Authored by Ashley Figueroa.
In each therapy session, clients share many stories relating to their treatment goals. These stories can range from their family upbringing, past relationships, work-life balance, academic stress, trauma, self-esteem, etc. While therapists are trained to identify patterns and support personal growth, they remain human and inevitably interpret the clients’ stories through their own subjectivity. In some instances, therapists may either misinterpret or completely miss out on a key part of a client’s story. What if therapists had access to a program that was able to identify themes and patterns across months or years of session transcripts? Could we use Python to program a tool that would help bridge the gap between the therapist’s understanding and the client’s storytelling?
As a marriage and family therapist (MFT) in the MFT doctoral program at Florida State University, I am interested in exploring ways to increase the impact of therapy sessions. Marriage and family therapists do not only see married couples and families; they work with clients across a range of demographics. My research focus is on families with one or more children who have been diagnosed with a pediatric illness. What makes this field unique is how we see and think in relational terms. With a relational lens, we are looking at the broader context surrounding a client; we are also exploring their whole system such as their patterns of communication, family dynamics, life stages, and intergenerational patterns. Using relational concepts allows me to identify how clients are being influenced by and referencing other systems, specifically the connections between the client, the key people, and contexts in their lives.
When I learned about the Digital Research Incubator program, I was curious about how Python could be integrated into the therapy field. Can a math-based program work in human development and family science? I realized this type of programming functions in a similar way to what therapists do in the room. Using Python, we repeatedly look for and narrow down themes across transcripts, just as I connect common themes across multiple sessions with real clients. We are constantly navigating people and their emotions. As I thought more about it, I recognized how I show up in the room: humans naturally filter conversations and events through their lived experiences to make sense of the world. Therapists do the same, always checking in with clients while acknowledging that pieces of the conversation can still be missed. This is exactly where my proposed project comes into play.
For my project, I wanted to explore how Python could be used as a tool to support therapists in capturing, analyzing, and reflecting on the narrative content of therapy sessions, with the goal of improving treatment planning and narrative continuity. Ideally, this tool will help therapists strengthen their clinical skills and their therapeutic alliance with their clients.
Given that this is a sixteen-week program, it is important for me to have a structured schedule when learning the Python modules and to help complete this project. For my idea, I must collect data, clean the data, analyze it, and hopefully publish a paper. Right now, the data collection has been tricky because I need to make sure that the transcripts are HIPAA compliant. Instead of gathering transcripts from ongoing cases at the Center for Couple and Family Therapy (CCFT) at Florida State University, I decided to look for published transcripts. If I were to use ongoing cases, I would have to find a way to clean the original transcripts before analyzing them, which would take me a lot longer than if I used a preexisting transcript. Overall, my idea is a big one, so I think the best plan for me is to start small and work my way up. Being able to run a search like this would allow for enhanced recall of client experiences and may help the therapist see the larger trends in the development of this theme.
A challenge I have been running into is finding accessible transcripts that I can use for this project, and this has made the process feel a bit stalled at times. I have been going back and forth between transcribing YouTube video clips and getting in contact with the FSU Library to see if there are any databases I can use. At the same time, I have also been challenged by learning how to incorporate loops into my computational analysis of transcripts, especially when trying to identify the same patterns across multiple transcripts. At first, I kept asking myself, “How does this even make sense? What exactly is a loop?” To clarify, a loop is a piece of code that allows you to repeat a set of instructions multiple times. To fully understand it, I reread the module several times, consulted another AI source to hear the explanation in different words, and practiced creating loops within the module itself. During this part of the module, I was super frustrated with myself, but this was where I had to give myself grace for learning something new. Once I finalize the dataset that I will be using, I can focus on applying my code to the transcript and see what happens from there.
Besides the technicalities of Python, I found myself practicing a lot of patience throughout this process. I was confident going into this that I was going to pick up on these skills fast. In reality, I have had a difficult time trying to teach myself to code. It is like learning math all over again after five years of not using it.. This program has pushed me to my limit and stressed my perfectionist side to the max. Looking back at the past few weeks, I have learned to be more patient and give myself grace. Doing this has allowed me to be more receptive to the material and create a better learning environment for myself.
So far, I am grateful for this experience and the challenge it has given me. I have found myself stepping out of my comfort zone and trying things I would not have attempted before. In this program, while learning Python, practicing flexibility and curiosity has been key. These two skills are essential in my work as a therapist as well. Not every client will receive the same approach in the same way, which is why being flexible is a must. Curiosity also prevents me from making assumptions about my clients. Practicing these skills while learning Python has strengthened them in my work as a therapist. Since I am only halfway through the program, I hope to continue building my Python skills and make more progress on my project. In the end, I hope this tool will be an amazing application for therapists to access when they need some direction in their sessions.