Trauma-related Cognitions: ‚Äčthe Role of Beliefs and Attributions in the Experience of Potentially Traumatic Events

Sissi H. Ribeiro

Advisor: Keith D Renshaw, PhD, Department of Psychology

Committee Members: Sarah Fischer, Rachel Goodman

Online Location, Online
November 29, 2023, 09:00 AM to 11:00 AM


Potentially traumatic events (PTEs) are very common, with around 90% of people experiencing at least one PTE over the course of their life. Most individuals who experience a PTE naturally recover after experiencing transient short-term negative reactions, but a small percentage go on to experience significant consequences, to include posttraumatic stress disorder (PTSD). Several models have been posed to understand determinants of post-PTE response, including how PTE-related cognitions may shape responses to such an event. 

The current project seeks to expand our understanding of cognitive mechanisms in this process, specifically attributions, or the causal explanations that people develop for an event. Attributions are generally categorized into five overarching dimensions: internal/external (something internal to themselves vs. an external factor or situation), stable/unstable (something likely to persist over time vs. something transient), controllable/uncontrollable (something under one’s control vs. not under one’s control), personal/universal (something that applies to only the person making the attribution vs. applies to most people), and global/specific (something that applies to most areas of one’s life vs. only applicable to one area). Numerous studies have shown that more severe PTSD symptoms are related to attributions that are stable, internal, uncontrollable, and global, but inconsistencies in results and assessment methods exist across studies, and few studies to date have included the personal/universal dimension. 

To further examine these issues, I collected data from 337 college students who had directly experienced, witnessed, or had a close family member exposed to a PTE, for which they identified at least one perceived cause for that event. Participants were recruited from February to December of 2020 and completed an online survey including several self-report questionnaires with some open-text measures, focused on participants’ PTE exposure, PTSD symptoms, and attributions regarding the worse PTE they identified. For this dissertation, I examined these data in two related but separate studies. 

In Study 1, I directly compared two methods of assessing PTE attributions. Both methods were based on asking participants to write, in their own words, up to three unique causes to which they attributed the event. For each cause they listed, participants were then asked to answer 5 Likert scale items, each assessing one of the 5 attributional dimensions on a scale from 1 (one end of a dimension) to 7 (the other end of the dimension). Later, the open-text responses were coded by research assistants who were trained to identify the same 5 attributional dimensions through an attributional coding scheme. After coding was complete, I conducted five one-way ANOVAs (one per dimension), comparing participants’ self-reported item score for that dimension across responses that were coded by the research team on one end of the dimension vs. the other end of the dimension (e.g., comparing self-reported scores on the internal/external item when the coding team rated a cause as internal to scores on the internal/external item when the coding team rated a cause as external). All five ANOVAs were significant, but inspection of the means revealed a much narrower differences of scores than was expected (internal/external difference = 1.92, stable/unstable difference = 0.49, global/specific difference = 0.77, controllable/uncontrollable difference = 2.27, universal/personal difference = 0.71). Subsequently, I created one response per participant for each dimension when a participant had listed more than one cause, by (a) averaging self-report scores for multiple causes together for the quantitative assessment and (b) having coders rate the entirety of all attributional statements together for the qualitative assessment. Similar to results for individual causes, all five one-way ANOVAs were significant, but the actual mean differences were narrow (internal/external difference = 1.46, stable/unstable difference = 0.53, global/specific difference = 0.54, controllable/uncontrollable difference = 2.36, universal/personal difference = 1.04). Overall, these two methods of assessment did not appear to provide highly similar information. 

To obtain a preliminary evaluation of the validity of the two assessment approaches, I examined the associations of the attributional dimensions with participants’ self-reported PTSD scores, via correlations and a multiple regression. Using the quantitative self-report scores from participants, I found that more personal, internal, and global attribution were associated with greater PTSD symptom severity, generally consistent with prior findings in the literature. In contrast, no significant relationships were found using data from the qualitative approach. Thus, it appears that how attributions for a PTE experience are assessed can produce different results, and the pattern of results in my study suggested that the quantitative approach may have more utility than the qualitative approach. 

In Study 2, I examined how trauma type (interpersonal vs. not interpersonal) related to different types of attributions, and if those associations might help illuminate the well-established finding that interpersonal traumas are associated with greater PTSD symptom severity than are non-interpersonal traumas, for survivors of PTEs. Given the results of Study 1, I used data from the quantitative measure of attributions. Participants’ target PTEs were categorized as interpersonal or non-interpersonal, according to guidelines from prior research, yielding a total of 141 interpersonal and 196 non-interpersonal target events. A significant one-way ANOVA confirmed that PTSD symptom severity was higher in those reporting interpersonal PTEs than in those reporting non-interpersonal PTEs. Five one-way ANOVAs comparing attributional dimension scores across interpersonal and non-interpersonal PTEs revealed that interpersonal PTEs were associated with significantly more internal attributions than external attributions and significantly more controllable attributions than uncontrollable attributions. No significant differences were obtained for the stable/unstable, global/specific, or personal/universal dimensions. Finally, I examined whether participants’ attributions partially mediated the relationship between trauma type and PTSD symptom severity. Results of a path analysis revealed a similar pattern of associations as detected in the one-way ANOVAs, with interpersonal trauma associated with internal and controllable attributions, and internal, controllable, and personal attributions associated with greater PTSD symptom severity. Notably, interpersonal trauma maintained its association with PTSD symptoms severity, suggesting that attributions did not account for variance in this association. 

Altogether, attributions seem to play a role in how survivors perceive PTEs and the subsequent impact of those events, and the method of assessment has a large influence on the empirical evaluation of this phenomenon. My results suggest that qualitative assessment of survivors’ prompted attributions for a PTE may not be as informative as survivors’ own self-report of those attributions; however, it is important to note that my assessment relied on prompted attributions, rather than spontaneous attributions. Further research is needed to understand how qualitative coding of spontaneously produced attributions might compare to prompted self-report of attributions. Results also suggest that those who were exposed to an interpersonal event are more likely to develop internal and controllable attributions for the event and report a greater severity of PTSD symptoms.