Me, Myself, and I: MultiLevel Modeling of Self-Focused Language Use and Depression in the Context of Divorce

Research demonstrates that self-focus and self-focused language is associated with depression. However, little is known about the direction of causality between self-focused language and depression, nor how this association may operate over time. In this honors thesis I investigate how self-focused language, as operationalized by participants’ use of first-person singular pronouns during a divorce-related stream-of-consciousness (SOC) recording, may slow or speed mood symptom recovery following a marital separation; similarly, I examine how selfreported mood symptoms are prospectively associated with the use of first-person singular language. 105 recently separated participants (n = 105, 38 men) completed this study and made three SOC recordings about their divorce over a period of 9 months; the SOC recordings were then transcribed and submitted to the Linguistic Inquiry and Word Count (LIWC) computer program, which derives language use data across several dimensions. It was hypothesized that a reciprocal association between self-focused language and depression would be discovered through multilevel modeling analysis. Results indicate that first-person singular I Words have a lagged effect on BDI scores, but that BDI scores do not have a lagged effect on I Words. Discussion centers on this finding in the context of previous literature