Social media data analytics: How to apply them in education


To date, social media data anaytics have mostly been used to help marketers refine their pitches as they get a sense of how they are perceived by the public. When properly processed and analyzed, the data can also lead to a wealth of insights to influence future curriciulums and lesson plans.

So far, social media data remains largely untapped in the education industry. At last April’s #EduAnalyticsDC conference, there was some discussion about social media data analytics but no serious proposals, according to my interview with Bill Rand, director of University of Maryland Center for Complexity in Business.

“There is some resistance in educational administration to thinking that [social media analytics are] a valuable signal,” Rand explained. “There isn’t really research being done in that space to the extent that there should be because of that resistance.”

Twitter signaling college choices

Rand conducted an experiment in which he monitored Twitter for four months in the spring to glean insights about the college decision-making process for high school juniors and seniors. Rand looked at 57 keywords including #collegedecision and #collegeopportunity, which netted some 10 million tweets. After rooting out the false positives, Rand narrowed the number to 1.1 million tweets for 25 keywords.

After running sentiment analysis, Rand found that “by far, most of the conversations around this topic were fairly negative” and indicated stress or uncertainty. The most positive tweets centered around being accepted to a college or referred to the community a college offered. The most negative referred to college visits and the food offered at the college cafeterias. Other insights included positive tweets related to college acceptance to LGBT lifestyles.

Social data applications for K-12

As a public network, Facebook and Twitter make the most sense for data analysis in the grade-school education space. According to a Pew Research Center study, 71 percent of teens age 13 to 17 use Facebook and 33 percent in the same age range use Twitter.

Despite the educational system’s reluctance to use such data, Rand thought of a few instances where it could be beneficial. Looking at national Twitter or Facebook data, researchers might be able to discern which math topics, for instance, are causing the most consternation for students and what their pain points are. Rand also noted that educators can narrow the search to their particular geographic area, such as within 10 miles of their school, to see what the students are saying.

“You could identify topics that are being discussed and see where people are experiencing frustration,” he explained. “It would be particularly useful in the context of curriculum changes. If you’re going to adopt a new curriculum like Common Core, you might be able to get feedback from students at a much quicker rate than you might by the traditional process of doing a lot of evaluations.”

Analytics use by colleges

Right now, most of the experimentation with social media analytics appears to be happening at the college level. Ithaca College in upstate New York, for instance, has been collecting social media data on prospective students since 2007 to use in admission decisions, according to The Hechinger Report. The goal is to use signals in social media, like the number of friends applicants have or how many photos they have on social media, to judge how successful a student will be in college.

Ithaca uses a Facebook-like site that it created to glean such data. It’s conceivable that a high school could do the same, providing an environment for students to sound off about topics that they might not be willing to discuss offline. Data analysis could also identify troubled students or academic issues that affect the entire student body. In addition, they could help determine topics of interest to help educators create more personalized learning programs. Social media data analytics is a wide-open canvas for educators who are willing to take a closer look.

Transform teaching and learning with social data analytics

Follow @IBMAnalytics

This entry was posted in Big Data. Bookmark the permalink.