Abstract

Anxiety has long been associated with changes in attention processing, emotional salience detection, and emotional affect. Traditional methods of measuring anxiety frequently utilize self-report symptom measures, behavioral performance measures, and psychophysiological indicators. The current study integrates behavioral accuracy, self-report measures of psychological distress and well-being, and event-related potential markers of attention processing into an exploratory multimodal psychological network analyses to identify meaningful patterns across these indicators. The final sample of nonpatients included 36 anxious (low depression) and 24 control (low anxiety and depression) participants who completed an Emotion Attention Blink task while EEG data were collected to assess early, more automatic (P100) and later, more evaluative processing (P300) of emotional information within the context of an Emotion Attention Blink task. Exploratory analyses included zero-order, partial-correlation Gaussian Graphical Models, and collapsed-node network models designed to reduce redundancy and improve network stability. Network robustness was evaluated using bootstrap stability procedures, and supplementary EMOTION-ATTENTION RELATIONSHIPS IN ANXIETY machine learning analyses were conducted to identify variables with prominent discriminatory features across groups. Across increasingly refined psychological network models, several organizational patterns emerged that differentiated anxious and healthy control participants. Zero-order and partial-correlation networks did not demonstrate significant differences in overall global network structure or strength. Group differences emerged within the collapsed-node network, which revealed significant differences in network structure despite nonsignificant differences in global strength. Across network models, qualitative network organization, modularity trends, node-level metrics, and edgewise comparisons consistently suggested greater integration among psychological distress, well-being, attentional processing, and behavioral performance variables within anxious networks. In contrast, healthy controls demonstrated relatively greater differentiation among emotional, attentional, and behavioral domains of functioning. These findings suggest that anxiety may be characterized by altered organization of emotional, attentional, behavioral, and psychophysiological systems, with greater integration among these domains relative to healthy functioning. The results highlight the potential value of multimodal network approaches for identifying organizational features and potential intervention targets that may not be apparent when these systems are examined independently.

Date of publication

Summer 8-5-2026

Document Type

Dissertation

Language

english

Persistent identifier

http://hdl.handle.net/10950/5133

Committee members

Sarah Sass, Premananda Indic, Lauren Kirby

Degree

Doctoral of Philosophy in Clinical Psychology

Download

Share

COinS