that kicked-off my teaching.

We just managed to squeeze in a class in the CETL (probably my last, as the space is now booked up for teaching right through to January 2009) and had lovely talks from Blay Whitby (on research ethics) and Mike Herd the director of the Innovation Centre.

If I do this again next year I need to remember not to presume knowledge in advance.  I was astounded to discover that I had made the mistake of presuming that everyone knew the distinction between qualitative and quantitative research.  It presumed a lot of underlying philosophical knowledge that CS students might never have been exposed to.  My shock when I realised was tangible, and resulted in some restructuring of the course.

Now all I need to do is face the marking (and possibly spend some time designing clear lesson plans, based on my experiences, to follow for next time).

Agile 2008

I thought I should write a quick note to say how excited I am about the 'stage' concept at Agile 2008.

"Agile 2008 has adopted the metaphor of a music festival that provides multiple stages to attract audiences with common interests. The stages within our program are designed and organized by experts (acting as stage producers) who are truly passionate about their particular areas. Each stage will have a feel of a smaller, focused mini-conference whilst providing the conference attendee with a wide choice of stages to choose from."

I will be helping out with some small aspects of Brian Marick's Designing, Testing, and Thinking with Examples Stage. This particularly appealed to me after my experiences with teaching, and fits well with the model used at the Sussex Creativity Zone (CETL).

"...it won’t merely accept the need for examples, it will glory in them as one of the primary ways we learn, teach, communicate, test, design, code, and decide how to act in the world. The stage is therefore open to any kind of session that puts the concrete example front and center."
and
"We encourage risky sessions. If your session has a chance to succeed spectacularly, we won’t mind that it might fail disastrously. The success of the stage will be more about how high the peaks were than what the average session ratings were."

I am rather excited to see how well pushing for concrete examples, as the logical step beyond experience reports, works in practice.

I don't know yet whether or not I can attend, but I really hope so.

However beautiful...

As mentioned in my last post I am now entering the major analysis phase of my research, a phase set to last me a full year.  I have over fifty hours of transcribed audio (interviews and meetings) which I now need to sit down and face.

• December is the narrative extraction phase.  In order to focus on narrative elements in the data, these structures first need to be extracted, broken down into component structures and tagged accordingly.  By extracting narrative structures it is possible to have a clear unit of analysis, as is preferred in qualitative analysis, and limiting an otherwise nebulous and unwieldy dataset.  As mentioned narrative in this instance is being recognised as a discursive element (whether told, recounted or hypothesised) around a causally-linked set of events (i.e. with a temporal structure), whether true, fictitious or partly told.  The data gathered supports a consideration of both types of narrative; the contrast is between stories told in interview (typically event narratives following Labovian structure) and those co-constructed in meetings (shared social stories co-constructed typically following a ‘small story’ structure).
• January to March is the open-coding phase.  This is (fingers-crossed) where some clear hypotheses can be developed and a taxonomy of stories collected will be developed.
• March to May and June to August are the two following analysis phases where hypotheses can be explored more deeply in the data, and some nice story network analysis can be conducted. (Informatics likes it when you can produce visualisations...)

The overall approach directly supports multiple-viewpoint analysis of the topics under investigation and allows for cross-comparison of stories told in interviews and meetings (i.e. under different interactional circumstances).  It also lends itself to a focus on the different types of narrative found in different circumstances, which may have an impact on our understanding of developer coordination in different meeting structures. 

It is a beautiful strategy and a big pile of work.  I have taken the decision to upgrade from my trusty N6 to NVIVO 7 which will be better for visualisations.  I look forward to learning this new software as well.

I look forward to being able to see this work take shape, such that I can see past the methodology to the results.  I have taken heed of my fortune cookie quote from HackDay:

    "However beautiful the strategy, you should occasionally look at the results" (Winston Churchill)

Christmas Geek Dinner - Year of the Social

It was lovely to attend the Christmas Geek Dinner.  I was able to arrive late and be surprised about what had been arranged.  Bliss.

I'm remaining nothing but glad that I took the sensible route and deliberately stepped back a bit from co-organising the girl geek dinners.  I was severely over-committed, spread thin.   Although I had put a lot of work into setting up the dinners, now they're established and my time is precious, the time came to redistribute my time more effectively.

The night itself was great fun, with an interesting talk from Aral*, lovely food, gifts and a fun raffle.  It was a tribute to all the hard work put in by Simon, Devi and Rosie.

Although it was strange to not be involved, especially given my tendency towards volunteering for everything, attending as normal attendee was a sheer pleasure, and I hope to do so again soon!

---

* Amused that 2007 really was the year of the social, both locally (more socials than ever) and globally (more focus on social and communication media).

Mentoring

After much thinking I have accepted the mentoring that is available to me as a CommercialiSE Fellow. 

Recently my old boss pointed out that I would gain a lot more focus on, and complete more quickly, my PhD if I had some clarity about what I was going to do with myself afterwards.  It is apparently a  great motivator, and can help with that little problem I have with trying to appease too many groups at once.

There were three clear options:

• Go out get a job, make money
  • Appealing.  I have been a student or employed researcher for many years now, and money is suddenly something that I am aware I would like...
• Do consultancy and/or build a business from the consultancy and any output from my PhD
  • Risky but potentially rewarding.  I lack a clear idea of how to develop a business from what I am currently doing, but have some entrepreneurship training.
• Stay in academia
  • My original plan.  A relaxed and varied work environment that will tax me, and never bore me, for the rest of my career if I chose.  Academia is one of the few  true 'career' paths remaining these days.  I already have experience in research and teaching (even if I don't love the latter I know I am more than adequate at it).

I'm still not decided, but on meeting my potential mentor I realised that she was someone who would challenge me.  This is something I have been lacking recently, and the idea of someone who is capable of keeping me in check was fascinating.  As it stands any business I could build would not be scalable, but it may be that something could be productised over the next year and a half from my work.  We shall explore and by the end I will know whether this option is a workable one.

Our agreement stands: as long as I do the work.

Next I'll be venturing into the heady world of market analysis.

Reviewing my Research

So I defended the next stage of my research as challenged.  Kinda.

I clarified a lot of things which I had forgotten to say and learned some important lessons:

• Be explicit
• Focus

I shaped an analysis plan for this coming year which effectively showed I need to:

• Work very hard
• Focus

Unfortunately some issues are still unresolved.  As I am looking for my hypothesis to arise from my data I am currently in the worrying position of being able to produce a thesis suitable for social science and of interest to practitioners, but that in itself isn't necessarily a passable informatics thesis. 

This could be a serious problem, especially if my data proves less amenable than I hope.  Sadly my data doesn't support incremental analysis - it is all or nothing, so I cannot fail fast.

I know I can do this, but I am mildly amused that I have effectively just argued that I be allowed to conduct waterfall research.  Single loop: Gather, Analyse, Deliver.

I have also promised to focus.  This means cutting back on all my other commitments and getting on with my research for a change.

We shall see next Easter how well this has all succeeded.  Until then, wish me luck.

d.Construct

In trying to reflect on d.Construct 2007 I fear I have less to say than I would like.

From the volunteer’s perspective, a role I have filled many times at many other conferences, it was by far the slickest organised conference I have ever attended.  Smoother even than the ones I used to organise back in the day.  I’m glad to have helped for the little I could.

As a conference it also had the nicest T-shirts.  This must be true as I actually plan to wear mine again.  The gift bag was also darned good.  These things so often get overlooked.

I remember that I was on a high from having my first pint of cider in a month the night before.

I remember I was jet-lagged, having flown back from the states and burning man the day before.

I remember the microphone not working just when I needed it to. 

I remember really enjoying the talks.*

I remember drinking afterwards.

I remember drinking some more.

I remember wondering if I should have had some dinner.

And now I seem to have a notepad full of drawn faces, but I don’t recall who drew them.

This probably all summates to a very good conference (v.g.c.).  Shame I don’t seem to have much in the way of notes.

---

* I unfortunately clashed slightly with the talk that related to my research area, which somewhat wound me up through not mentioning all the masses of work being done in the area.  It wasn’t a bad talk, in fact it had incredibly well-done slides, just not for me.  Couldn’t be helped.

I did however develop an instant crush on Matt Webb for being the first person in my recollection to ever mention Grice in a presentation.   I wish I had not just hit jetlag nap-time so I could have focussed more on the talk overall as it seemed potentially really fascinating; sadly sleep was battering at my concentration.

Annual Review: My Challenge

Good: Dissemination
Bad: Clarity

That’s the long and short of it.  I’m doing well, we think, but my lack of clarity about my research aim at this stage is still very worrying.  For us all.

Teaching research methods means I am doubly aware of what a horrific state my research is in.  How can I have a full data set and no hypothesis?  What about hunches? What would be the shape of my thesis?

[Intro]
[Literature]
Magical gap which somehow forms:
  [Methodology]
  [Exp 1]
  [Exp 2]
[Conclusions]

Is this the time to re-read all of phd comics to find where I am in the process?

Review which papers have shaped my thinking?

Write an abstract for my research?

Sketch out thesis shape?

Clearly I need to plan for a conference paper which will capture the story of analysis.  I need to break down the next year of analysis; without a testable hypothesis I will need to consider my approach carefully. Analyse a part of the data, or analyse the full data set on a partial axis or theme?  What would such themes be?  Do they entail having already done a partial analysis?  I need to break this down.  If someone wanted to replicate my approach what instructions might I leave?

Ugh.

The following faux-abstract is the sticking point:

“Are agile processes reflected in their social aspect?  A study of developers’ stories in a company applying agile software development practices.  Through their narratives we consider:

• Do agile practices shape stories?
• Do these stories show communication of reflection, willingness to change, other core values?
• How do we interpret programmer narratives?
• Etc.”