#STFC in tweets

My twitter feed in the last weeks of the decade was often dominated by reactions to and discussions of the funding cuts to physics and astronomy research in the UK announced on 16 December. On the day of this announcement I created a permanent archive for all tweets hashtagged #stfc to keep a record of all that was said on the issue – as regular twitter searches only go back around 10 days. On the last day of the year I exported all the tweets going back to 10 December and I’ve been playing around with visualising them. Here are a few samples.

Here’s a plot of the number of tweets in the 12 days from 15 December to 16 December, binned per hour. The red lines indicate the approximate time of the announcements of the funding cuts on 16 December, and of the postdoctoral fellowships cancellation on 21 December. Click on the chart for a closer look and get the data here. At the height of the buzz, there were almost 120 tweets in one hour, and on the 16th around 500 tweets were tagged #stfc. And while the traffic has quieted somewhat now, the STFC crisis is still an active topic of discussion.

If you thought it was only a small group of people making a lot of noise, think again. Over 330 people tweeted about STFC in the whole 3-week period, and even the busiest tweeters contributed no more than 4% of the total. Click on each pie segment to see the usernames and numbers.

The archive is accessible to anyone and continues to aggregate #stfc tweets at the same location. I’ve placed the data file with the tweets I used to generate these images in a public location, as well as the idl script I wrote to process the data and generate the timeline (the script uses a few functions from the idl astronomy library). The pie chart was created using the Google Pie Chart gadget in Google docs. The word cloud was generated by Wordle, removing common English words plus “rt” and “stfc”.

Trackbacks

  1. [...] This post was mentioned on Twitter by Stuart, Sarah Kendrew. Sarah Kendrew said: New blog post, #STFC in tweets – http://tinyurl.com/y9wha6f [...]

  2. [...] purposes, and valuable information can be extracted quite easily. The post I wrote about it is here with links to datasets and idl scripts I wrote (yes, not the most useful of languages, I know) [...]