4 - (SYM 44) Clinical Takeaways Differ Across Nomothetic to Idiographic Models: Comparing Networks of PTSD Symptoms, Emotion Dysregulation, and Alcohol Use in Women Experiencing Interpersonal Violence

A longstanding problem in clinical research is the trade-off between nomothetic methods, which examine interindividual variation between people, and idiographic approaches, which focus on intraindividual variation (i.e., N=1). While nomothetic information can identify intervention targets that may be broadly applicable to groups of patients, it can fail to capture person-specific heterogeneity in clinical processes. Recent improvements in the accessibility of idiographic analytic tools have led to increases in the use of idiographic methods. However, the extent to which key clinical takeaways can differ across nomothetic to idiographic frameworks is unclear. These considerations are particularly relevant in studies of underserved and minoritized populations, as complex relations between variables may not generalize across individuals with different intersecting social identities. In the present study, we employed temporal network analyses to examine bivariate associations among PTSD symptoms, interpersonal violence, negative and positive emotion dysregulation, and alcohol use in a diverse sample of women experiencing interpersonal violence (N=106; M_age=40.56 years; 40.3% white, 31.3% Black, 11.8% Hispanic or Latino/a/x, 8.3% American Indian/Alaska Native, 5.6% multiracial; 63.2% unemployed). Participants completed an experience sampling protocol of three daily interactive voice technology calls for 30 days (T=90). We estimated network models representing a spectrum of fully nomothetic (multilevel graphical vector autoregressive model [mlVAR]), hybrid (group iterative multiple model estimation [GIMME]), and fully idiographic (N=1 graphical VARs) analyses to compare how clinical takeaways differed across the analytic frameworks. The nomothetic mlVAR identified a group-level model with several significant paths, with highest centrality for negative emotion dysregulation and interpersonal violence in the temporal and contemporaneous networks, respectively. In contrast, GIMME and the N=1 models did not support an overall group model and indicated that most participants had their own unique network structure. Thus, results suggest that personalized interventions may be needed to address complex associations between interpersonal violence and related mental health concerns. This study highlights that clinical conclusions can vary significantly across the spectrum of nomothetic to idiographic models. Considerations for study designs in complex data and representative samples will be discussed.