Date of Award

8-2021

Level of Access Assigned by Author

Open-Access Thesis

Degree Name

Doctor of Philosophy (PhD)

Department

Psychology

Advisor

Emily A. P. Haigh

Second Committee Member

Matthew P. Dube

Third Committee Member

Jordan LaBouff

Additional Committee Members

Rebecca MacAulay

Craig A. Mason

Abstract

The majority of research on emotion regulation processes has been restricted to controlled laboratory settingsthat use experimental paradigms to investigate short-term outcomes (Berking & Wupperman, 2012). A true understanding of emotion regulation requires an unobtrusive, ecologically valid assessment of the construct as it naturally unfolds in the environment. Digital phenotyping, or moment-by-moment quantification of individual-level human behavior using data from smartphone sensors (Torous & Onnela, 2016), is a novel method for evaluating human behavior in naturalistic settings. The present project is the first to implement digital phenotyping in the investigation of emotion regulation.

The central aim of the study was to evaluate whether smartphone-based digital phenotyping data predicted individual differences in emotion regulation in both in-lab and naturalistic settings. During an in-lab session, unselected adult participants (N = 69) completed self-report questionnaires measuring trait emotion regulation as well as state affect/emotion regulation following a neutral mood induction, negative mood induction, and recovery period. Smartphone-based digital phenotyping data were collected during a 7-day follow-up period using the Beiwe Research Platform (Onnela & Rauch, 2016), an open-access mobile- and cloud-based research tool for collecting digital data via smartphones.

Results showed that variation in mobile power state level and GPS distance were significantly associated with variation in negative state affect and state cognitive reappraisal over time. Clustering and classification analyses showed power state level and GPS distance over time to accurately, and with high sensitivity and specificity, classify two trait emotion clusters. Variation in power state level and GPS distance together with trait and state emotion regulation was not associated with current depressive symptoms. Overall, the findings provide initial data on the use of digital phenotyping data in predicting individual differences in state and trait emotion regulation in both in-lab and naturalistic settings. The results suggest that operationalizations of digital phenotyping data and modeling approaches are particularly important factors to consider when implementing digital phenotyping methodology in the study of mental health processes such as emotion regulation.

Download

Share