The present study relates mental health to characteristics of work and family roles in a small sample of employed female clerical workers who are married and/or have children living at home. The main goals of the investiga- tion are (a) to examine the separate and joint influence of qualities of work and family roles on women's mental health, and b) to explore conditions under which the effects of work and family roles are maximized.

Chronic exposure to high levels of stress in childhood pose risk for mental health problems. However, the effects of mild daily stress on youth psychological functioning is poorly understood. The three studies in this dissertation utilize intensive repeated data (e.g., daily diaries) to examine how children react to and recover from minor negative events on the same day or the next day. The first study examined children’s emotion reactivity to and recovery from school problems, and assessed their cross-sectional associations with internalizing and externalizing problems in 83 5th and 6th graders. Study used repeated ratings of school problems and positive and negative emotion completed several times a day over five consecutive weekdays. Youths reported more negative emotion and less positive emotion at school and at bedtime on days when they experienced more problems at school. Youths who tended to report more negative emotion on stressful days at school had more symptoms of depression, even after controlling for average levels of exposure to school problems. Youths who tended to recover by bedtime had fewer internalizing problems. The second study examined same day and next day mood responses to school problems over the course of 40 consecutive weekdays in a sample of 47 8-13-year-old youths. On average, youths reported more negative mood and less positive mood on days that they experienced more school problems. School problems were not linked to mood on the next day. Children who tended to report more negative mood or less positive mood on days when they experienced more school problems showed more internalizing problems three years later when they were 11-17 years old. The third study used data from the same sample of 47 children to test the within-and between-person effects of daily negative events – peer problems, academic problems and interparental conflict – on diurnal cortisol, a physiological indicator of stress reactivity. Three indices of diurnal cortisol were derived from saliva samples collected four times a day across eight days: same day diurnal cortisol slope, same day bedtime cortisol, and cortisol at wakeup the next morning. On average, children who reported more peer problems showed flatter slopes of cortisol decline from wakeup to bedtime. However, children secreted more cortisol at wakeup following days when they reported more peer or academic problems than usual. Interparental conflict was not significantly associated with diurnal cortisol. In sum, this dissertation showcases a novel application of intensive repeated methods in developmental psychopathology research. Using this methodology, studies found individual differences in reactivity to and recovery from daily problems, which in turn were associated with youth internalizing problems.

The studies that make up this dissertation adopt process-oriented approaches to examine the mechanisms by which daily stressors result in disruptions to interpersonal relationships. To do so, all three studies made use of daily diaries or momentary survey assessment techniques to measure changes in stressors, cognition, behavior and affect across multiple time points. Each study also examined individual- and couple-level factors that were hypothesized to affect recovery from stressful events. The first study examined the day-to-day effects of difficult, highly demanding days on marital behavior over the course of 56 days in a diverse sample of 47 couples. The behaviors measured were marital anger and two types of withdrawal: hostile withdrawal and reductions in affection and disclosure, or “retreat.” A self-reported desire to withdraw from the family explained increases in marital anger, hostile withdrawal and retreat when participants experienced overload; negative mood, however, only explained increases in marital anger and hostile withdrawal. Husbands’ tendency to express anger or retreat on overloaded days was associated with poorer overall marital satisfaction. Using the same sample of 47 families, the second study examined the interaction between parents’ daily conflicts with each other and with their children. Marital conflicts predicted increases in negative parenting behavior, and parent-child conflicts predicted increases in marital anger. Negative mood partially mediated the majority of these associations, suggesting that additional variables may exist that explain the transfer or “spillover” of conflict across family dyads. Conflict spillover was exacerbated by children’s externalizing behavior and fathers’ neuroticism. The third study tested the hypothesis that certain emotional and behavioral responses to stressors impact whether individuals will go on to generate new stressors—specifically, interpersonal difficulties—in a sample of 137 college students. By assessing stressor occurrence, mood and behavior four times a day for five days, this study found that strong negative emotional reactions to stressors increased the likelihood of interpersonal problems later that day. This was particularly true when individuals engaged in avoidance while experiencing severe distress. Individuals with poorer psychological functioning reported more interpersonal stressors overall, but were not more likely to generate interpersonal stressors immediately following stressful events.

Positive emotions broaden our thinking and help us build resources (Fredrickson, 1998, 2001), and different specific positive emotions appear to have unique adaptive functions (e.g. Stellar et al., 2017). Yet, clinical science has typically focused on the reduction of negative emotions in treatment (Carl et al., 2013). Expanding the study of positive emotions, especially specific ones like compassion and gratitude, in relationship and clinical science can reveal new processes that support healthy relationships and improve individual well-being (Repetti & McNeil, 2018).This dissertation asked how positive emotions are linked to relationship functioning and mental health, by investigating daily expressions of compassion and gratitude, and increases in positive emotions, in the contexts of marriage and psychotherapy. The question is addressed in three studies which encompass two samples of married couples and one clinical sample, cross-sectional and longitudinal designs, and observational and participant-report methodologies. Using intensive repeated measures collected over 5 weekdays, Study 1 found evidence of daily synchronization of spouses’ expressions of gratitude, and compassion synchrony was moderated by marital quality. Compassion may require more effort to express than gratitude and therefore its reciprocation is only characteristic of higher quality marriages. In Study 2, naturalistic observations of couples from the Center on Everyday Lives of Families (CELF) were coded for expressions of everyday compassion. Couples offered compassion frequently, on average twice per hour. Husbands who perceived their marriages as lower quality offered more compassion; wives offered more compassion when their perceived marital quality was high but their spouses’ was low. Higher rates of compassion were also expressed by both husbands and wives in couples where wives reported more depressive symptoms and husbands reported more neuroticism. Personal distress appeared to be associated with more frequent expressions of marital compassion. Study 3 examined the trajectories of positive and negative emotions as well as treatment outcomes over the course of psychotherapy in a psychology training clinic. Both increases in positive emotions and decreases in negative emotions independently predicted concurrent declines in symptom distress and improvement in relationship functioning, suggesting both are indicators of therapeutic change. However, changes in positive emotions were also predictive of future improvements in symptoms and relationship functioning, suggesting they may be precursors to therapeutic change.

Daily interactions may function as opportunities for family members to impact one another in ways that replenish or deplete coping resources. Evidence of synchronization in affect, behavior, and arousal has been observed in parent-infant dyads and coregulaton of mood and physiology has also been demonstrated in marital partners. However, little is known about whether pre-adolescents and their parents may serve a similar regulating function for one another. The current study draws upon three repeated measures samples to test transmission of daily mood states and physiology in pre-adolescent children aged 7-13 and their mothers and fathers. Synchrony hypotheses predict bidirectional transmission between parent and child, and child and parent in both daily mood and diurnal cortisol. The hypothesis that mother-child dyads would show synchrony in daily positive mood was supported by all three samples. Father-child dyads did not evidence positive mood synchrony. Father-child and mother-child dyads evidenced negative mood synchrony in one of three samples. In both of the sample that included diurnal cortisol measures, pre-adolescents showed synchrony with both mothers and fathers. Overall, the current project supports a reciprocal view of parent-child influences in both physiology and mood during the pre-adolescent period of development.

While a robust literature links childhood exposure to stressful family environments, hypothalamic pituitary adrenal axis (HPA-axis) functioning, and mental and physical health outcomes, there is a paucity of research examining associations between parent-child relationship qualities and HPA-axis activity during middle childhood and adolescence. In addition, despite assumptions implicit in the allostatic load literature, little is known about the temporal stability of children’s diurnal cortisol and how children’s cortisol responses to acute stressors are concurrently related to their diurnal cortisol profiles. To address these gaps in the literature, two studies using multi-method, repeated-measures designs were carried out in an ethnically diverse community sample of 47 children aged 8 to 13. Study 1 examined naturalistic associations between two dimensions of the parent-child relationship (parent-child attachment and daily parent-child conflict) and two indicators of diurnal HPA-axis activity (children’s diurnal cortisol slopes and end-of-day cortisol levels) at the between- and within-person levels of analysis. Children’s reports of secure attachment to their fathers moderated the association between fathers’ reports of daily father-child conflict and children’s daily bedtime cortisol levels: Children who reported lower levels of secure attachment to their fathers had higher bedtime cortisol levels on higher conflict days compared to children who reported higher levels of attachment to fathers. Children’s reports of secure attachment to their mothers predicted their diurnal cortisol slopes such that higher levels of secure attachment were associated with flatter slopes. Study 2 used multilevel-model derived intraclass correlation coefficients (ICCs) to assess short-term temporal stability in four metrics of diurnal cortisol -- waking cortisol levels, the cortisol awakening response (CAR), diurnal slope, and bedtime cortisol levels -- over a range of 2 to 8 sample days. Associations between each diurnal cortisol metric and children’s reactivity to and recovery from the Trier Social Stress Task for Children (TSST-C) were also explored. Overall, children’s diurnal cortisol metrics were moderately stable, with highest stability estimates observed in bedtime cortisol levels and lowest estimates observed in the CAR. Overall, increasing the number of sample days did not improve stability. Better cortisol recovery from the TSST-C was significantly correlated with higher waking cortisol levels, while cortisol reactivity to the laboratory tasks was not associated with any of the diurnal cortisol variables.