As a former student of journalism, a current social media professional, and as somebody with an unhealthy appetite for numbers, I've found a way to satiate myself this weekend with an amateurish combination of all three.
Using a week of tweets from Sunny Hundal's @RipplaNews Twitter account (more on that here), I've drawn together a rather simple data-set to provide a base for examining what news makes an impact on the social sphere, and what encourages people to share the news they share.
Pie-charts (yum), bar-charts and other associated data-visualisations to follow:
Over-all seven-day share by news organisation:
I think it's fair to say that I'm somewhat surprised by this. I had expected the Daily Mail to have had a bigger share, and The Sun to be roughly on par with the Guardian. Bravo to the BBC for taking almost half of the social shares.
Daily number of trending tweets listed by news organisation:
Given the pie-chart above, this probably isn't terribly unexpected. That said, I would have expected a higher spike on Monday for the Guardian, given Charlie Brooker's weekly comment piece - but that might say more about my personal taste than anything else.
Number of article images measured against number of shares:
This scatter chart has been skewed somewhat by a single 40-image article. It featured four times as a trending news article. It was part of the Daily Mail's coverage of Whitney Huston's death. That aside, it is clear that the most frequent shares come with articles that contain between 0-2 images.
This is a misleading statistic though, given that few online articles (Daily Mail excluded) will contain three or more images, making it more likely that 'most-shared' articles would fall within this 0-2 image band. There are a range of variables here that I simply don't have the time to explore, but if you fancy it...
So, what does this mean?
I'm not really sure, although, if you're vaguely interested in this sort of measurement, please feel free to use the data (linked below), expand on it, improve it, and present it in a more sensible fashion
Get the raw data: You can view the raw data on Google Docs.
About this data: This data covers the top trending online news stories by measuring their social shares. The source for this data is Sunny Hundal's Rippla News application (http://rippla.com/). Details on how Rippla News works can be found here: http://rippla.com/ripples.php. This data has been recovered from the updates on the @RipplaNews twitter feed. This blog post (and linked data) is an example of a very simplistic way of examining the Rippla News data - It does not take full advantage of the rich data available on the Rippla News website.
Collection Timeframe: This data covers one week of retrievable tweets (78 Tweets), from 12:00pm Sunday 12 February to 12:00pm Sunday 19 February
Disclaimer: I make absolutely no claim that this is an exhaustive piece of reliable research, nor do I claim to be a researcher in any sort of professional sense. The data collection method is imperfect, and the classification of the different strands of information is , in parts, highly flawed (For example, the terms in the "Category" column) - This data is also reliant on what I would assume would be referred to as a small sample set. The intention was to catalogue trending news in order to gain an insight in to what news tends to trend, and to perhaps allow me (or anybody else that cares to explore it) to identify themes in the social share trends of online news.
Sources: All sources are clearly referenced by way of hyperlinks in the raw data.
Re-use of data: If you find this data to be in any way interesting, please feel free to re-use, refine and redistribute this data freely in part or in whole. A credit to @pdarigan would be nice, but is by no means necessary. I would ask that you do reference the Rippla News application, either by way of a link to the Twitter feed or website (That's where the real effort went in, and I can't claim any credit at all for that).
No comments:
Post a Comment