Integrating Ecological Momentary Assessment and Passive Sensing to Investigate Environmental Determinants of Psychological Well-Being: A Systematic Review of Methods and Procedures

Background and Purpose: Current understanding of the environmental determinants of psychological well-being primarily stems from studies adopting a static approach with healthy individuals. Methodological challenges emerge when we attempt to explore person-environment interactions in real-time, particularly in understanding how these interactions vary among individuals with various health limitations. We conducted a systematic review to synthesize the literature on how EMA and passive sensing has been used and integrated to examine the environmental determinants of psychological well-being, identify the methodological strengths and limitations,and provide recommendations for conducting inclusive EMA research.

Methods: We performed the search in PsycINFO, CINAHL, MEDLINE, PubMed, and Web of Science to identify studies that utilized EMA methods to examine environmental determinants of psychological well-being. Risk of bias assessment was conducted with a recently developed assessment rubric specifically tailored for evaluating the methodological strengths and weaknesses of EMA studies. Data were extracted and analyzed for three primary domains: EMA administration, prompting design, and passive sensing technologies.

Results: We found that most of studies are conducted with healthy individuals while the EMA methods and procedures exhibit a wide variability. 27 studies (90%) are smartphone-based, and 20 out of these 27 studies administered EMA through a mobile application. 77% of the studies integrated EMA and passive sensing technologies such as GPS tracking and accelerometer. Regarding prompting strategy, more than half of the studies adopted semi-random signal contingent. Study duration and the frequency of prompts vary, ranging from 2-30 days and 3-8 prompts per day. We also identified different reminder mechanisms and latency across studies.

Conclusions and Implications: We highlighted the importance of conducting inclusive EMA research that yields reliable and valid results with minimal participant burden. More thoughtful considerations should be given to the design of each EMA component in future studies. The use of technology should prioritise improving data collection efficiency and reducing participant burden over solely gathering extensive data from multiple sensors. A checklist outlining key considerations and questions to guide EMA study designs was provided to facilitate future inclusive EMA designs for enhanced participant compliance, improved data validity and generalizability of findings, therefore informing policy effectively.