Using Technology to Elevate Therapeutic Practice
When asked about his current research, Research Assistant Professor Mike Tanana can barely contain his excitement. And if you know Dr. Tanana’s philosophy on life and work, it makes sense. There’s a quote from Steven Dubner, of Freakonomics fame, that Dr. Tanana tells his PhD students over and over—“People who love what they do have a ridiculously unfair advantage.” In the end, they will be better at that thing because they love it, because they’ll be thinking about that when they could be thinking of something else.
So what is it that Dr. Tanana loves? Among his greatest passions are using technology to automate feedback, creating data visualizations to make outcomes more accessible, and machine learning. And he’s involved in the perfect project to bring all that together. With funding through the National Science Foundation, he is developing systems that automatically provide performance based feedback for psychotherapists. Essentially, Dr. Tanana and the team of researchers he works with have developed algorithms that have learned how to rate therapists for basic therapeutic skills, like asking open ended questions, using non-confrontational language, reflecting, and displaying empathy.
Why do this? Because therapists need feedback to improve their skills.
While a mechanism does exist to provide therapists specific feedback on how to improve their therapeutic skills, the process is so laborious, time consuming, and slow that the only people who do this are researchers. Even for therapists who want to improve their skills, Dr. Tanana explains, “there’s very little to guide behavior in one direction or another.” Many other highly skilled professions have immediate feedback loops. “If you ask a surgeon which tool causes more damage to tissue, the surgeon can look directly at the tissue and see which physically caused more damage.” But with therapy, that feedback loop doesn’t exist in the same way. “In the real world, this type of feedback doesn’t exist,” says Dr. Tanana. “And when there’s no feedback, people will assign meaning to ancillary, meaningless cues.”
This digital technology is meant to fill that gap, to systematically break down the session in order to show how well these skills are being performed. Ultimately, the goal is much more about skill development and skill practice. “This isn’t meant to be a punitive performance evaluation,” says Dr. Tanana. “We want to hone and build the craft of psychotherapy.”
Though this research is still in its early stages, Dr. Tanana explained that the implementation of the process is going well, and he’s eager to see where the research leads and what new issues he’ll get excited about next.
A view of a sample recording and playback system with automatically generated transcript:
Machine-generated feedback from that same session: