Twitter in the Christchurch Earthquake, Pt. 2

Following on from my overview of key tweeting patterns in the Christchurch earthquake in February, in the previous post, I wanted to further explore Alfred Hermida’s idea of Twitter as ‘ambient journalism‘ – or more to the point, examine whether there are any obvious differences between Twitter’s coverage of the immediate earthquake and its long-term discussion of related events.

So, based on the activity patterns we’ve identified in the previous post, I’ve split the #eqnz data into two lots: from the event itself, on 22 Feb. 2011, to 00:00 on 25 Feb., and from 00:00 on 25 Feb. to the end of our dataset (which for the purposes of this analysis runs until 8 March 2011). That’s necessarily an arbitrary division – but 24 Feb. was the last day on which #eqnz activity reached more than 1,000 tweets per hour on multiple occasions; after that, the Twitter discussion continued at a much reduced level (or, as is also likely, diversified into a number of additional hashtags which we’re not tracking here). (Note that there was another brief spike above 1,000 tph in the evening of 1 March, though, as another major aftershock rattled the city.)

If we think about the days immediately after the event as the time when Twitter’s function is most likely to turn from ambient into actual, first-hand, citizen journalism (as well as secondary news sharing), and the longer period after those days as a time during which Twitter use is less news-oriented, then: is there any evidence of this difference in the data? Let’s look first at the major Twitter contributors during those periods:

 
Some major differences here, obviously. I’ve already mentioned in the previous post that the official Canterbury Earthquake account @CEQgovtnz, which I believe was set up after the previous earthquake in September, didn’t commence posting again until 24 Feb., so it doesn’t feature in the first period we’re looking at here; it is clearly the most active #eqnz account in the time after 25 Feb., though, by some margin. In second place is a similar account, @eqnz_live, which is run by the volunteer group CrisisCampNZ and also provides a constant stream of recovery advice. Further down the list, the City Council’s @ChristchurchCC and the Ministry for Social Development’s @msdgovtnz also feature.

By contrast, the list of most active tweeters in the immediate aftermath is dominated by a mix of individuals and news organisations – for the most part, providing first-hand news updates or retweeting what information they have come across. So, this would seem to support the ‘ambient journalism’ idea: in the immediate aftermath, Twitter users’ focus turns to making sense of what’s happening, by sharing information and amplifiying (through retweets) the material coming from key news organisations; later, attention turns from news to more practical advice on how to respond to and recover from the disaster.

Let’s illustrate this a little further, by looking at the amount of @replies (including old-style retweets) sent to Twitter accounts by #eqnz participants:

   
Here, too, we see some obvious differences. Before 25 Feb., it’s news organisations like @nzherald and @NZStuff who get the most @replies (including, I’m willing to bet, a vast amount of retweets as users share the breaking news being posted by these organisations); additionally, interested individuals who are active in passing on information about what’s happening – such as @anthonybaxter or @georgedarroch – also receive substantial numbers of @replies or retweets. There are also some immediate advisories (especially from @TelecomNZ and @vodafoneNZ) getting a substantial number of retweets, of course – and a few more general messages of support (especially that from @stephenfry) are also highly visible.

After our 25 Feb. cutoff, things look somewhat different. The top nine spots on the list are all occupied by official sources of various kinds – from @CEQgovtnz and @ChristchurchCC to @nzcivildefence and @NZRedCross, with a few key media organisations also still prominent in the mix. In fact, the only individual Twitter user in the top ten is @NZTopModelColin, one of the judges on the TV show New Zealand’s Next Top Model. The stronger focus on public advice and information during this second phase of the response is also documented by the presence of a number of other more specific sources: the Earthquake Commission @EQCNZ, the @BritishRedCross, the airline @FlyAirNZ, @ANZ_NZ bank, and @NZLotteries.

We should have no doubt that the @reply rankings are strongly influenced by retweets, of course – and this supports the point that #eqnz’s focus after the first few days of the disaster gradually shifts from sharing news updates to passing on more specific information that is of immediate use to locals trying to deal with the aftermath of the earthquake. This, of course, also means that the discussion on #eqnz is no longer of general interest, which explains the dropping number of posts using the hashtag. (We have no easy way of identifying where Twitter users are located – unless they are amongst the small percentage who geotag their tweets -, but if we did, I would expect that the geodata would also show the #eqnz community to be far more strongly localised to the Christchurch region during this second phase than it was in the hours and days following the quake itself.)

So much for part 2 – in part 3, we’ll look at some content patterns…