Personal Information Management!

My goodness. I've just finished organising all the paper copies of literature that I've read as background for this project.

They amount to 2 whole lever-arch files full. Plus a big wallet of PhD print-out, and a folder containing my personal notes on the project (see previous post). That is about twice the volume of paper produced by the entire of the first half of my Masters course (the taught part!).

Of course the estimate is skewed somewhat because I got rid of paper copies of lecture notes that were duplicated online or articles I could get from e-prints etc. So they probably amount to almost the same volume of "printout-inches"!

I'm quite pleased that I've read about the same volume of stuff (measured rather crudely, to be sure) for the self-directed part of the Masters.

Well, my next step is to re-shape the outline of the thesis. This I hope to do by making up lots of cards with headings of varying specificity (oh, sorry, I mean granularity! Must keep up...) and then arranging them in some sort of logical order to demonstrate the narrative (of my argument, not the project chronology).

After that, back to the coding scheme (long sigh...) because I think it can be improved again! I dug up some more references on Information Literacy (that course was almost a year ago so I can be forgiven for forgetting) which shed light on the categories I want to use, and which I was on the verge of essentially duplicating! Gosh, what a lot can be gained (and what effort spared!) by just reviewing the past.

More piloting...

...although the dissertation has taken a back seat to my new job, I am back in the comfy chair and fiddling around with previous studies' coding schemes to try to make the perfect research instrument.

It's surprisingly hard work, as all the categories and sub-categories seem to blend into each other after a while and I find myself second- and third-guessing my decisions. I guess I am experiencing the famous "coder fatigue".

Still, the 2nd or 3rd (depending on how you define it) pilot study is going well.

The Pilot! Pt. 2

And... we have pilot study. I have my basic results and a newly-formed coding scheme. After all these weeks! It's definitely been a question of building up my confidence and ability to just "go for it", realising that research is meant to be a step into the unknown and it's OK to change things around.

Oh, well, off home now before they close the library. I'm quite enjoying this dissertation lark now.

Pre-pilot

I did a very rough-and-ready pilot with a new coding scheme including some oh-so-scary qualitiative categories... ouch.

I take back everything I've been secretly thinking about content analysis being an easy option.

I randomised my list and ended up picking a highly specialised blog which was not easy to categorise!

The categories for "objective" measures are pretty easy to allocate, but I seriously wonder whether even a suitably qualified coder would pick the same headings as me for the more warm and fuzzy categories.

Oh, well, more piloting tomorrow, it's my free day for university work. Ah, the bliss of a 9-5 routine with a bit of space for disserting.

Back in business!

Hurrah, it's been a good day!

My first day of my new part-time (30 hours per week is part-time?) job in a real live university library, and I've also put together a workable coding framework for the first pilot of my content analysis.

Now I'm off to choir practice for a nice change-is-as-good-as-a-break.

Freshness

Looks at first glance that most of the blogs are actually quite "fresh" - i.e. they have a post in the last 3 months, if not this month or even this week.

At least this criterion filters out pretty much all of the abandonned blogs. Maybe a stricter criterion or a random sample is needed to ensure feasibility. My time will be limited once I start work next month.

I'm beginning to lean towards using the same kind of scheme as Lee and Bates' paper, since that seems to be a pretty acceptable technique, even in a PhD paper. As long as I am applying their technique to a new dataset, a new geographical area and a new sector, I think I am very much producing something original.

First supervision

I had my first supervision this morning and I'm just writing up the action points which came out of it. Things seem a bit clearer now and I'm quite happy that I'm more or less on the right track.

I need to work on clarifying the statement of my research questions and getting them out of my head and onto paper. I was trying to link the actual content of the blog posts to SCONUL's information literacy model but hadn't developed any clear understanding of how to do that. I think it may be more possible than I expected, given the next point:

The sample is potentially huge, since it's more or less the whole known population of these blogs, correct to the current date! However, if I cut it down to only what's current, i.e. updated in the last month, fortnight or week, it could end up being quite manageable. Hence I would only be looking at a few tens of blogs, and a more in-depth content analysis, with clear categories and some synonyms maybe, could be feasible.

So I plan to brain-storm everything I am planning so far as a mind-map, assess previous studies' coding instruments for relevancy, and go through my current draft dissertation write-up, and maybe aim to do a pilot study of the currency of the blogs to see what size sample I get left with. If it's good I may try out the SCONUL model as a source of categories (somehow!).

Coding and coders

Having just received the coding scheme from the previous study on which this project is based, I'm quite happy that I can put something together once my supervisor finally materialises.

There are two poles I could aim at; either a large sample and a very statistical overview; or a small sample and a more qualitative investigation.

On the other hand, even with a small sample I could focus more on form (design features) of the blogs, or on their content (post topics and references).

Either way, feasibility is a major factor. One way to avoid the question almost entirely is to base my coding scheme mostly on what's gone before to aim for a good comparison; this would also mean a fairly design-oriented taxonomy which could make for less time-consuming coding.