Social-Ecological Determinants of Affective States Among College Students: A Cross-Cultural Ecological Momentary Assessment Study
Abstract
Background: College students experience a complex and diverse array of stressors during their college life, making them vulnerable to severe mental health issues. Momentary affective states are linked to major mental health disorders and overall well-being; however, the current social-ecological model of affect lacks a temporal dimension, failing to account for within-person dynamics of affective states over time. This study disentangles the complex temporal dynamics among the momentary, multilevel social-ecological factors and affective states in the real-world setting of college students across the United States (US) and India. However, establishing invariance of the affect model across time, people, and cultures is a critical prerequisite for cross-cultural comparisons of the within-person dynamics. This dissertation examined the invariance of affect structure across time, people, and groups (i.e., US and Indian college students). Thereby, at the psychological level, this study investigated the role of momentary self-control in the association between state mindfulness and affective states; and at the social and environmental level, this study examined the joint effects of physical activity, and social and physical context on momentary affective states among the US and Indian college students. Cross-cultural differences were also examined in these momentary associations. Methods: The study comprised 218 college students (Mean age=19.99, SD=1.14) from the US (n=158) and India (n=60). Participants received ecological momentary assessments (EMA) 6 semi-random times per day over 7 consecutive days between 2021 and 2023 via a smartphone application. Momentary surveys measured positive affect (3 items), negative affect (4 items), mindfulness (3 items), and self-control (1 item) on a 0-4 Likert scale, along with current physical activity levels (lying down, sitting, standing, or moving), social context (alone or with others), and physical context (indoors or outdoors). Latent Markov factor analysis (LMFA) and multigroup multilevel confirmatory factor analyses (MMCFA) examined measurement invariance of affect structure; dynamic structural equation models (DSEM) investigated the role of momentary self-control; and multilevel structural equation model (MSEM) tested the joint effects of physical activity, and social and physical context on affective states. Results: LMFA and MMCFA of 5,820 EMA surveys suggested that the two-factor model of affect, with correlated factors (i.e., positive and negative affect) and residuals, had the best fit at within-person and between-person level. LMFA found that the model was invariant across time and people. MMCFA extended these findings by showing configural and metric invariance at the within-person level. DSEM revealed that mindfulness not only enhances affective states by improving self-control but also that affective states influence mindfulness by altering self-control levels; however, this mediating role of self-control was not different across the US and Indian samples. MSEMs revealed that higher-than-usual levels of being not alone (vs. alone) and moving (vs. sitting) were associated with both increased positive affect and decreased negative affect, while being outdoors (vs. indoors) and standing (vs. sitting) were associated only with increased positive affect. Regional differences in these independent effects indicated that while both regions experienced a lower negative affect being outdoors, the effect was more pronounced for Indian college students. Joint effects showed that students reported the lowest negative affect while engaging in higher levels of physical activity (i.e. moving) with someone (vs. alone), and while being outdoors with someone (vs. alone). Highest positive affect and lowest negative affect were also reported while sitting outdoors (vs. indoors). Conclusions: This dissertation extended the social-ecological model of affect by examining the intra-individual dynamics of affective states within a day in the real-world setting of college students across the US and India. The metric invariance of the two-factor model of affect at the within-person level allows a more reliable cross-cultural and temporal comparisons of affective dynamics. The within-person association of psychological, behavioral, and contextual factors with affective states can inform the timing, context, and key targets for just-in-time adaptive interventions. For instance, frequent practice of group mindful walking in outdoor settings may yield the maximum benefits for reducing negative affect.