Twitter

Approved: 
Yes
Related Articles: 
Links to Web 2.0 Resources

Twitter

An extremely active part of the social media world is Twitter. Many digital projects and digital experts maintain Twitter accounts to increase their visibility and reach. How do you measure the impact of Twitter, however?

While there are several tools available, the appropriate tool depends in part on the quantity of tweets you anticipate needing to analyse.

 

TAGS

For low volume Twitter data (i.e. a few thousand tweets during your collection period), we have found the TAGS: Twitter Archiving Google Spreadsheet (https://tags.hawksey.info/) designed by Martin Hawksey to be an extremely simple (and free) way to create a spreadsheet of tweets. Hawksey maintains the code for TAGS, and so the latest information on how to set up and use the program is available on his site. Essentially, however, the steps are as follows:

  1. Follow the link to TAGS: https://tags.hawksey.info/
  2. Use the link on that site to open a copy of the spreadsheet in your own Google Drive account.
  3. Follow the instructions on how to authenticate to Twitter.
  4. Put search terms, hashtags, or Twitter handles in the search box.
  5. Run your collection to make sure it works.
  6. Follow the instructions on the site to set up your sheet to collect data periodically (once an hour, once a day, etc.), which it will do automatically even if you don't log in to your Google account. It will collect all the new tweets it finds, and they will be there waiting for you when you next open the sheet.

TAGS includes some analytic tools in the spreadsheet, or you can do you own analysis, either within Google or by downloading all the data to a program such as Excel.

 

NodeXL

NodeXL (http://nodexl.codeplex.com/) is another free tool that runs as a plug-in to Microsoft Excel which also allows you to import data directly from Twitter (as well as YouTube, Flickr, Facebook, and the Web). NodeXL was designed with social network analysis in mind, so it has many tools to do things like visualize your Twitter network as a network of 'nodes' (dots) and 'edges' (lines connecting the dots).

The NodeXL community has also created and shared many free resources to learn to use NodeXL and experiment with existing datasets. See http://nodexl.codeplex.com/wikipage?title=NodeXL%20Teaching%20Resources.

 

Twitter API

The Twitter API (https://dev.twitter.com/) provides access to a portion of the Twitter stream. Using specialized tools, you can access less than 1% of the Twitter stream. However, for most projects, the volume of tweets you would need to access would never exceed 1% of the tweets in any given day, so you would essentially be able to collect all the tweets relevant to your topic.

Tools to access the API of Twitter (or any other social media site with an API) are frequently written by developers for specific projects. An increasing number of developers are able to do this work, so it is worth considering whether your team already has access to someone with the expertise needed to build a small app that will gather all the data you need.

 

Twitter Firehose

The so-called Twitter Firehose (all the tweets sent) is only available from commercial sources. These include GNIP (http://gnip.com/) and DataSift (http://datasift.com/). As commercial providers, the cost of access can be expensive, so you should be sure to contact these sources before you plan your budget if you think you will need to include data from a very large number of past tweets.

Another company worth contacting is texifter (http://texifter.com), which was founded by an academic and may be able to negotiate academic deals for access to only those data from the Twitter firehose which are relevant to your needs.