The Best and Worst Public Transit Systems, According to Twitter


transit-twitter-ft

Jens Schott Knudsen/Getty Images



If Twitter’s good for one thing, it’s complaining. It’s the easy, free way to convert quiet muttering into griping everyone and anyone can hear, without the need to scream out loud and brand yourself a lunatic. So it’s no surprise that when Lisa Schweizter, an associate professor at USC’s Sol Price School of Public Policy, started collecting tweets for a study on how people talk about public transit agencies, she found gobs of evidence to work with.


The results of her study, published this month in the Journal of the American Planning Association, ranked 10 of the largest public transit agencies in the US and Canada by how well regarded they are on Twitter. Based on Schweitzer’s “mean sentiment score” and more than 60,000 tweets collected between 2010 and 2014, Twitter was nicest to Vancouver’s Translink, which was followed by Portland, Oregon’s TriMet and Toronto’s TTC. The harshest tweets concerned systems in Chicago, Philadelphia, Boston, and New York. For comparison, Schweitzer calculated scores for public figures (the sentiment score ranged from William Shatner to Osama Bin Laden), airlines, police departments, and welfare programs (the full chart is at the bottom of this post).


Schweitzer used text mining to pick out positive and negative words from the tweets (and manually added terms including brokedown, wtf, scam, epicfail, pervy, and unsuck). Machine learning helped spot things like parody accounts and unusually frequent tweeters. Schweitzer and her graduate students also analyzed some 5,000 tweets by hand, to ensure they lined up with the computer system’s interpretations. Reasons for complaint included delays, facilities, staff conduct, public mismanagement, and the class, race, and gender of other riders.


Here’s the funny thing: The transit system’s scores don’t line up with service quality (judged by on-time performance). But the unsurprising fact that public griping doesn’t necessarily match reality doesn’t make the data useless. Because Schweitzer did find one factor that predicts “mean sentiment”—the way the transit agencies themselves behave on Twitter.


Tweets from the people running subways and buses in New York, Los Angeles, Washington, D.C. and other cities weren’t included in the data set, but Schweitzer took them into consideration, and found they matter a great deal. Some agencies use Twitter to pump out impersonal blasts announcing service disruptions. Others make the effort to respond to user complaints and interact with them. That second approach, it turns out, makes a major difference in how the agency is perceived online. “Transit companies that respond to other social media users have statistically more favorable opinions expressed about the transit agency for just about every measure I considered,” Schweitzer writes.


The clearest example is provided by Philadelphia’s SEPTA, which in late 2011 started a @SEPTA_SOCIAL account for dialog with riders. After one year, Schweitzer found, its score went from -0.3 to nearly 0 (on a scale from -1 to 1). That’s a 70 percent improvement.


So what’s the takeaway? If you’re looking for a low investment way to improve your public image on Twitter, use Twitter as a tool for conversation, not one-way communication. It may seem that someone complaining to 18 followers that their train is late doesn’t matter, but Schweitzer makes the point that social media does influence broader public perceptions.


“If planners seek to support strong public transit systems as a key element in building equitable and sustainable communities, they should encourage positive public sentiment about the service, in part by encouraging public transit agencies to use interactive social media approaches.”


Or, you know, make the trains run on time.


CorrectedFigure1

USC Sol Price School of Public Policy




No comments:

Post a Comment