At California State University, Monterey Bay, I work as a McNair Scholar and research assistant in Dr. Christine Valdez’s Trauma Outcomes and Recovery (TOAR) lab. The TOAR lab focuses on improving treatments for trauma survivors and understanding risk factors for posttraumatic stress disorder (PTSD). In my 3 years of experience in the TOAR lab I have contributed to a number of interesting projects.
One large ongoing effort in the TOAR lab is adapting a public speaking intervention, the Communities Healing and Transforming Trauma (CHATT) program, for use with former gang members in their communities. To achieve this goal the TOAR lab recruited a sample of former gang members (n = 32) to complete semi-structured phone interviews and a series of questionnaires to better understand the unique needs and concerns of this hard-to-reach population. As part of this project I reviewed completed interview transcripts for accuracy, transcribed audio from other interviews, and completed qualitative data analysis. I was trained in qualitative research methods like grounded theory and phenomenological thematic analysis. I used these new skills to identify themes in the interview data around posttraumatic growth and barriers for participants in receiving mental health services. Through this coding process I became aware of the logistical and cultural barriers that prevented this population from receiving mental health services. My experience on this project taught me that effective interventions should be not only mechanistically efficacious, but also appropriately acceptable and accessible to the people they aim to help.
From 2018-2020 I was heavily involved in a study investigating the use of virtual reality (VR) as a perspective-taking tool to enhance empathy in students and aid psychology educators in achieving social-justice major learning outcomes set by the American Psychological Association. Students (n = 81) enrolled in a psychology service-learning course were recruited to complete a 10-15 minute VR experience designed to “simulate” a path to homelessness (“Becoming Homeless: A Human Experience”). The VR experience takes users through job loss, eviction, anti-homelessness ordinances, and harassment. Participants filled out self-report surveys measuring affect and attitudes regarding homelessness before and after the VR activity, changes in these measures were assessed with paired sample t-tests. Our team found that this brief VR activity had moderate to large within-person effects in self-report measures which suggested enhanced empathy relative to pre-VR levels. The students also overwhelmingly reported increased empathy and a new desire to help people experiencing homelessness in the open-ended responses. My role in this project involved scheduling participants, running the study protocol, training new research assistants, contributing to data analyses design, writing all of the R code, and leading the manuscript writing process. While surely limited by lack of long-term follow-up, this project was an exciting proof-of-concept embedded in a live college course pointing the way to more VR pedagogy and social justice interventions.
More recently during the COVID-19 pandemic, I have focused on developing my statistical toolkit while contributing to two projects that utilize clinical data sets from an urban community clinic.
The first project examined the relationship between exposure to traumatic events, self-attitudes (e.g. self-criticism and shame), and psychiatric symptom severity in a sample presenting for PTSD treatment (n = 73) and in a second sample presenting for hoarding disorder treatment (n = 81). In this first project I taught myself advanced missing data techniques (full information maximum likelihood and multiple imputation), introductory structural equation modeling, and the R package lavaan using free online materials so that I could independently write and run path analyses in R. Our research team found that self-attitudes like self-criticism and shame had stronger associations with hoarding disorder symptom severity in the hoarding disorder sample than these same self-attitudes had with PTSD severity in the PTSD sample. On the other hand the total number of traumatic experiences was more strongly related to symptom severity in the PTSD sample, as compared to the hoarding sample. These results suggest that self-criticism and shame may be especially important in the etiology, maintenance, and treatment of hoarding disorder relative to PTSD.
A second project investigated how trauma-related cognitions at treatment intake are related to rates of PTSD symptom change over 16 weeks of therapy in a community clinic (n = 56) using a Bayesian multi-level growth model. Trauma-related cognitions were measured across 4 dimensions at an intake appointment: overaccommodation, accommodation, assimilation, and optimism. Overaccommodation refers to an excessive and over-generalized adjustments to beliefs in response to experiencing trauma (e.g. “I have no future”). Assimilation is when traumatic information is made to fit into existing belief structures (e.g. “I knew better than to do what I did” ). Accommodation occurs when traumatic information/cognitions adjust belief structures in a balanced way (e.g. “Sometimes bad things happen to good people”). Optimism measured the degree to which participants endorsed positive cognitions (e.g. “Most people are basically caring”). Our findings suggest that levels of overaccommodation and assimilation reported at intake had sizable relationships with the average rate of PTSD symptom change over the course of treatment in our sample. A 1-point increase in overaccommodation mean score was related to a ~30% slower rate of symptom improvement. On the other hand, a 1-point increase in assimilation mean score was related to a ~20% faster rate of symptom improvement. I found this project to be especially rewarding as I was able to independently learn and apply many new-to-me statistical techniques like longitudinal multi-level modeling, multiple imputation for multi-level missing data, and Bayesian approaches. I really look forward to learning more about Bayesian approaches to data analyses. I appreciate the ability to incorporate prior knowledge and make direct inferences about parameter values. Bayesian approaches also seem useful in clinical research where sample sizes can be low. Finally, I valued the opportunity to learn more about more cognitive approaches to PTSD treatment like Cognitive Processing Therapy.
Over the last 3 years, working in the TOAR lab has fostered my passion for intervention development, given me valuable research skills, facilitated my scholarly independence, and introduced me to fantastic friends and colleagues!
Trevor M. S. Stevens 