How Can Visualization Help Everyday Twitter Users?
The role of visualization for increased knowledge on Twitter.
Social media has seen a proliferation of use in the past decade, with giants such as Twitter being one of the main ones. With its growth, so has the need for clarification of topics in Twitter social media threads grown. While it may provide a cognitive thrill for a user to scroll through endless threads of Tweets, it is often difficult to discern what exactly a user is looking at or what the context or meaning of a singular tweet is without context clues.
Social Media Threads and Topic Confusion
First, what is a “social media thread”? A social media thread is one where people can share posts or comments about common elements. For this study, the common element of the Twitter threads was a video of a vehicle driving through a crowd of protestors in Buffalo, NY.
In social media threads, many different people can leave comments about various parts of the common element. In the case while the common element is the vehicle driving through a group of protestors, many comments could be about other topics such as police brutality or injuries.
Many of the comments in a thread cover different aspects or topics related to the original tweet. This makes it hard for users to determine the topic of individual posts in a social media thread. Therefore the inevitable question arises, how can the context or meaning of a single tweet or reply be made clear?
So how can individual tweets or replies on a thread be made clear?
To answer this question, Franziska Huth and her research team embarked on a study to figure out whether a word-sized visualization can help users understand the meaning of a single tweet or reply without any additional information. The team studied 3 different ways of showcasing tweet topic areas. These were, in order of simplest to the most complex, text labels (1), categorical color coding (2), and word-sized visualizations (3).
Understanding the Different Approaches
For context, it is vital to understand the difference between the three different approaches that the researchers tested. The first approach shows the number of topics with a text label that can then be clicked on for more information, and the second approach uses a background color for each tweet to show what topic the tweet covers. The second approach also requires a user to click on the tweet to see what color the topic represents.
The third and most complex approach is a bar chart that is laid horizontally. To read the bar chart, one can simply look at the chart to determine which topics are most common in the tweet because if more of the chart is one color, then it means that the topic is more common than the other ones. A user can also hover over the bar chart to see more information if they want to. It is important to note that each approach is not that large, meaning that it does not take up more than a small paragraph of text.
How Researchers Determined the Best Visualization for Users
As seen in the screenshots above, the team tested various approaches to solving this problem. While each approach diverges in its own way, with each being significantly different in appearance, each approach attempts to figure out what works and what does not when trying to give Twitter users an overview of topic diversity on Twitter.
To figure out how each of these approaches compared, an online study was conducted with 64 participants, and each participant was assigned one of the three different approaches. Many of the participants were familiar with visualizations (77%) and between 18–27 years old (81%). Additionally, 55% of the participants were Twitter users already. The results of how well they identified the topics quickly and accurately of the tweets were then compared.
What Were the Researchers’ Findings?
While the study attempted to find that word-sized visualizations (3) had a clear advantage over the other two approaches, this was not found. The researchers found that there was no clear advantage of word-sized visualizations in this study. However, this does not mean that word-sized visualizations are not helpful in all cases, and more research is needed.
Summary
Twitter is an excellent platform for quick and easy access to information. However, the meaning and topic of tweets can be lost in social media threads. Researchers attempted to create different solutions to solve this problem, including word-sized visualizations that showcased topics of tweets and other representations of topic diversity. Despite the researchers’ goal to show that word-sized visualizations are the best way of showcasing topic diversity, this was not the case meaning that the researchers’ other approaches were just as effective.
This blog post is inspired by the following paper:
- Franziska Huth, Miriam Awad-Mohammed, Johannes Knittel, Tanja Blascheck, Petra Isenberg. Online Study of Word-Sized Visualizations in Social Media. IEEE EuroVis 2021
Click here to view a PDF of the paper.
