Article Critique

Exploring the Complexity of Mothers’ Real-Time Emotions While Caregiving

Margaret L. Kerr University of Wisconsin–Madison

Hannah F. Rasmussen University of Southern California

Katherine V. Buttitta Claremont Graduate University

Patricia A. Smiley Pomona College

Jessica L. Borelli University of California, Irvine

Existing research suggests that parenthood is both emotionally rewarding and demanding, yet little work has examined multiple facets of parents’ emotions. The current study examines the complexity of parents’ emotions by examining the intensity, variability, and emodiversity of mothers’ positive and negative emotions across caregiving and noncaregiving contexts. Data were collected from 136 mothers of young children using ecological momentary assessment (EMA) to measure their real-time experiences of positive and negative emotion during a 10-day period. Results demonstrated that mothers reported higher intensity in positive emotion, and greater emodiversity in both positive and negative emotion when caring for their children compared with times when they were not caring for their children. As the first study to explore the complexity of parents’ real-time emotional experiences, this work has important implications for interpreting the existing literature, and for developing interventions that enhance parents’ emotional experiences in the service of improving the quality of the parent– child relationship.

Keywords: parenting, emotions, intraindividual variability, emodiversity

Nearly 80% of adults become parents at some point in their lives; most consider it the most important thing they do (Erickson & Aird, 2005). Although parenthood is a nearly universal experi- ence, it is not well understand how caring for children influences emotional experiences. Existing research has found that parent- hood is associated with a variety of psychological outcomes (Nel- son, Kushlev, & Lyubomirsky, 2014), yet its impact on the dy- namic characteristics of emotion—including the intensity, variability, and diversity— has received little empirical attention. Although emotions have been examined primarily as stable trait-

like characteristics, emerging evidence suggests that emotion dy- namics may vary within individuals and across contexts (Koval & Kuppens, 2012). In this article, we extend this work to mothers of young children by exploring how dynamic characteristics of emo- tion vary across caregiving and noncaregiving contexts.

Emotions in Parenthood

The majority of extant research on parents’ emotions focuses on global levels or retrospective accounts of emotion (Nelson et al.,

X Margaret L. Kerr, Department of Human Development and Family Studies, University of Wisconsin–Madison; Hannah F. Rasmussen, De- partment of Psychology, University of Southern California; Katherine V. Buttitta, Department of Behavioral and Organizational Sciences, Clare- mont Graduate University; X Patricia A. Smiley, Department of Psychol- ogy, Pomona College; X Jessica L. Borelli, Department of Psychological Science, University of California, Irvine.

Katherine V. Buttitta is now at Brazelton Touchpoints Center, Boston Children’s Hospital, Boston, Massachussets.

This work was funded by the Pomona College Faculty Research Grants (Jessica L. Borelli and Patricia A. Smiley), David L. Hirsch III and Susan H. Hirsch Research Initiation Grant (Jessica L. Borelli), American Psy- choanalytic Association Grant (Jessica L. Borelli), the Claremont Graduate University Dissertation Fellowship Award (Katherine V. Buttitta), and the

National Science Foundation Graduate Research Fellowship Program (Hannah F. Rasmussen). We thank the families who participated in this project and research assistants that helped collect and process the data. Other findings from the larger investigation from which this data originated have been disseminated through publications and conference posters and presentations. No prior analyses or publications include the same ideas or findings as presented in this paper. The current article is unique in its focus on dynamic characteristics of emotion, variables which have not been included in prior papers or presentations. The datasets used in these analyses have been made publicly available online at https://osf.io/uk8r2/ ?view_only�5322854efc684a7484c122e825029ad5.

Correspondence concerning this article should be addressed to Margaret L. Kerr, Department of Human Development and Family Studies, Univer- sity of Wisconsin–Madison, 1300 Linden Drive, Madison, WI 53706. E-mail: margaret.kerr@wisc.edu

T hi

s do

cu m

en t

is co

py ri

gh te

d by

th e

A m

er ic

an P

sy ch

ol og

ic al

A ss

oc ia

ti on

or on

e of

it s

al li

ed pu

bl is

he rs

. T

hi s

ar ti

cl e

is in

te nd

ed so

le ly

fo r

th e

pe rs

on al

us e

of th

e in

di vi

du al

us er

an d

is no

t to

be di

ss em

in at

ed br

oa dl

y.

Emotion © 2020 American Psychological Association ISSN: 1528-3542 http://dx.doi.org/10.1037/emo0000719

545

2021, Vol. 21, No. 3, 545-556

This article was published Online First January 9, 2020.

2014). More recently, however, researchers have begun to focus on parents’ experiences of emotion in real-time, with a particular focus on intensity of emotion. This limited body of research generally indicates that mothers experience higher mean levels, or intensity, of positive emotion when caring for or interacting with their children compared with when they are not with their children (e.g., Delle Fave & Massimini, 2004; Impett, English, & John, 2011; Nelson, Kushlev, English, Dunn, & Lyubomirsky, 2013). Increases in positive emotion extend to the timepoints following caregiving as well (Taquet, Quoidbach, de Montjoye, Desseilles, & Gross, 2016), and may be specific to certain types of activities, such as direct interactions or play (Nelson-Coffey, Killingsworth, Layous, Cole, & Lyubomirsky, 2019).

Importantly, the majority of existing studies (Nelson et al., 2013, 2019; Taquet et al., 2016) have not explicitly examined negative emotion. However, there is some basis to expect that parents also experience heightened levels of negative emotion when caring for their children. For example, in global reports of emotion, mothers and fathers report more anger, stress, and anxiety than nonparents (Deaton & Stone, 2014; McLanahan & Adams, 1987; Ross & Van Willigen, 1996). Similarly, parenting activities elicit more negative emotion than any other activity besides work (Kahneman, Krueger, Schkade, Schwarz, & Stone, 2004). Al- though studies in this area are limited, this research supports the notion that parents may actually experience heightened levels of both positive and negative emotion when caring for their children compared to when they are not with their children. This premise is supported by at least one study, which found that mothers reported higher levels of both positive and negative emotions while spend- ing time with their children (Offer, 2014).

Dynamic Characteristics of Emotion

Gaining a comprehensive understanding of parents’ emotions requires more than just understanding mean levels, or intensity of emotion. Emotions are dynamic in nature, in that they change from one moment to the next and fluctuate in response to environmental stimuli (Kuppens, Oravecz, & Tuerlinckx, 2010). Hence, asking parents to report on traitlike aspects of their emotional experience across a long period of time (e.g., a day) may offer a poor representation of their true, dynamic, and highly variable, experi- ence of emotion. Fortunately, ecological momentary assessment (EMA) procedures allow researchers to capture not just intensity of emotion in real time, but also other dynamic characteristics. In the current study we focus on two additional characteristics of emotion: variability and emodiversity. These two dynamic char- acteristics of emotion capture the range of emotions experienced in different ways. Emotional variability captures within-person changes in emotion intensity across a short interval of time (Eid & Diener, 1999; Nesselroade, 1991), whereas emodiversity examines variety and abundance of discrete emotions (Quoidbach et al., 2014). Exploring both variability and emodiversity, in addition to intensity, more precisely captures the complexity of parents’ ex- periences by allowing us to examine how emotions fluctuate and shift in response to caregiving.

Variability assesses the range in intensity of a particular emo- tion, or set of emotions (i.e., positive, negative), expressed over a period of time (Eid & Diener, 1999). In contrast to intensity, which assesses the average strength of emotions (i.e., mean levels),

variability measures the extent to which a discrete emotion fluc- tuates over time (i.e., average deviations from the mean). More simply, variability is a measure of emotion stability or consistency. It is typically assessed via the intraindividual standard deviation (iSD) of emotions assessed on a daily or momentary basis across a specific time period (Eid & Diener, 1999; Ram & Gerstorf, 2009). Most studies have examined interindividual differences in variability, finding that higher positive and negative emotion vari- ability are associated with worse physical health and psychological well-being (e.g., Gruber, Kogan, Quoidbach, & Mauss, 2013; Hardy & Segerstrom, 2017; Houben, Van Den Noortgate, & Kup- pens, 2015), as well as with psychopathology, such as borderline personality disorder (Trull et al., 2008). On the other hand, studies examining within-person differences in variability have demon- strated that negative emotion variability temporarily increases fol- lowing stressful life events, such as relationship dissolution or loss (Bisconti, Bergeman, & Boker, 2004; Sbarra & Emery, 2005), and that greater positive emotion variability in response to a romantic break-up resulted in better adjustment 1 month later (Sbarra & Emery, 2005). The latter finding suggests that in some situations, increases in variability may be beneficial and signal an adaptive reaction to environmental change. This is consistent with emotion theory, which suggests that emotions vary as a function of stimuli or demands of the environment (Frijda, 2007; Izard, 2009).

Emodiversity, on the other hand, is a more recent concept that refers to the relative variety and abundance of discrete emotions experienced over a period of time (Quoidbach et al., 2014). Whereas variability assesses the range in positive and negative emotion intensity (i.e., how emotion intensities fluctuate around their means), emodiversity captures the array of discrete emotions experienced (i.e., variability across categories). That is, emodiver- sity captures the richness (i.e., number of discrete emotions) and evenness (i.e., relative frequency or abundance) of emotion expe- riences. For example, a person with higher positive emotional diversity might experience several discrete positive emotions, such as joy, contentment, and gratitude on multiple occasions and at similar intensities, whereas a person with low emotional diversity may primarily experience intense joy, with few reports of other emotions or at low intensities. Emodiversity takes a unique ap- proach to emotion dynamics through its focus on variety in dis- crete emotions. Greater emodiversity is indicative of a richer and more complex emotional life and is thought to promote positive adaptation to situations and circumstances (Lindquist & Barrett, 2008; Quoidbach et al., 2014) as well as greater physical and psychological health (Benson, Ram, Almeida, Zautra, & Ong, 2018; Ong, Benson, Zautra, & Ram, 2018; Quoidbach et al., 2014). Experiencing many differentiated or discrete emotions may be beneficial for emotion regulation by informing more adaptive behavioral responses to environmental stimuli. Emotion theory also suggests that experiences of more differentiated emotions are more adaptive than more broad, global emotions (e.g., Barrett, Gross, Christensen, & Benvenuto, 2001; Schwarz, 1990).

Emotion Dynamics Across Parenting Contexts

The assessment of real-time emotions invites the possibility of examining how intensity, variability, and emodiversity of emo- tions differ across contexts. While the majority of extant studies of variability and emodiversity have focused on interindividual dif-

T hi

s do

cu m

en t

is co

py ri

gh te

d by

th e

A m

er ic

an P

sy ch

ol og

ic al

A ss

oc ia

ti on

or on

e of

it s

al li

ed pu

bl is

he rs

. T

hi s

ar ti

cl e

is in

te nd

ed so

le ly

fo r

th e

pe rs

on al

us e

of th

e in

di vi

du al

us er

an d

is no

t to

be di

ss em

in at

ed br

oa dl

y.

546 KERR, RASMUSSEN, BUTTITTA, SMILEY, AND BORELLI

ferences, more recently researchers have turned their focus to how emotion dynamics vary within-person across daily contexts (e.g., work vs. home; e.g., Adolf, Voelkle, Brose, & Schmiedek, 2017; Koval & Kuppens, 2012), based on the premise that changes in emotions are partially attributable to interactions with the environ- ment. Indeed, emotion varies or fluctuates in response to momen- tary events or stimuli such as social interactions or stress (Koval & Kuppens, 2012; Scott, Ram, Smyth, Almeida, & Sliwinski, 2017; Zautra, Berkhof, & Nicolson, 2002). Although a few studies have examined parents’ emotion intensity when caring for their children compared with other activities (Kahneman et al., 2004; Nelson et al., 2013; Offer, 2014; White & Dolan, 2009), no studies have compared parents’ variability or emodiversity across contexts. Given that emotion dynamics are an integral aspect of emotional well-being (Trull, Lane, Koval, & Ebner-Priemer, 2015) and par- enting presents a variety of environmental stimuli and stressors (Deater-Deckard, 2004), understanding how emotions may vary across caregiving and noncaregiving contexts is an important area of inquiry. Further, exploring multiple dynamic characteristics of emotion may offer a more nuanced assessment of how parents’ emotions vary across caregiving and noncaregiving contexts.

Parents serve as protectors, teachers, and caregivers for their children, and these diverse roles place demands on their own emotion regulation abilities (Rutherford, Wallace, Laurent, & Mayes, 2015). As parents guide children in developing their own emotion regulation skills (Eisenberg, Cumberland, & Spinrad, 1998), they may need to regulate both their children’s and their own emotions, making parenting a particularly complex emotional experience. For example, a mother may experience greater emo- tional variability as her negative emotion steeply intensifies upon identifying a negative emotion in her child. Similarly, a mother may experience greater emodiversity as she recognizes her child’s fear about separating from her; in reflecting the fear back to her child, she may experience both sadness and guilt about his emo- tional experience, and her role in it, while simultaneously feeling pleased to have recognized his feeling so that she can respond to it, as well as excitement and gratitude for having some time to herself away from her child. These examples illustrate emotional complexity in parenting, and the challenges that emerge in parent- ing situations, particularly those involving responding to children’s distress when parents are also experiencing their own negative emotions. In fact, studies show that parents often suppress or amplify their own emotions or emotional reactions to benefit their child, at the expense of their own well-being (Le & Impett, 2016; Rasmussen et al., 2017). Emotional complexity may be revealed by assessing variability and emodiversity, in addition to intensity, in emotional experience across caregiving contexts, and by com- paring it with differences across other close relationship contexts, such as romantic relationships.

Current Investigation

In the current study, we examine differences in intensity, vari- ability, and emodiversity of mothers’ emotional experiences across contexts by comparing moments of caregiving (when mothers reported being with their child or children) and moments when they were not caring for their children (children were not present). We aim to extend existing work on parents’ emotional experiences by (a) examining parents’ emotions as they occur in real-time; (b)

assessing both positive and negative emotions; (c) examining multiple characteristics (intensity, variability, and emodiversity) of mothers’ emotions across contexts; and (d) conducting exploratory analyses comparing emotion dynamics when parents were and were not with their romantic partners.

For Hypotheses 1 through 3, we compared real-time positive and negative emotions in mothers during moments of caregiving and noncaregiving. Based on previous work showing heightened levels of positive and negative emotion in parents (e.g., Deaton & Stone, 2014; Nelson et al., 2013; Offer, 2014), we predicted that mothers would report higher intensity of both positive and nega- tive emotion when caring for their children compared to having no children present (Hypothesis 1). Second, based on preliminary evidence suggesting that parenting is an emotionally complex experience (Eisenberg et al., 1998; Rutherford et al., 2015), we predicted that parents would report greater variability (Hypothesis 2) and emodiversity (Hypothesis 3) in both positive and negative emotions when caring for their children.

In addition, to determine whether these effects were specific to parenting versus other close relationships, we conducted an addi- tional set of analyses comparing intensity, variability, and emodi- versity of positive and negative emotions during occasions when participants were with their romantic partners, based on data collected at the same time. In making this comparison, we were seeking to find a relationship that similar to parent– child relation- ships, was likely to have a high level of emotional connection, high levels of physical touch, and involve a high degree of sharing of space. For the sake of our analyses, we inferred these character- istics on the basis of the person’s role in our participants’ lives (romantic partners), because romantic partners vary in terms of these characteristics, just as parents and children vary in the degree to which they exhibit these characteristics (e.g., the degree to which they engage in physical touch). We also reasoned that including mothers’ romantic partners would provide a similar point of comparison in that mothers provide attachment-related caregiving to both their children and their partners (Kunce & Shaver, 1994). Further, similar to mothers’ roles with their chil- dren, romantic partners engage in some emotion regulation, or coregulation (Butler & Randall, 2013), with their partner that is not as typical in other relationships, such as friends and coworkers. However, this comparison was also based on the premise that the parent– child relationship is unique and was designed to test whether the role of “parent” creates emotional complexity that is specific to the parent– child relationship. As compared with the romantic partnership, wherein the mother can presumably also receive care, in the parent– child relationship, she can only provide care. Given that this comparison was not a primary focus of the investigation, we did not have directional hypotheses and con- ducted our analyses in an exploratory fashion.

Method

Participants and Design

The current study used baseline assessments from a larger investigation examining the efficacy of an in-home relationship wellness intervention with mothers of toddlers (approved by Pomona College IRB, #4/29/2016JB-MP). Mothers with at least one child between 18 and 27 months (M � 20.9, SD � 2.4) were

T hi

s do

cu m

en t

is co

py ri

gh te

d by

th e

A m

er ic

an P

sy ch

ol og

ic al

A ss

oc ia

ti on

or on

e of

it s

al li

ed pu

bl is

he rs

. T

hi s

ar ti

cl e

is in

te nd

ed so

le ly

fo r

th e

pe rs

on al

us e

of th

e in

di vi

du al

us er

an d

is no

t to

be di

ss em

in at

ed br

oa dl

y.

547MOTHERS’ REAL-TIME EMOTIONS WHILE CAREGIVING

invited to participate in the study. To recruit families, advertise- ments for “a study on emotion and relationships between parents and children” were placed on Internet classifieds, social network- ing sites, and around the community. The final sample of mothers (N � 136) were between 20 and 43 years of age (M � 30.8, SD � 5.0), and were ethnically and socioeconomically diverse (see Table 1). Because this was part of a larger study, the sample size was predetermined. However, our sample size is in line with other EMA studies examining emotion dynamics (e.g., Benson et al., 2018; Koval & Kuppens, 2012). The majority of mothers (53.7%) had additional children, which ranged from newborn to young adult but were typically of school-age (M � 5.5 years, SD � 4.1). Additional demographics are reported in Table 1.

Procedure

Participants provided informed consent and were given instruc- tions regarding the EMA phone surveys. The EMA surveys were administered randomly five times a day for 10 days using a platform called Survey Signal. This platform uses text messages to alert participants and provide a link to the online survey. In order to prevent retrospective recall, participants were asked to respond with regard to how they felt and what they were doing immediately before they entered their responses, not when they received the signal. Participants who did not have a smartphone (n � 6)

borrowed one from the laboratory during the survey period. Par- ticipants had a 30-min window to respond to each survey before it closed.

We obtained 5,027 complete responses from the original sample of 151 mothers (about a 67% response rate). In line with other EMA studies (e.g., Johnson et al., 2009), we identified a minimum compliance criterion prior to any data analyses. Par- ticipants who completed less than 10 surveys (�20% response rate) were excluded from data analyses (n � 6). We also excluded nine mothers who never reported being away from their children during the study period, as we were interested in comparing caregiving and noncaregiving contexts. The final response sample included 4,672 responses from 136 mothers, which corresponds to a 69% response rate. Out of 50 total possible responses, participants responded to an average of 34.4 signals (SD � 8.5).

Measures

Positive and negative emotion. For this study, we chose a list of positive and negative emotion words that tapped a wide range of emotional valence and intensity. The list of emotions included nine positive emotion words (attentive, excited, interested, proud, strong, content, grateful, accomplished, and joyful) and 11 nega- tive emotion words (empty, frustrated, guilty, hopeless, lonely, sad, stressed, worried, irritable, ashamed, angry). Participants were asked to rate the extent to which they were feeling each emotion at the present moment on a 5-point scale (1 � very slightly or not at all; 5 � extremely).

In line with generalizability theory, we conducted multilevel factor analyses to confirm that the data allowed for reliable as- sessment of within-person changes in emotion across time (Bolger & Laurenceau, 2013; Cranford et al., 2006). Within-person reli- ability, or Omega coefficients, were 0.82 for negative emotion and 0.85 for positive emotion.

Presence of children/partner. During each EMA assessment, participants were presented with a list of options and asked to indicate whom they were with at the time of the response. The list included: I’m alone, my child in study, my other child/children, my partner, other family member(s), friend(s), acquaintance(s), co- worker(s), stranger(s), other. For the purposes of this study, we considered mothers to be caregiving when they were in the pres- ence of any of their children, regardless of whether they were directly interacting with their children or simply near them but not interacting. This is consistent with other studies that define a parenting context as being in the presence of one’s child, regard- less of any direct interaction. (Meier, Musick, Flood, & Dunifon, 2016; Musick, Meier, & Flood, 2016) In fact, one EMA study found that the majority of parenting activities (76%) do not involve direct interaction with children (e.g., supervising play time, cook- ing them dinner; Offer, 2014). Measures of emotion dynamics at times when mothers’ partners were present or absent were also analyzed to determine whether differences were unique to the parent– child relationship.

Plan of Analysis

Measures of emotion. First, we calculated emotional inten- sity (iMean) by combining all positive and negative emotions at

Table 1 Sample Characteristics

Construct Percentage

Race White 68.1% Asian/Pacific Islander 5.2% American Indian 1.5% Black 2.2% More than one race 11.9% Other 11.1%

Ethnicity Hispanic 39.0% Not Hispanic 61.0%

Single parent Yes 8.9% No 91.1%

Number of children One 46.3% Two 47.1% Three 5.1% Four 1.5%

Work outside home Yes 52.9% No 47.1%

Education High school or less 8.7% Some college 25.2% Community college 13.4% Bachelor’s degree 31.5% Graduate degree 21.3%

Income Less than $40K 31.0% $41K to $60K 15.4% $61K to $80K 16.9% $81K to $100K 14.0% $100K to $120K 11.0% Over $120K 12.5%

T hi

s do

cu m

en t

is co

py ri

gh te

d by

th e

A m

er ic

an P

sy ch

ol og

ic al

A ss

oc ia

ti on

or on

e of

it s

al li

ed pu

bl is

he rs

. T

hi s

ar ti

cl e

is in

te nd

ed so

le ly

fo r

th e

pe rs

on al

us e

of th

e in

di vi

du al

us er

an d

is no

t to

be di

ss em

in at

ed br

oa dl

y.

548 KERR, RASMUSSEN, BUTTITTA, SMILEY, AND BORELLI

each timepoint and calculating two mean scores (one positive and one negative) for each individual observation. These scores were then compared across caregiving contexts using multilevel mod- eling, which is described in detail below.

Second, we used the intraindividual standard deviation (iSD) to represent variability, which was calculated separately for caregiv- ing and noncaregiving contexts.

The formula for iSD is:

iSDi � �� (yit � yi.)2Ti � 1 where yit is the observed score at occasion t for person i and yi is the individual’s mean across all occasions. There were multiple positive and negative emotion words, and we calculated an iSD for each individual emotion separately, and then created mean iSDs for positive and negative emotions (Wang, Hamaker, & Bergeman, 2012). It is important to note that six participants reported only one occasion when they were not with their children. Because the calculation for iSD requires more than one data point, this reduced the analytic sample size for iSD to 130 participants.

Third, we computed separate emodiversity scores for positive and negative emotions using Shannon’s entropy metric, which is useful for capturing both the richness (total number of individual emotions) and evenness (relative abundance of different emo- tions). Richness is useful when considering the range of emotion experiences, while evenness is useful for examining differences in intensities across individual emotions; for the purposes of this study we were interested in both richness and evenness.1 We calculated separate emodiversity scores for caregiving and non- caregiving occasions using the continuous scale scores. For these analyses, emotion scores were recoded into a 0 (not at all) to 4 (extremely) scale.

The formula for Shannon’s entropy is:

Entropyi � �� j�1

m

pij(ln pij)

where m reflects the number of discrete emotion types and pij indicates the proportion of the mother’s experiences that endorsed each discrete emotion. Shannon’s entropy produces scores from 0 to ln(m), with higher scores indicating greater diversity (Benson et al., 2018; Quoidbach et al., 2014). For instance, a person with a high entropy score may have often endorsed all 11 negative emotions at similar intensities, whereas a person with a low en- tropy score may have mainly indicated experiencing one or two (e.g., stressed), and indicating a score of 0 (not at all) on all others.

For our comparison analyses, the above procedures were repli- cated to obtain iMean, iSD, and entropy scores during occasions when mothers reported being with and without their romantic partners.

Hypothesis testing. To account for observations nested within individuals, we used SAS PROC MIXED to run multilevel models with maximum likelihood estimation for models compar- ing emotion intensity. In the two-level model, individual surveys (Level 1) were nested within participants (Level 2). We specified spatial power error structure (SP(POW)), a form of autocorrelation that accounts for the inherently unequal time intervals character- istic of the study design (Bolger & Laurenceau, 2013; Schwartz & Stone, 2007). We also explored linear effects of time, using both

time elapsed since start of study, time of day, and day of the week. None of the time variables had any impact on the results so were not included in the final analyses.

Consistent with procedures outlined by Bolger and Laurenceau (2013), for analyses involving iMean, or intensity, we partitioned between-person and within-person variance and controlled for variance at the between-person level. In other words, these anal- yses controlled for the proportion of responses where mothers reported caring for their children compared with the rest of the sample. This allowed us to isolate the within-person effects, in order to accurately compare mothers’ emotion intensity across two contexts (i.e., caregiving or noncaregiving).

The multilevel equation for emotion intensity is:

iMeanit � �00 � �01X.j � �10(Xij � X.j) � u0j � u1j � εij

where X.j represents the between-person component of the care- giving variable and Xij � X.j represents the within-person compo- nent. The average emotion intensity and the association between caregiving and emotion intensity were allowed to differ across persons, as indicated by the u0j and u1j parameters, respectively.

For models of variability and emodiversity, values during care- giving and noncaregiving contexts were compared using paired samples t tests and then repeated measures ANCOVA for models controlling for mean levels of emotion. Based on recommenda- tions by Wang, Hamaker, and Bergeman (2012), in analyses of variability we controlled for mean levels of positive and negative emotion. To do so, we aggregated mean emotion scores separately for caregiving and noncaregiving occasions across all surveys in the 10-day period. In line with recent recommendations by Mest- dagh et al. (2018), we also explored the relative iSD as an alter- native variability index that is not confounded by the mean. For emodiversity, we explored models with and without controlling for mean levels of emotion, consistent with previous studies of emo- diversity (Benson et al., 2018; Ong et al., 2018).

On average, mothers reported being with their children during 31% of their responses (SD � 19%), with 37% (n � 50) of mothers being with their children during 80% or more of the daily signals. As a result, the number of occasions were often unequal across groups with 52% of the sample having eight or less re- sponses without their children present. To account for these often unequal groups and provide confidence in our results, we calcu- lated average difference in emodiversity and variability across contexts with 10,000 bootstrap resamples of 10 observations in each context (with replacement) and report these results along with the t test and ANCOVAs.

Results

Means, standard deviations, and zero-order correlations for pri- mary study variables are reported in Table 2. Aggregated mean

1 Because the number of caregiving and noncaregiving occasions varied widely, we also conducted our analyses using the Gini coefficient (Gini, 1912), which only captures evenness, but operates independent of sample size and tends to be more normally distributed (Benson et al., 2018). Although the entropy scores were negatively distributed, the pattern of results was the same across the two indices (Gini and Shannon’s entropy) so Shannon’s entropy was retained and reported as the more informative metric.

T hi

s do

cu m

en t

is co

py ri

gh te

d by

th e

A m

er ic

an P

sy ch

ol og

ic al

A ss

oc ia

ti on

or on

e of

it s

al li

ed pu

bl is

he rs

. T

hi s

ar ti

cl e

is in

te nd

ed so

le ly

fo r

th e

pe rs

on al

us e

of th

e in

di vi

du al

us er

an d

is no

t to

be di

ss em

in at

ed br

oa dl

y.

549MOTHERS’ REAL-TIME EMOTIONS WHILE CAREGIVING

scores (across all observations) of intensity, variability, and emo- diversity for positive and negative emotion did not differ as a function of parent age, number of children, race, ethnicity, or income; therefore, these demographic variables were not included in analyses.

Hypothesis 1: Associations Between Emotion Intensity and Caregiving

Results from the fixed effects models revealed that, on av- erage, mothers reported greater intensity of positive emotion during caregiving compared to times when they were not caring for their children, �̂10 � 0.11, SE � 0.03, p � .001, after controlling for between-person differences in caregiving (see Table 3, Figure 1). Importantly, the random slope for caregiving on positive emotion was significant, suggesting that this effect varies across participants. Fixed effects models for negative emotion, however, revealed that mothers reported no differ- ences in intensity of negative emotion when they were caregiv- ing compared with when they were not, �̂10 � �0.02, SE � 0.01, p � .09, after controlling for between-person differences in caregiving (see Table 3, Figure 1).

Hypothesis 2: Associations Between Variability and Caregiving

Analyses revealed that after controlling for mean levels of positive emotion, there were no significant differences in vari- ability of positive emotion between caregiving and noncaregiv- ing occasions using the iSD metric (see Table 4). This was consistent with the relative variability index, which indicated no significant difference in variability between caregiving and noncaregiving contexts, t(129) � �0.92, p � .36. However, the bootstrapped 95% confidence interval for did not overlap zero for either iSD, 95% CI [0.001, 0.09], or relative variability, 95% CI [0.01, 0.04], suggesting that when comparing groups of even sizes there was greater variability in positive emotion during caregiving occasions.

For negative emotion, however, mothers reported higher vari- ability in negative emotion during caregiving compared to noncaregiving occasions, F(1, 128) � 5.12, p � .025, after controlling for mean levels of negative emotion. The boot- strapped 95% confidence interval, however, overlapped zero,

95% CI [�0.04, 0.05], suggesting that the difference may be partially explained by differences in the number of occasions (caregiving vs. noncaregiving) in each group. Results from a paired samples t test indicated that the relative variability index was lower during caregiving than noncaregiving, t(125) � �2.99, p � .003, although the bootstrapped 95% confidence

Table 2 Means, Standard Deviations, and Zero-Order Correlations for Study Variables

Construct N M (SD) 1 2 3 4 5 6 7

1. Parent age 136 30.8 (5.0) — 2. Number of children 136 1.62 (.7) .17� — 3. Pos emo intensity 136 3.0 (.8) .02 .02 — 4. Neg emo intensity 136 1.4 (.3) .00 .01 �.26� — 5. Pos emo variability 136 0.8 (0.2) �.13 �.15† �.22� 0.35� — 6. Neg emo variability 136 0.6 (0.3) �.11 �.03 �.22� 0.86� 0.55� — 7. Pos emodiversity 136 0.8 (0.1) �.01 �.14 0.65� �.05 �.11 �.08 — 8. Neg emodiversity 136 0.4 (0.1) �.03 0.00 �.24� 0.59� 0.43� 0.75� �.07

Note. emo � emotion; pos � positive; neg � negative. † p � .10. � p � .05.

Table 3 Multilevel Models Examining Intensity of Emotion Across Contexts

Predictor

Positive emotion

Negative emotion

Estimate SE Estimate SE

Caregiving models

Fixed effects Intercept 3.03� 0.07 1.29� 0.02 Caregiving (between) 0.23 0.34 0.002 0.10 Caregiving (within) 0.11� 0.03 �.02† 0.01

Random effects Level 2 (between-person)

Intercept 0.51� 0.05�

Intercept � Caregiving 0.01 �.01�

Caregiving (slope) 0.04� 0.01�

Level 1 (within-person) Variance 0.19� 0.05�

SP(POW) 0.97� 0.85�

Residual 0.19� 0.05�

Romantic partner models

Fixed effects Intercept 3.07� 0.06 1.28� 0.02 With partner (between) �.02 0.41 �.11 0.13 With partner (within) 0.06� 0.02 �.03� 0.01

Random effects Level 2 (between-person)

Intercept 0.47� 0.05�

Intercept � Partner 0.01 �.003 With partner (slope) 0.02� 0.004�

Level 1 (within-person) Variance 0.17� 0.06�

SP(POW) 0.97� 0.78�

Residual 0.19� 0.04�

Note. SP(POW) � spatial power error structure. SP(POW) indicates first-order autoregressive serial correlation. Variance indicates uncorre- lated errors across assessments. All random effects are represented as variances. † p � .10. � p � .05.

T hi

s do

cu m

en t

is co

py ri

gh te

d by

th e

A m

er ic

an P

sy ch

ol og

ic al

A ss

oc ia

ti on

or on

e of

it s

al li

ed pu

bl is

he rs

. T

hi s

ar ti

cl e

is in

te nd

ed so

le ly

fo r

th e

pe rs

on al

us e

of th

e in

di vi

du al

us er

an d

is no

t to

be di

ss em

in at

ed br

oa dl

y.

550 KERR, RASMUSSEN, BUTTITTA, SMILEY, AND BORELLI

interval of mean differences in relative variability based on 10,000 resamples of 10 occasions per group indicated the opposite trend. Together, these results suggest that the negative variability findings were not stable and varied depending on the metric (iSD or relative variability) and type of analyses used.

Hypothesis 3: Associations Between Emodiversity and Caregiving

Consistent with our hypothesis, mothers reported more emodi- versity in positive emotion during caregiving, t(135) � �3.95, p � .001. This effect remained significant in a repeated measures ANCOVA, controlling for mean levels of positive emotion (see

Table 4). Mothers also reported higher emodiversity, in negative emotion during times they were caring for their children compared with when they were not caregiving, t(135) � �5.35, p � .001, which remained significant in ANCOVA models controlling for mean levels of negative emotion (see Table 4). In other words, of the nine possible positive emotions and 11 possible negative emotions, mothers endorsed a wider variety and more even distri- bution/intensities of both emotion types on surveys when they were caring for their children (see Figure 2). The bootstrapped 95% confidence interval was [0.02, 0.04] for positive emotion and [0.08, 0.21] for negative emotion. These effects held across boot- strapping analyses, providing confidence that our results cannot be explained solely by differences in the amount of observations across groups.

Exploratory Analyses: Presence of Romantic Partner

To examine the uniqueness of our findings with respect to the parent– child relationship, we conducted all analyses again, this time comparing occasions when participants’ romantic partners were present and when they were not. In the overall sample of 136 mothers, nine participants did not have a romantic partner or were not with them during any of their signal responses, so were excluded from these analyses, resulting in a sample size of N � 127.

For emotion intensity, results indicated that mothers reported higher intensity in positive emotion, �̂10 � 0.06, SE � 0.02, p � .01, and lower intensity in negative emotion, �̂10 � �0.03, SE � 0.01, p � .01, when they were with their romantic partner compared with times when their partner was not present, after controlling for between-person differences in presence of ro- mantic partner (see Table 3). Results for emotion variability indicated that after controlling for mean levels of emotion, there were no differences in positive, F(1, 124) � .47, p � .50, or negative variability, F(1, 124) � 1.14, p � .29, when mothers

Figure 1. Emotion intensity scores for a single participant across the study period. The black line represents positive emotion and the red (gray) line represents negative emotion. The background is shaded according to whether it was a caregiving (darker shade) or noncaregiving (lighter shade) occasion. Missing data is represented by white space. See the online article for the color version of this figure.

Table 4 Repeated Measures ANCOVAs Comparing Variability and Diversity Across Contexts

Caregiving Noncaregiving Outcome M (SD) M (SD) F value �p2

Positive emotion Variability .79 (.22) .76 (.22) .01 .00 Emodiversity 2.13 (.10) 2.09 (.15) 16.77� .11

Negative emotion Variability .41 (.24) .38 (.32) 5.12� .04 Emodiversity 1.66 (.48) 1.40 (.66) 10.03� .07

With partner Not with partner

Positive emotion Variability .75 (.25) .79 (.23) .47 .00 Emodiversity 2.10 (.15) 2.12 (.10) 16.35� .12

Negative emotion Variability .35 (.26) .41 (.23) 1.14 .01 Emodiversity 1.42 (.62) 1.66 (.47) 4.29� .03

Note. All analyses controlling for mean level of emotion. � p � .05.

T hi

s do

cu m

en t

is co

py ri

gh te

d by

th e

A m

er ic

an P

sy ch

ol og

ic al

A ss

oc ia

ti on

or on

e of

it s

al li

ed pu

bl is

he rs

. T

hi s

ar ti

cl e

is in

te nd

ed so

le ly

fo r

th e

pe rs

on al

us e

of th

e in

di vi

du al

us er

an d

is no

t to

be di

ss em

in at

ed br

oa dl

y.

551MOTHERS’ REAL-TIME EMOTIONS WHILE CAREGIVING

reported being with their romantic partner compared with when they were not with their partners. Comparing mean differences across contexts using bootstrapped 95% confidence intervals, however, suggested that mothers report lower variability in positive emotion, 95% CI [�.10, �0.01] when with their ro- mantic partners compared to when they are not. This was consistent with the relative variability index, t(121) � 6.67, p � .001, which is not confounded by the mean. For emodiversity, results from paired samples t tests indicated that mothers re- ported less emodiversity in positive emotions when with their romantic partners, t(126) � 2.72, p � .008. This finding re- mained significant in a repeated measures ANCOVA, control- ling for mean levels of positive emotion (see Table 4). Mothers also reported less emodiversity in negative emotion when with their partners compared with when they were not with their partners, t(126) � �5.44, p � .001, which remained significant when controlling for mean levels of negative emotion (see Table 4). Both positive and negative emodiversity results held across bootstrapping, with 95% confidence intervals of [�0.03, �0.005] for positive emotion and [�0.26, �0.13] for negative emotion.

Discussion

This study provides insight into the complexity of mothers’ emotion experiences by comparing the intensity, variability, and emodiversity of mothers’ real-time emotions during caregiving and noncaregiving moments. Our results indicated that mothers reported more intense positive emotion and greater diversity in

discrete positive and negative emotions when caring for their children (see Figures 1 and 2). Further, there was some evidence that positive and negative emotion variability was higher during caregiving, although this was not consistent across metrics. With the exception of positive emotion intensity, these findings were specific to caregiving and were not observed when mothers were with their romantic partners. As the first study to examine multiple facets of real-time positive and negative emotional experience in mothers, our results support the notion that parenting is an emo- tionally complex experience and different in nature from the close partner relationship.

Our finding that higher intensity of positive emotion is as- sociated with caregiving occasions is consistent with other work examining parents’ real-time emotions (Delle Fave & Massi- mini, 2004; Impett et al., 2011; Nelson et al., 2013). Our more novel findings that caregiving is associated with more positive and negative emodiversity, and with more negative emotion variability, require further interpretation. The emodiversity findings seem to suggest that caregiving is associated with a richer emotional experience in terms of both positive and neg- ative emotions. Caregiving may create the type of experience that is more layered in terms of discrete emotional experience— perhaps through exposure to their children’s emotional experi- ence, which changes moment-to-moment and evolves rapidly across development, mothers are exposed to a kind of growth experience and development themselves, which if carefully observed, could result in a more varied or textured emotional landscape. Young children’s keen exploration of their physical

Figure 2. Emodiversity scores for a single participant during caregiving and noncaregiving occasions. The number of times each emotion was experienced is indicated by the length of each petal and the proportion of times the emotion was rated at low (darker shades) to high (lighter shades) intensities is indicated by the color of each petal. Both positive and negative emodiversity is higher for this mother when she is caring for her child compared with when she is not. See the online article for the color version of this figure.

T hi

s do

cu m

en t

is co

py ri

gh te

d by

th e

A m

er ic

an P

sy ch

ol og

ic al

A ss

oc ia

ti on

or on

e of

it s

al li

ed pu

bl is

he rs

. T

hi s

ar ti

cl e

is in

te nd

ed so

le ly

fo r

th e

pe rs

on al

us e

of th

e in

di vi

du al

us er

an d

is no

t to

be di

ss em

in at

ed br

oa dl

y.

552 KERR, RASMUSSEN, BUTTITTA, SMILEY, AND BORELLI

and emotional environments may prompt parents to pay closer attention to their own emotional experience, making more ap- parent the many different emotions they experience. Extant literature has established that having more diverse emotional experiences is positively associated with health and well-being (Benson et al., 2018; Quoidbach et al., 2014), indicating it is possible that parents experience some emotional benefit from caregiving. However, less work has examined emodiversity across contexts suggesting a need for further investigation.

Our findings around variability in emotion, however, are less straightforward. There was some evidence suggesting that vari- ability in positive and negative emotions was greater during caregiving compared to noncaregiving occasions. However, we used multiple metrics to examine variability to account for differences in the number of observations between contexts and to confirm results were not confounded with the mean levels of emotion. When looking across these different tests and metrics, the findings did not remain consistent, suggesting that any differences observed may have been spurious or unduly influ- enced by characteristics of the data from this specific sample. Given that few studies have compared variability across con- texts, these results are informative in illuminating the chal- lenges in capturing the real-time, contextual differences in variability. While our results did not remain stable across tests, we also do not feel we can rule out the possibility that differ- ences in variability does indeed exist across caregiving con- texts. Future research is needed in order to capture more occa- sions when parents are without children, as well as identifying ways of assessing emotion in an ongoing (streaming) way, as true assessments of variability would be moment-to-moment. Perhaps using a larger response scale would capture more nuanced variability, which would allow for a more detailed observation of differences in variability across real-time con- texts. Other ideas include constructing a “living laboratory,” perhaps using behavioral coding of parents’ emotion across a longer time span (e.g., a day) and comparing windows of time (e.g., 2 hr with, 2 hr away) when they were with their children and away from their children.

It is of particular interest that we did not find a similar pattern of emotional experience when mothers were with their romantic partners. Mothers did report more intense positive emotions when with their romantic partners (as compared with when they were away from romantic partners) as well as when with their children (as compared with when they were away from their children). This is consistent with other literature suggesting that close relationships are beneficial for one’s well-being (e.g., Dush & Amato, 2005). However, mothers actually demon- strated the opposite trend in emodiversity—they reported less positive and negative emodiversity when with their partners, and there was some evidence supporting that mothers reported lower variability in positive emotion when with their romantic partners as well. In contrast to being with one’s child, the presence of a romantic partner may actually diminish emotional complexity because partners are likely to share emotions, pos- itive or negative, which can have some benefits as well as some drawbacks (for a review see Butler, 2015).

The parent– child relationship may be different from other interpersonal relationships in that caregiving places unique demands on parents’ emotion regulation (e.g., Rasmussen et al.,

2017; Rutherford et al., 2015)—to some extent, effective care- giving may actually require holding one’s own emotional needs at bay (at least in the moment) to attend to the child’s needs. Although there is evidence for emotion coregulation in adult relationships (e.g., Butner, Diamond, & Hicks, 2007; Saxbe & Repetti, 2010; Sbarra & Hazan, 2008), these partnerships are more egalitarian in nature, in that each partner may serve a regulatory function for the other. Children require guidance from their parents in helping them identify and cope with emotions they experience (Eisenberg et al., 1998; Rutherford et al., 2015). Given this relationship difference, experiencing greater intensity, range, and types of emotions may reflect caregiving sensitivity with children. During moments of care- giving, there may be more emotion stimuli to attend to (both their own and their children’s); therefore, experiencing a wider range of intensity, and types of emotions may indicate that typically, parents are emotionally responsive to their environ- ments and, importantly, adapting to their children’s needs (e.g., Lindquist & Barrett, 2008; Quoidbach et al., 2014). While future research is needed to understand the implications of emotion dynamics across caregiving contexts for caregivers’ mental health and for children’s feelings of security, our results provide support for the idea that caregiving is emotionally complex, yet generally positive.

Limitations

Several limitations of this study deserve notice. First, our study relied on mothers’ subjective reports of their own emo- tions, which are subject to social desirability bias. Second, the random effects in the multilevel models examining emotion intensity were significant, indicating that while there was an overall association between context and emotion, this model showed significant variability. Specifically, the random slope of caregiving varied from �0.9 to 0.31, indicating that for some mothers the effect of context was positive and for some it was negative. There are likely to be several individual differences that influence the quality of emotion experience while caregiv- ing. For example, a recent study showed that attachment avoid- ance was related to differences in positive emotion when care- giving (Kerr, Buttitta, Smiley, Rasmussen, & Borelli, 2019), with more avoidant mothers reporting less positive emotion during caregiving. Identifying specific factors that influence differences in emotion dynamics across contexts, such as care- giving, is an important area of future inquiry.

Next, the generalizability of our findings is limited given that we only included mothers, and research indicates that mothers and fathers report different experiences of parenthood (Nelson et al., 2014; Nelson-Coffey et al., 2019). Further, comparing caregiving and noncaregiving contexts provides a broad over- view of parenting; we were unable to complete more fine- grained analyses that would allow us to consider the nature of the interactions between mother and child. For example, be- cause mothers were likely to be with more than one child at a time, we were unable to isolate interactions with a specific child or determine how child age impacted mothers’ emotions. Sim- ilarly, our study did not take into account children’s emotions or temperament. Because of the bidirectional nature of the parent– child relationship (Belsky, 1984), obtaining a complete picture

T hi

s do

cu m

en t

is co

py ri

gh te

d by

th e

A m

er ic

an P

sy ch

ol og

ic al

A ss

oc ia

ti on

or on

e of

it s

al li

ed pu

bl is

he rs

. T

hi s

ar ti

cl e

is in

te nd

ed so

le ly

fo r

th e

pe rs

on al

us e

of th

e in

di vi

du al

us er

an d

is no

t to

be di

ss em

in at

ed br

oa dl

y.

553MOTHERS’ REAL-TIME EMOTIONS WHILE CAREGIVING

of parents’ emotions during caregiving would ideally include a consideration of children’s real-time emotions and behavior as well. Another limitation is that our comparison condition was mothers’ emotions when with or without their romantic part- ners, who we presumed were their attachment figures. We selected romantic partners in order to control for close inter- personal connections; however, in future studies of this topic, it may be useful to include a comparison condition that involves caregiving because the partner relationship also involves care- seeking—for instance, in studies in which parents also have older children, researchers could compare times when parents are with and without their older children to times when parents and with and without a young child. Doing so would enable researchers to isolate the unique emotional experience of par- enting a young child.

Finally, many of the mothers in our sample were rarely sepa- rated from their children during signal responses. Over the course of the 10 days, 37% (n � 50) of our mothers reported being with their children during 80% or more of the daily signals. While multilevel models can account for this unbalanced data, and our analyses comparing caregiving and noncaregiving contexts should be interpreted with caution. Although, the emodiversity results held across 10,000 bootstrapped resamples, this may have contrib- uted to the inconsistent findings we observed for emotion variabil- ity. Conceptually, it is difficult to obtain a clear picture of mothers’ emotions when not caring for their children if they were rarely separated from them.

Future Directions

This study goes beyond prior work in capturing the nuances of parents’ emotional experiences, yet we recognize that the results presented here represent only a snapshot of the parenting experience. In fact, assessing parents’ emotional experiences at this level reveals the extent of unexplored questions in this area. For example, future research could extend our work on care- giving/noncaregiving contexts with parents, to compare emo- tion dynamics in parents and nonparents, or across the transition to parenthood. Further, longitudinal data that capture real-time experiences at multiple points in development will allow us to map the trajectory of parents’ emotional experiences as their children become more autonomous. Future work should also consider how individual differences, such as personality or attachment style, as well as the nature of day-to-day activities that parents are engaged in, impact parents’ emotions.

Conclusion

This study significantly contributes to existing literature on parents’ emotional experience by exploring multiple facets of real-time emotions. By showing that mothers report more in- tense positive, and more diverse positive and negative emo- tional experiences when caring for their children than when not caring for them, we have demonstrated that the emotional experience of parents is complex and nuanced. Our work has important implications for developing interventions that could enhance parent– child relationships, through helping parents become aware of and cope with emotion change during care- giving experiences.

References

Adolf, J. K., Voelkle, M. C., Brose, A., & Schmiedek, F. (2017). Capturing context-related change in emotional dynamics via fixed moderated time series analysis. Multivariate Behavioral Research, 52, 499 –531. http:// dx.doi.org/10.1080/00273171.2017.1321978

Barrett, L. F., Gross, J., Christensen, T., & Benvenuto, M. (2001). Know- ing what you’re feeling and knowing what to do about it: Mapping the relation between emotion differentiation and emotion regulation. Cog- nition and Emotion, 15, 713–724. http://dx.doi.org/10.1080/ 02699930143000239

Belsky, J. (1984). The determinants of parenting: A process model. Child Development, 55, 83–96. http://dx.doi.org/10.2307/1129836

Benson, L., Ram, N., Almeida, D. M., Zautra, A. J., & Ong, A. D. (2018). Fusing biodiversity metrics into investigations of daily life: Illustrations and recommendations with emodiversity. The Journals of Gerontology: Series B, 73, 75– 86.

Bisconti, T. L., Bergeman, C. S., & Boker, S. M. (2004). Emotional well-being in recently bereaved widows: A dynamical systems approach. The Journals of Gerontology Series B, Psychological Sciences and Social Sciences, 59, 158 –167. http://dx.doi.org/10.1093/geronb/59.4 .P158

Bolger, N., & Laurenceau, J. P. (2013). Intensive longitudinal methods. New York, NY: Guilford Press.

Butler, E. A. (2015). Interpersonal affect dynamics: It takes two (and time) to tango. Emotion Review, 7, 336 –341. http://dx.doi.org/10.1177/ 1754073915590622

Butler, E. A., & Randall, A. K. (2013). Emotional coregulation in close relationships. Emotion Review, 5, 202–210. http://dx.doi.org/10.1177/ 1754073912451630

Butner, J., Diamond, L. M., & Hicks, A. M. (2007). Attachment style and two forms of affect coregulation between romantic partners. Personal Relationships, 14, 431– 455. http://dx.doi.org/10.1111/j.1475-6811.2007 .00164.x

Cranford, J. A., Shrout, P. E., Iida, M., Rafaeli, E., Yip, T., & Bolger, N. (2006). A procedure for evaluating sensitivity to within-person change: Can mood measures in diary studies detect change reliably? Personality and Social Psychology Bulletin, 32, 917–929. http://dx.doi.org/10.1177/ 0146167206287721

Deater-Deckard, K. (2004). Parenting stress. New Haven, CT: Yale Uni- versity Press. http://dx.doi.org/10.12987/yale/9780300103939.001.0001

Deaton, A., & Stone, A. A. (2014). Evaluative and hedonic wellbeing among those with and without children at home. Proceedings of the National Academy of Sciences of the United States of America, 111, 1328 –1333. http://dx.doi.org/10.1073/pnas.1311600111

Delle Fave, A., & Massimini, F. (2004). Parenthood and the quality of experience in daily life: A longitudinal study. Social Indicators Re- search, 67, 75–106. http://dx.doi.org/10.1023/B:SOCI.0000007335 .26602.59

Dush, C. M. K., & Amato, P. R. (2005). Consequences of relationship status and quality for subjective well-being. Journal of Social and Personal Relationships, 22, 607– 627. http://dx.doi.org/10.1177/ 0265407505056438

Eid, M., & Diener, E. (1999). Intraindividual variability in affect: Reli- ability, validity, and personality correlates. Journal of Personality and Social Psychology, 76, 662– 676. http://dx.doi.org/10.1037/0022-3514 .76.4.662

Eisenberg, N., Cumberland, A., & Spinrad, T. L. (1998). Parental social- ization of emotion. Psychological Inquiry, 9, 241–273. http://dx.doi.org/ 10.1207/s15327965pli0904_1

Erickson, M. F., & Aird, E. G. (2005). The motherhood study: Fresh insights on mothers’ attitudes and concerns. Research Report, Mother- hood Project and the Mothers’ Council of the Institute for American Values. Retrieved from http://www.motherhoodproject.org

T hi

s do

cu m

en t

is co

py ri

gh te

d by

th e

A m

er ic

an P

sy ch

ol og

ic al

A ss

oc ia

ti on

or on

e of

it s

al li

ed pu

bl is

he rs

. T

hi s

ar ti

cl e

is in

te nd

ed so

le ly

fo r

th e

pe rs

on al

us e

of th

e in

di vi

du al

us er

an d

is no

t to

be di

ss em

in at

ed br

oa dl

y.

554 KERR, RASMUSSEN, BUTTITTA, SMILEY, AND BORELLI

Frijda, N. H. (2007). The laws of emotion. Mahwah, NJ: Laurence Erl- baum.

Gini, C. (1912). Variabilità e mutabilità [Variability and Mutability]. Reprinted. In E. Pizetti & T. Salvemini (Eds.), Memorie di metodologica statistica. Rome, Italy: Libreria Eredi Virgilio Veschi.

Gruber, J., Kogan, A., Quoidbach, J., & Mauss, I. B. (2013). Happiness is best kept stable: Positive emotion variability is associated with poorer psychological health. Emotion, 13, 1– 6. http://dx.doi.org/10.1037/ a0030262

Hardy, J., & Segerstrom, S. C. (2017). Intra-individual variability and psychological flexibility: Affect and health in a National U.S. sample. Journal of Research in Personality, 69, 13–21. http://dx.doi.org/10 .1016/j.jrp.2016.04.002

Houben, M., Van Den Noortgate, W., & Kuppens, P. (2015). The relation between short-term emotion dynamics and psychological well-being: A meta-analysis. Psychological Bulletin, 141, 901–930. http://dx.doi.org/ 10.1037/a0038822

Impett, E. A., English, T., & John, O. P. (2011). Women’s emotions during interactions with their grown children in later adulthood: The moderat- ing role of attachment avoidance. Social Psychological & Personality Science, 2, 42–50. http://dx.doi.org/10.1177/1948550610379505

Izard, C. E. (2009). Emotion theory and research: Highlights, unanswered questions, and emerging issues. Annual Review of Psychology, 60, 1–25. http://dx.doi.org/10.1146/annurev.psych.60.110707.163539

Johnson, E. I., Grondin, O., Barrault, M., Faytout, M., Helbig, S., Husky, M., . . . Swendsen, J. (2009). Computerized ambulatory monitoring in psychiatry: A multi-site collaborative study of acceptability, compliance, and reactivity. International Journal of Methods in Psychiatric Re- search, 18, 48 –57. http://dx.doi.org/10.1002/mpr.276

Kahneman, D., Krueger, A. B., Schkade, D. A., Schwarz, N., & Stone, A. A. (2004). A survey method for characterizing daily life experience: The day reconstruction method. Science, 306, 1776 –1780. http://dx.doi .org/10.1126/science.1103572

Kerr, M. L., Buttitta, K. V., Smiley, P. A., Rasmussen, H. F., & Borelli, J. L. (2019). Mothers’ real-time emotion as a function of attachment and proximity to their children. Journal of Family Psychology, 33, 575–585. http://dx.doi.org/10.1037/fam0000515

Koval, P., & Kuppens, P. (2012). Changing emotion dynamics: Individual differences in the effect of anticipatory social stress on emotional inertia. Emotion, 12, 256 –267. http://dx.doi.org/10.1037/a0024756

Kunce, L. J., & Shaver, P. R. (1994). An attachment-theoretical approach to caregiving in romantic relationships. In K. Bartholomew & D. Perl- man (Eds.), Advances in personal relationships: Vol. 5. Attachment processes in adulthood (pp. 205–237). London, UK: Kingsley.

Kuppens, P., Oravecz, Z., & Tuerlinckx, F. (2010). Feelings change: Accounting for individual differences in the temporal dynamics of affect. Journal of Personality and Social Psychology, 99, 1042–1060. http://dx.doi.org/10.1037/a0020962

Le, B. M., & Impett, E. A. (2016). The costs of suppressing negative emotions and amplifying positive emotions during parental caregiving. Personality and Social Psychology Bulletin, 42, 323–336. http://dx.doi .org/10.1177/0146167216629122

Lindquist, K. A., & Barrett, L. F. (2008). Emotional complexity. In M. Lewis, J. M. Haviland-Jones, & L. F. Barrett (Eds.), Handbook of emotions (pp. 513–530). New York, NY: Guilford Press.

McLanahan, S., & Adams, J. (1987). Parenthood and psychological well- being. Annual Review of Sociology, 13, 237–257. http://dx.doi.org/10 .1146/annurev.so.13.080187.001321

Meier, A., Musick, K., Flood, S., & Dunifon, R. (2016). Mothering experiences: How single parenthood and employment structure the emo- tional valence of parenting. Demography, 53, 649 – 674. http://dx.doi .org/10.1007/s13524-016-0474-x

Mestdagh, M., Pe, M., Pestman, W., Verdonck, S., Kuppens, P., & Tuer- linckx, F. (2018). Sidelining the mean: The relative variability index as

a generic mean-corrected variability measure for bounded variables. Psychological Methods, 23, 690 –707. http://dx.doi.org/10.1037/ met0000153

Musick, K., Meier, A., & Flood, S. (2016). How parents fare: Mothers’ and fathers’ subjective well-being in time with children. American Sociolog- ical Review, 81, 1069 –1095. http://dx.doi.org/10.1177/00031224 16663917

Nelson, S. K., Kushlev, K., English, T., Dunn, E. W., & Lyubomirsky, S. (2013). In defense of parenthood: Children are associated with more joy than misery. Psychological Science, 24, 3–10. http://dx.doi.org/10.1177/ 0956797612447798

Nelson, S. K., Kushlev, K., & Lyubomirsky, S. (2014). The pains and pleasures of parenting: When, why, and how is parenthood associated with more or less well-being? Psychological Bulletin, 140, 846 – 895. http://dx.doi.org/10.1037/a0035444

Nelson-Coffey, S. K., Killingsworth, M., Layous, K., Cole, S. W., & Lyubomirsky, S. (2019). Parenthood is associated with greater well- being for fathers than mothers. Personality and Social Psychology Bulletin, 45, 1378 –1390. http://dx.doi.org/10.1177/0146167219829174

Nesselroade, J. R. (1991). Interindividual differences in intraindividual change. In L. M. Collins & J. L. Horn (Eds.), Best methods for the analysis of change: Recent advances, unanswered questions, future directions (pp. 92–105). Washington, DC: American Psychological As- sociation. http://dx.doi.org/10.1037/10099-006

Offer, S. (2014). Time with children and employed parents’ emotional well-being. Social Science Research, 47, 192–203. http://dx.doi.org/10 .1016/j.ssresearch.2014.05.003

Ong, A. D., Benson, L., Zautra, A. J., & Ram, N. (2018). Emodiversity and biomarkers of inflammation. Emotion, 18, 3–14. http://dx.doi.org/10 .1037/emo0000343

Quoidbach, J., Gruber, J., Mikolajczak, M., Kogan, A., Kotsou, I., & Norton, M. I. (2014). Emodiversity and the emotional ecosystem. Jour- nal of Experimental Psychology, 143, 2057–2066. http://dx.doi.org/10 .1037/a0038025

Ram, N., & Gerstorf, D. (2009). Time-structured and net intraindividual variability: Tools for examining the development of dynamic character- istics and processes. Psychology and Aging, 24, 778 –791. http://dx.doi .org/10.1037/a0017915

Rasmussen, H. F., Borelli, J. L., Smiley, P. A., Cohen, C., Cheung, R. C. M., Fox, S., . . . Blackard, B. (2017). Mother-child language style matching predicts children’s and mothers’ emotion reactivity. Behav- ioural Brain Research, 325, 203–213. http://dx.doi.org/10.1016/j.bbr .2016.12.036

Ross, C. E., & Van Willigen, M. (1996). Gender, parenthood, and anger. Journal of Marriage and the Family, 58, 572–584. http://dx.doi.org/10 .2307/353718

Rutherford, H. J., Wallace, N. S., Laurent, H. K., & Mayes, L. C. (2015). Emotion regulation in parenthood. Developmental Review, 36, 1–14. http://dx.doi.org/10.1016/j.dr.2014.12.008

Saxbe, D., & Repetti, R. L. (2010). For better or worse? Coregulation of couples’ cortisol levels and mood states. Journal of Personality and Social Psychology, 98, 92–103. http://dx.doi.org/10.1037/a0016959

Sbarra, D. A., & Emery, R. E. (2005). The emotional sequelae of non- marital relationship dissolution: Analysis of change and intraindividual variability over time. Personal Relationships, 12, 213–232. http://dx.doi .org/10.1111/j.1350-4126.2005.00112.x

Sbarra, D. A., & Hazan, C. (2008). Coregulation, dysregulation, self- regulation: An integrative analysis and empirical agenda for understand- ing adult attachment, separation, loss, and recovery. Personality and Social Psychology Review, 12, 141–167. http://dx.doi.org/10.1177/ 1088868308315702

Schwartz, J. E., & Stone, A. A. (2007). The analysis of real-time momen- tary data: A practical guide. In A. A. Stone, S. Shiffman, A. A. Atienza, & L. Nebeling (Eds.), The science of real-time data capture: Self-reports

T hi

s do

cu m

en t

is co

py ri

gh te

d by

th e

A m

er ic

an P

sy ch

ol og

ic al

A ss

oc ia

ti on

or on

e of

it s

al li

ed pu

bl is

he rs

. T

hi s

ar ti

cl e

is in

te nd

ed so

le ly

fo r

th e

pe rs

on al

us e

of th

e in

di vi

du al

us er

an d

is no

t to

be di

ss em

in at

ed br

oa dl

y.

555MOTHERS’ REAL-TIME EMOTIONS WHILE CAREGIVING

in health research (pp. 76 –113). New York, NY: Oxford University Press.

Schwarz, N. (1990). Feelings as information: Informational and motiva- tional functions of affective states. In E. T. Higgins & R. M. Sorrentino (Eds.), Handbook of motivation and cognition: Foundations of social behavior (Vol. 2, pp. 527–561). New York, NY: Guilford Press.

Scott, S. B., Ram, N., Smyth, J. M., Almeida, D. M., & Sliwinski, M. J. (2017). Age differences in negative emotional responses to daily stres- sors depend on time since event. Developmental Psychology, 53, 177– 190. http://dx.doi.org/10.1037/dev0000257

Taquet, M., Quoidbach, J., de Montjoye, Y.-A., Desseilles, M., & Gross, J. J. (2016). Hedonism and the choice of everyday activities. Proceed- ings of the National Academy of Sciences of the United States of America, 113, 9769 –9773. http://dx.doi.org/10.1073/pnas.1519998113

Trull, T. J., Lane, S. P., Koval, P., & Ebner-Priemer, U. W. (2015). Affective dynamics in psychopathology. Emotion Review, 7, 355–361. http://dx.doi.org/10.1177/1754073915590617

Trull, T. J., Solhan, M. B., Tragesser, S. L., Jahng, S., Wood, P. K., Piasecki, T. M., & Watson, D. (2008). Affective instability: Measuring

a core feature of borderline personality disorder with ecological momen- tary assessment. Journal of Abnormal Psychology, 117, 647– 661. http:// dx.doi.org/10.1037/a0012532

Wang, L. P., Hamaker, E., & Bergeman, C. S. (2012). Investigating inter-individual differences in short-term intra-individual variability. Psychological Methods, 17, 567–581. http://dx.doi.org/10.1037/ a0029317

White, M. P., & Dolan, P. (2009). Accounting for the richness of daily activities. Psychological Science, 20, 1000 –1008. http://dx.doi.org/10 .1111/j.1467-9280.2009.02392.x

Zautra, A. J., Berkhof, J., & Nicolson, N. A. (2002). Changes in affect interrelations as a function of stressful events. Cognition and Emotion, 16, 309 –318. http://dx.doi.org/10.1080/02699930143000257

Received August 17, 2018 Revision received October 18, 2019

Accepted November 25, 2019 �

T hi

s do

cu m

en t

is co

py ri

gh te

d by

th e

A m

er ic

an P

sy ch

ol og

ic al

A ss

oc ia

ti on

or on

e of

it s

al li

ed pu

bl is

he rs

. T

hi s

ar ti

cl e

is in

te nd

ed so

le ly

fo r

th e

pe rs

on al

us e

of th

e in

di vi

du al

us er

an d

is no

t to

be di

ss em

in at

ed br

oa dl

y.

556 KERR, RASMUSSEN, BUTTITTA, SMILEY, AND BORELLI