Written by: Josh Rosenberg
Primary Source: Joshua M. Rosenberg, November 3, 2015
This week, Spencer Greenhalgh and I are attending the AECT conference for the first time. We’re here to present on some Twitter research we’re doing (among other things), so we thought it would be fun to take a quasi-real time look at what people are tweeting. We’re collecting tweets from the hashtags #aect, #aect15 (the official hashtag), and #aect2015 using TAGS—the Twitter-Archiving Google Sheet&mdash. It’s a pretty handy tool that lets you collect Twitter data without too much hassle. You can see the collection of tweets here.
This post focuses on the official first day of the conference—Tuesday, November 3rd—though our TAGS sheet has actually collected tweets as far back as October 25th. There have been a number of workshops and meetings today, but things are still getting started, so let’s start with something pretty straightforward: a wordcloud. Spencer and I both use R for our statistical computing needs, so we downloaded our AECT tweets, cleaned the data a bit, and ran some wordcloud code that we had handy (here’s a link to the code if you’re interested; we borrowed much of it from Chin-Hsi Lin). The code then gave us the following wordcloud; like all wordclouds, the bigger words are those that appeared more frequently:
As Spencer and I expected, we don’t have a lot to work with after just one day of pre-conference activities (just 170 tweets before removing duplicates, and the wordcloud only registers words that appeared at least 5 times). However, there are still some cool takeaways from this. Some of them we can use to learn about what people are tweeting about at AECT, but others are more useful as lessons on things to do (or avoid) when working with data like these, which may come in handy if you’re interested in exploring Internet research methods.
First, we see a few Twitter handles in there. Now, the wordcloud isn’t snagging the usernames of the people sending out tweets, so a Twitter handle only shows up if someone is being mentioned or retweeted a lot. This is fantastic, because anyone can be a prolific tweeter, but it takes a little more to become an influential tweeter. So far at AECT, it looks like Chuck Hodges, Trey Martindale, and Tonia Dousay are among our top contenders.
Second, we can look at some common words for today’s tweets. “Classrooms,” “edtech,” and “research” tell us some about what we’re hoping to accomplish at AECT, and words like “today” and “see” might mean (as we would expect) that people are just now arriving and are excited to see familiar (and new) faces.
Finally, what are “election” and “orange” doing on here? Well, it turns out that we don’t have a monopoly on the hashtag #aect. I haven’t figured out quite what it is, but there was an election somewhere in New Zealand in late October where the acronym AECT seemed to play an important role. This is another thing to keep in mind when doing Twitter research based on particular hashtags: Spencer and I have inadvertently collected a whole bunch of Danish tweets about food thanks to overlapping hashtags.
Latest posts by Josh Rosenberg (see all)
- Getting started with ‘open science’ through blogging - October 1, 2017
- A person-in-context approach to student engagement in science (article in JRST) - September 29, 2017
- Comparing MPLUS and MCLUST output - September 14, 2017