When people use the microblogging service Twitter, they can connect the subject of their post to a stream of postings by other Twitter users on the same subject by using a “hashtag.” A hashtag is a word, number or any string of text preceded by the pound sign, #. Twitter automatically converts hashtags into hyperlinks, and clicking on a hashtag link leads one to a page filled with the most recent posts that also use that hashtag.

In the state of Maine, one hashtag used by journalists and citizen commentators is #MEpolitics, short for “Maine Politics.” What can be said about the people who’ve been using that hashtag lately? Are #MEpolitics users separate from one another or connected in a giant game of Telephone? Who’s most listened to? Who’s ignored?

To answer those questions, I’ve used the free and open-source NodeXL template extension of Microsoft Excel, with which it’s possible to load all recent Twitter posts (“tweets“) using a hashtag into a spreadsheet for analysis. NodeXL considers all “replies” and “mentions” of one account by another account, or “retweets” of one account’s post by another account, to be indicative of a social tie (or “edge,” in NodeXL’s network parlance) connecting two accounts. Being the recipient of a retweet, reply or mention is an indication of popularity, and the number of those retweets, replies or mentions received by an account is called “in-degree“.

Through a useful bit of coding, NodeXL takes an edge list of all the ties between Twitter accounts using a hashtag like #MEpolitics and converts it into a graphic image of a network (referred to as a “sociogram“).

In this sociogram, drawn from the 1,108 tweets using the #MEPolitics hashtag from August 24 to 28, 2012, there are 239 unique Twitter accounts. Each of these accounts is represented as a dot if an account’s in-degree is only 0 or 1 during the 8/24-8/28 period. If an account has an indegree of 2 or more, then it is represented with the account’s graphic icon. Of all 239 #MEPolitics tweeters over the past 5 days, 28 accounts merit the big graphic. These are the big shots of the network to whom others respond most often. Listed by username, they are:

If you live in Maine and pay attention to the Maine political scene, you should be able to to use the combination of graphic icon and username to tell who a number of these more popular tweeters are and with whom they’re conversing. What patterns do you see in this network?

There’s much more a body can tell from a simple list of tweets than I’ve touched on today. One could find central communicators in the #MEPolitics network, uncover groups who speak most often to one another, cliques who shut others out, bridges between otherwise disconnected regions of the network, and isolates who speak to and are heard by no one. One could look at the content of the tweets too. What sort of statements about are being uttered about Maine politics? Which tweets or phrases diffuse through the network, and which drop dead?

These sorts of observations will be the subject of future posts on the subject of analyzing tweets, but for now I’d like you to notice what draws these various kinds of measurements together. None of these measures rely on understanding the character of the individual account. All of them draw upon patterns of social relations. That is why the study of social media is sociological. That, in turn, is why a solid sociological footing is essential for those who are interested in entering the burgeoning social media profession.