« Making sense of the ALT-C change | Main | Key trends in education? - a crowdsource request »

September 10, 2010

Comparing Twitter usage at ALT-C - 2009 vs. 2010

I've just been taking a very quick and dirty (and totally non-scientific) look at the Summarizr Twitter summaries for #altc2009 and #altc2010. Here are some observations (which may or may not be either significant or correct):

Firstly, twittering was up roughly 35% this year on last - not surprising I guess (but definitely worth noting). Oddly this was despite a fall in the number of twitterers, though I think that might be explained by the fact that the older hashtag has been in use much longer? There is also a slight, but probably not significant, shift towards a fewer number of heavy Twitter users.

Secondly, #digilit, #falt10 and #elearning appear in the top ten list of tweeted hashtags this year, none of which appeared last year. I suggest that #digilit and #falt10 reflect a growing interest in digital literacy more generally and a greater awareness of the fringe event by the main conference respectively. Conversely, #elearning seems a bit odd (to me) since I would have expected that term to be disappearing? The #fail hashtag also appears this year but not last which might be taken to indicate some problems with technology (or something else)? On the other hand, #awesome appears this year but not last, which either indicates some success somewhere or the fact that we are all getting more American in our use of language? If so, shame on us!

Thirdly, the top 10 tweeted URLs this year include a majority of blog posts (there were none last year) which either indicates that people are tending to blog earlier (and more) or that there is more of a culture of re-tweeting posts to blog entries than there used to be (or both)?

Finally, the list of most commonly appearing words in the Twitter stream for both years is largely useless as an indication of what was being talked about. However, I note that 'mobile', 'learner' and 'lecture' all make an appearance this year but not last ('lecture' for obvious reasons, given the first keynote) whereas discussion about the 'vle' seems to be shrinking.

As I say, all of this is anecdotal and unscientific, so treat accordingly.

TrackBack

TrackBack URL for this entry:
http://www.typepad.com/services/trackback/6a00d8345203ba69e20133f4132633970b

Listed below are links to weblogs that reference Comparing Twitter usage at ALT-C - 2009 vs. 2010:

Comments

One suggestion I have been meaning to make is the need to make it easier to specify date ranges when analysing hashtag usage. In order to compare like with like you should probably compare #altc2009 and #altc2010 during the week of the event.

It is possible to specify date ranges in Twapper Keeper. Would it be possible to add similar filtering to Summarizr?

Hmm... have just discovered that the short cuts (in Mac) for Close (Firefox) window & Grab's "grab screen' are the same, as I've just found a web site that could answer (to a degree) the question of where people tweeted from. And when.

It's crowdeye.com & seems to go back a max of 14 days - and you can narrow things down to a 10 mile radius.

http://picasaweb.google.com/lh/photo/3MfGQJFEH1pTkK6p-z8khg?feat=directlink - the screen grab in the end didn't highlight the tallest column, saying it was 1,761 or similar tweets.
The number didn't seem to vary when I changed the location, but the people showing up did.

Edit: Change that link to:
http://picasaweb.google.com/lh/photo/QzXv09nSC9CY7W1GfnhVug?feat=directlink - for near Nottingham
http://picasaweb.google.com/lh/photo/y2DkhGgF8CDVbnDg2oShng?feat=directlink for anyone.

Not sure why it didn't think Josie & Steve were in Nottingham, could have sworn I'd seen them.

But, Ewan wasn't!

Edit: Change that link to:
http://picasaweb.google.com/lh/photo/QzXv09nSC9CY7W1GfnhVug?feat=directlink - for near Nottingham
http://picasaweb.google.com/lh/photo/y2DkhGgF8CDVbnDg2oShng?feat=directlink for anyone.

Not sure why it didn't think Josie & Steve were in Nottingham, could have sworn I'd seen them.

But, Ewan wasn't!

I have been following a few event hash tags recently and it occurs that RTs clutter up the hash tag column. Maybe event etiquette could request RT without hash tag thus propagating beyond event without clutter.

The comments to this entry are closed.

About

Search

Loading
eFoundations is powered by TypePad