I’ve been a huge fan of Crisis Text Line since it launched back in August, 2013. I remember when I first saw it when its was in its infancy as a project at DoSomething. Nancy Lublin, who was CEO at the time, hosted a book party for “Measuring the Networked Nonprofit” and during a tour of the office, she introduced me to Bob Filbin, DoSomething’s Data Scientist at the time. Nancy said, “Bob, show her the crisis text line data set.” He showed me the project and explained the vision for the future. At last year’s Data on Purpose Conference at Stanford, Nancy gave the keynote and I was blown away by what they have achieved in so little time.
But today, Crisis Text Line reached another huge milestone to celebrate! The organization announced they will be making their data available to approved researchers, who must meet stringent requirements, including that research will help more people in crisis. Currently, the largest open dataset on mental health and crisis is a survey run every other year by the Center for Disease Control (CDC). Over 13 million text messages have been exchanged via Crisis Text Line since its launch in August 2013, qualifying this as the nation’s largest open set of crisis data.
According to the press release, founder and CEO, Nancy Lublin, is making good on a promise she made publicly from the very beginning: that the data itself could save lives. “From day one, this was the goal: to help people one-on-one and leverage the data for smart system change on a broad scale,” said Lublin. “I’m pretty darn excited that we’re making it happen.”
For the first time, researchers will have access to a large-scale crisis data, which will hopefully lead to more-informed policies, journalism, and interventions across the crisis and mental health space. For example, researchers can answer questions like: what is the relationship between weather patterns and depression? Or, if a texter mentions “bullying,” do eating disorders or other issues arise? This data offers a new perspective that complements existing qualitative and quantitative approaches.
Crisis Text Line is offering three levels of data to researchers – conversation, actor and message level – with each level increasing in detail. The conversation level dataset will allow researchers to explore questions such as, “what crisis issues occur most on holidays, such as Christmas and Valentine’s Day?” and “how are bullying and depression related?” The actor level dataset will allow researchers to answer questions like “For a texter experiencing depression, how do issues fluctuate over time?”
Enclave data will build on Crisis Text Line’s existing open data offerings. For the last 18 months, Crisis Text Line has made aggregate data available at www.crisistrends.org. This has already informed journalism, local policy makers, and community groups taking action to help people in crisis. Ultimately, Crisis Text Line believes opening this data will lead to better care for people in crisis across the U.S. “We are tremendously excited by the potential of these datasets,” said Lublin, “and we hope it enables researchers everywhere to engage in groundbreaking discoveries.”
Researchers interested in Crisis Text Line’s data enclave can find the application here: