You are on page 1of 13

Emotion

© 2022 American Psychological Association 2023, Vol. 23, No. 4, 1048–1060


ISSN: 1528-3542 https://doi.org/10.1037/emo0001132

Interpersonal Emotion Regulation Flexibility: Effects on Affect in Daily


Life

Ashley M. Battaglini, Katerina Rnic, Taylyn Jameson, Ellen Jopling, and Joelle LeMoult
Content may be shared at no cost, but any requests to reuse this content in part or whole must go through the American Psychological Association.

Department of Psychology, The University of British Columbia

Although the emotion regulation (ER) literature is vast, two emerging areas are particularly noteworthy.
First, as opposed to the traditional blanket characterization of ER strategies as adaptive or maladaptive,
theoretical models have highlighted the adaptability of greater ER flexibility (i.e., flexibly implementing
This document is copyrighted by the American Psychological Association or one of its allied publishers.

ER strategies based on the context). Second, instead of focusing on how individuals independently regu-
late emotions, researchers are increasingly examining how ER can occur with the help of another per-
son, a process known as interpersonal emotion regulation (IER). This study is the first to integrate these
two emerging areas of research and to apply the two main theories of ER flexibility to investigate the
effect of IER flexibility on negative and positive affect. A sample of 384 adults (Mage = 38.58 years,
SD = 13.82) residing predominantly in North America completed this 14-day daily diary study. As
expected, greater repertoire and greater responsivity to feedback were associated with more adaptive
affective outcomes (i.e., less negative affect and/or more positive affect). However, unexpected findings
also emerged: Greater context sensitivity did not significantly predict affect, and the covariation of
within-strategy variability and environmental variability predicted higher negative affect. Findings pro-
vide initial evidence that IER adaptiveness is influenced by one’s ability to flexibly implement IER strat-
egies (i.e., IER flexibility). The results also highlight the components of IER flexibility (namely greater
repertoire and responsivity to feedback) that predict daily adaptive affective consequences.

Keywords: emotion, affect, emotion regulation, emotion regulation flexibility, interpersonal emotion
regulation

Supplemental materials: https://doi.org/10.1037/emo0001132.supp

Individuals must regularly manage their emotions to effectively contexts. Instead, emotion regulation adaptiveness may be deter-
respond to their environment. The process of implementing strat- mined by one’s ability to flexibly implement emotion regulation
egies to modulate emotions is known as emotion regulation (Gross, strategies based on context, a process known as emotion regulation
1998). Importantly, there are two key emerging areas within this flexibility (Aldao et al., 2015; Bonanno & Burton, 2013).
body of literature. The first is the increased emphasis on emotion The second emerging area within the emotion regulation literature
regulation flexibility. Historically, researchers have categorized is the focus on adults’ interpersonal emotion regulation (Zaki & Wil-
emotion regulation strategies as adaptive or maladaptive based on liams, 2013). Previously, researchers had predominantly conceptual-
their consequences. However, researchers have criticized the blanket ized emotion regulation as an intrapersonal process, whereby
characterization of strategies as adaptive or maladaptive across individuals regulate their own emotions. However, individuals do
not always regulate emotions independently (Zaki & Williams,
2013). Indeed, emotion regulation can occur with the help of another
This article was published Online First September 1, 2022. person, a process known as interpersonal emotion regulation (IER;
Ashley M. Battaglini https://orcid.org/0000-0002-0498-6204 Zaki & Williams, 2013; also see Eisenberg et al., 1998, for a review
Katerina Rnic https://orcid.org/0000-0002-7132-4489 of the rich and longstanding interpersonal emotion regulation litera-
Taylyn Jameson https://orcid.org/0000-0002-8088-623X ture in children). IER can be divided into extrinsic and intrinsic IER.
Ellen Jopling https://orcid.org/0000-0001-5518-2364 In extrinsic IER, an individual attempts to regulate another person’s
Joelle LeMoult https://orcid.org/0000-0003-0931-2855 emotions, whereas in intrinsic IER, an individual engages in social
There has been no dissemination (e.g., conference presentation) of this contact to regulate their own emotions (Zaki & Williams, 2013).
data or its interpretation prior to this article’s submission. The data, code, Given our interest in understanding the association between imple-
and preregistration of the study’s hypotheses, design, and analytic plan can
menting IER strategies and one’s own affect, we investigated intrin-
be found on the Open Science Framework (Battaglini, 2022): https://osf.io/
yfz9r/?view_only=6db19703141f4c398dd125e65d94650b.
sic IER exclusively (e.g., Zelic et al., 2017); thus, going forward, we
Correspondence concerning this article should be addressed to Ashley M. use the shorthand “IER” to describe intrinsic IER.
Battaglini, Department of Psychology, The University of British Columbia, Despite the importance of both the emotion regulation flexibility
2136 West Mall, Vancouver, BC V6T 1Z4, Canada. Email: ashley.battaglini@ and IER literatures, to date, they have largely remained independ-
psych.ubc.ca ent. The present study amalgamates these literatures in order to

1048
1
INTERPERSONAL EMOTION REGULATION FLEXIBILITY 1049

introduce the construct of IER flexibility and to test its association that one engages in any given IER strategy over time; greater fluc-
with affect in daily life. Specifically, we defined IER flexibility tuations in the use of that strategy over time is quantified as higher
based on the two leading emotion regulation flexibilities theories within-strategy variability. Between-strategy variability is the
(Aldao et al., 2015; Bonanno & Burton, 2013) and then used this degree to which one engages in different IER strategies on a given
definition to investigate whether the flexible implementation of day (Blanke et al., 2020); greater variability in the degree different
four IER strategies (i.e., coreappraisal, codistraction, corumina- strategies are used in a given day is quantified as higher between-
tion, and coproblem-solving) was associated with short-term strategy variability. Both within- and between-strategy variability
changes in negative and positive affect. Toward this goal, we first of intrapersonal emotion regulation strategies predict lower nega-
describe the components of emotion regulation flexibility.
Content may be shared at no cost, but any requests to reuse this content in part or whole must go through the American Psychological Association.

tive affect (Blanke et al., 2020). However, researchers have not yet
examined whether repertoire of IER strategies is associated with
Emotion Regulation Flexibility daily adaptive affective outcomes.
The third component of Bonanno and Burton (2013) theory is
Presently, there are two central theories of emotion regulation feedback: when an individual incorporates internal feedback (i.e.,
flexibility. In the initial theory, Bonanno and Burton (2013) pos- their affective state) and then changes their emotion regulation
ited three interrelated components of emotion regulation flexibil- strategy use accordingly (Bonanno & Burton, 2013). Experimental
This document is copyrighted by the American Psychological Association or one of its allied publishers.

ity: context sensitivity, repertoire, and feedback. and cross-sectional studies on intrapersonal emotion regulation
Context sensitivity is the appropriate alignment of emotion regu- strategies show that higher feedback predicts greater life satisfac-
lation strategy with context (Bonanno & Burton, 2013). Empirical tion, whereas lower feedback is associated with elevated depres-
evidence supports the importance of context sensitivity for intraper- sive and anxiety symptoms (Birk & Bonanno, 2016; Chen &
sonal emotion regulation strategies. For example, whereas cognitive Bonanno, 2021). Similar to the other facets of Bonanno and Bur-
reappraisal had historically been characterized as a uniformly adapt- ton’s (2013) theory, researchers have not yet examined the associ-
ive strategy, more recent evidence suggests that its adaptiveness ation between feedback and affect in the context of IER strategies.
depends on context: It is adaptive in uncontrollable situations but Aldao et al. (2015) advanced the second main emotion regula-
maladaptive in controllable situations (Haines et al., 2016; Troy tion flexibility theory, which builds on the contributions of
et al., 2013). Indeed, emotion-focused coping strategies (i.e., strat- Bonanno and Burton (2013) by positing that emotion regulation
egies used to change emotions; Lazarus & Folkman, 1984; Troy flexibility is determined based on the joint fluctuation of emotion
et al., 2013), such as cognitive reappraisal, may be beneficial in regulation variability and environmental changes. One study to
uncontrollable contexts because changing emotions may hold more date has investigated the consequences of intrapersonal emotion
utility in these situations than problem-focused coping strategies regulation variability and affect. Blanke et al. (2020) found that
(i.e., strategies used to change the situation; Lazarus, 1993), such as intrapersonal within-strategy variability did not consistently pre-
problem-solving. Although researchers have not yet examined dict negative affect, whereas intrapersonal between-strategy vari-
whether controllability influences the adaptiveness of IER strategies, ability predicted lower negative affect. Strikingly, a study has yet
there is reason to investigate it as a possibility. For one, IER strat- to investigate the covariation of IER variability and environmental
egies have affective outcomes that tend to parallel those of their changes in predicting affect.
intrapersonal emotion regulation counterparts (e.g., Christensen Overall, emotion regulation flexibility is a promising direction
et al., 2020; Zelic et al., 2017); thus, it is possible that their adaptive- for future research, evidenced by studies that have supported the
ness is moderated by the same constructs. In addition, investigating importance of defining emotion regulation adaptiveness as the
the influence of controllability on IER would allow us to draw con- ability to flexibly implement emotion regulation strategies. How-
nections between inter- and intrapersonal emotion regulation given ever, a shift is needed in the application of emotion regulation flex-
the prevalence of controllability in the intrapersonal emotion regula- ibility theory to IER strategies.
tion literature (Haines et al., 2016; Troy et al., 2013). Moreover,
researchers have documented that receiving support from another Interpersonal Emotion Regulation
person is associated with adaptive outcomes in uncontrollable con-
texts (e.g., Valentiner et al., 1994). IER strategies were derived from the broader intrapersonal emo-
Repertoire, the second determinant of emotion regulation flexi- tion regulation literature and translated to a dyadic form. The most
bility as proposed by Bonanno and Burton (2013), describes the commonly studied IER strategies include corumination, codistrac-
range of emotion regulation strategies that one has at their dis- tion, coproblem-solving, and coreappraisal. Corumination is
posal. There are two types of repertoire considered: categorical defined as repeatedly discussing one’s problems and feelings with
and temporal variability. Categorical variability is operationalized another person (Rose, 2002). Codistraction is characterized as
as the number of strategies implemented. Empirical evidence indi- engaging in a conversation with another person about neutral in-
cates that greater categorical variability of intrapersonal emotion formation in order to direct one’s focus away from problems and
regulation strategies is associated with enhanced well-being and feelings (Zelic et al., 2017). Coproblem-solving occurs in an inter-
lower internalizing symptoms (Eldesouky & English, 2018; action where another person helps to solve one’s problem (Waller
Lougheed & Hollenstein, 2012; Southward & Cheavens, 2020). et al., 2014). Finally, coreappraisal occurs when another person
Temporal variability is assessed based on both within-strategy var- helps to change one’s interpretation of a situation in a way that
iability (e.g., day-to-day fluctuations in the implementation of impacts emotional response (Horn & Maercker, 2016).
each strategy) and between-strategy variability (e.g., variability in Researchers have documented that IER strategies (i.e., corumina-
the amount that different emotion regulation strategies are used tion, codistraction, coproblem-solving, and coreappraisal) are used
each day). For instance, within-strategy variability is the amount day to day (Liu et al., 2021) and may influence markers of
1050 BATTAGLINI, RNIC, JAMESON, JOPLING, AND LEMOULT

psychological well-being (e.g., affect, Niven et al., 2009; or clinical of coproblem-solving to be more beneficial (i.e., associated with
symptoms, Christensen et al., 2020) in ways that parallel their lower negative affect and higher positive affect) in controllable as
intrapersonal counterparts (Gross, 1998). For instance, coreap- compared to uncontrollable contexts. Although corumination is
praisal, coproblem-solving, and codistraction generally predict also considered an emotion-focused coping strategy, corumination
adaptive outcomes, as measured by either reduced negative affect tends, on average, to predict greater negative affect and reduced
(Stone et al., 2019; Zelic et al., 2017) or fewer clinical symptoms positive affect (Battaglini et al., 2021; Zelic et al., 2017). Given
(Christensen et al., 2020). In contrast, corumination generally pre- that uncontrollable events also predict greater negative affect (van
dicts maladaptive psychological outcomes, as measured by either der Stouwe et al., 2019), the compounding effect of corumination
Content may be shared at no cost, but any requests to reuse this content in part or whole must go through the American Psychological Association.

increased negative affect (Zelic et al., 2017) or greater clinical and uncontrollable events is expected to be associated with less
symptoms (Rose, 2002). However, no study to date has examined adaptive affective outcomes (i.e., higher negative affect and lower
the impact that the flexible implementation of IER has on affect in positive affect).
day-to-day life. Aim 2 investigated the association between repertoire range and
affect. Both categorical and temporal variability were examined, in
The Current Study line with past empirical and theoretical research (Blanke et al., 2020;
This document is copyrighted by the American Psychological Association or one of its allied publishers.

Bonanno & Burton, 2013; Eldesouky & English, 2018). We


This preregistered study is the first to establish the concept of
hypothesized that higher categorical and temporal variability in IER
IER flexibility and to test whether IER flexibility is associated
strategy use would predict lower negative affect and higher positive
with short-term (i.e., day-to-day) adaptive outcomes. While doing
affect compared to lower categorical and temporal variability.
so, we addressed the notable limitations of the field in two ways.
For Aim 3, we investigated the association between day-to-day
First, we included the most commonly studied interpersonal strat-
feedback and affect, in line with the description of feedback pro-
egies: coreappraisal, coproblem-solving, corumination, and codis-
vided by Bonanno and Burton (2013). We assessed whether affect
traction (Rose, 2002; Stone et al., 2019; Zelic et al., 2017).
the previous day would moderate the association between change
Second, whereas emotion regulation flexibility research has
in IER strategy use and affect the subsequent day. We hypothe-
focused on long-term longitudinal and experimental designs, we
sized that at higher levels of negative affect the day prior, greater
sought to strengthen ecological validity by studying IER flexibility
change in IER strategy use from one day to the next would predict
as it unfolds day to day and in a naturalistic setting. Naturalistic
reduced negative affect. We also hypothesized that at lower levels
designs are critical as they enable researchers to extend examina-
of positive affect the day prior, greater change in IER strategy use
tions of what participants do in the laboratory to assess what par-
from one day to the next would predict higher positive affect.
ticipants actually do in everyday life. Thus, we implemented an
However, given that it is unclear whether feedback influences al-
intensive 14-day longitudinal (i.e., daily diary) design to assess the
association of day-to-day IER flexibility with daily negative and ready adaptive levels of affect, we did not make an a priori predic-
positive affect. tion regarding moderation effects of lower levels of negative
This study had four research aims. The first three aims applied affect or higher levels of positive affect the day prior.
the components of emotion regulation flexibility posited by Aim 4 investigated the flexibility of emotion regulation concep-
Bonanno and Burton (2013) to IER, and the fourth aim applied the tualized as the covariation of IER variability and environmental
definition of emotion regulation flexibility posited by Aldao et al. change (defined as the number of events experienced, a change in
(2015) to IER. Aim 1 tested the association between context sensi- event type, or a change in the stressfulness of the event from one
tivity and affect. We predicted that the perceived controllability of day to the next; Aldao et al., 2015). We predicted that higher cova-
negative events would moderate the association between IER strat- riation of IER variability and environmental change would predict
egy use and affect. Considering the similarities between intraper- lower negative affect and higher positive affect (see Table 1 and
sonal emotion regulation and IER strategies (i.e., IER constitutes Figures S1–S4 in the online supplemental material for an outline
the properties of intrapersonal emotion regulation translated to a of the study hypotheses).
dyadic context), the limited IER research to date, and the fact that To test these hypotheses, we recruited adults at the start of the
this study was the first to investigate context sensitivity with IER COVID-19 pandemic to complete daily diaries each day online for
strategies, the hypotheses for Aim 1 were based on the intraperso- 14 days. During each daily diary, participants reported (a) their
nal emotion regulation and coping literature (Haines et al., 2016; affect; (b) whether an event occurred that day; (c) the pleasantness,
Lazarus, 1993; Troy et al., 2013). At lower levels of perceived controllability, and stressfulness of the event; and (d) IER strat-
controllability of an event, an event cannot easily be modified; egies implemented that day.
thus, changing one’s emotional response (i.e., implementing emo-
tion-focused coping strategies) is expected to be most adaptive Method
(Lazarus, 1993; Troy et al., 2013). As such, we expected emotion-
focused intrapersonal emotion regulation strategies of coreap- Participants
praisal and codistraction to be more beneficial (i.e., associated
with lower negative affect and higher positive affect) in uncontrol- The sample consisted of 384 participants (Mage = 38.58 years,
lable as compared to controllable contexts (Lazarus & Folkman, SD = 13.82; see Supplemental Table S1) who were recruited from
1984; Troy et al., 2013). Alternatively, problem-solving is a strat- April to July 2020 as part of the larger “COVID-19 and Wellbeing
egy that is centered around changing a situation (i.e., problem- Study.” Advertisements were posted online to recruit participants
focused coping; Lazarus, 1993) and is likely to be most beneficial living in Canada and the United States. However, individuals liv-
in contexts that can be controlled. Thus, we expected higher levels ing in other countries could participate if they were fluent in
INTERPERSONAL EMOTION REGULATION FLEXIBILITY 1051

English and were above the age of 18 years old. For research of the event on a scale from 1 (not at all stressful) to 5 (extremely
Aims 1 and 4, only a subset of the sample could be used to test the stressful).
specific aims. For Aim 1, only participants who reported negative
events (n = 235) could be included. For Aim 4, only participants
Interpersonal Emotion Regulation
who reported negative, neutral, or positive events (n = 334) could We assessed whether an event occurred that day and the degree
be included. An a priori power analysis was not conducted because to which participants implemented each IER strategy across all
the data were collected to answer a range of research questions, interpersonal interactions, including all in-person conversations,
not exclusively the research questions in this study. Our sample online or text messaging, and phone or video calls. Specifically,
Content may be shared at no cost, but any requests to reuse this content in part or whole must go through the American Psychological Association.

size, which is consistent with recent recommendations (Lakens, participants rated the amount that they implemented each of the
2021), was based on the number of participants recruited during four IER strategies (i.e., coreappraisal, coproblem-solving, codis-
the time the study was running, past daily diary research (e.g., Seo traction, and corumination) during their interactions that day. Par-
& Patall’s, 2021, work showing that a sample of 20 participants ticipants rated their use of each IER strategy on a scale from 1 (not
was enough to obtain a medium effect size at Level 1 using multi- at all) to 5 (very much so). See the online supplemental materials
level models [MLMs]), authors’ previous experience with inten- for additional details.
This document is copyrighted by the American Psychological Association or one of its allied publishers.

sive longitudinal research (also reported in Blanke et al., 2020;


Wenzel et al., 2021), and resource constraints. Additionally, the The COVID-19 Pandemic
sample size exceeded statistical recommendations for obtaining
accurate estimates (i.e., regression coefficients, standard errors, Participants for this study were recruited during the COVID-19
variances) using our analytic approaches (i.e., multilevel modeling pandemic. The World Health Organization declared the COVID-
and regression; Austin & Steyerberg, 2015; Maas & Hox, 2005), 19 outbreak a global pandemic on March 11, 2020, and data col-
which recommend a sample greater than 50 participants (or Level lection for this study took place between April 20, 2020 and July
2 units) for MLMs and at least two participants per variable in 20, 2020. The pandemic was associated with many significant life-
regression models. style changes: Physical distancing requirements were instated, and
nonessential in-person businesses and schools were closed in
Materials many places worldwide. To better understand how participants
interacted with others during the COVID-19 pandemic, partici-
Affect pants reported the amount of time they spent each day interacting
in person, via online or text messaging, and via phone or video
We used the 20-item version of the Positive and Negative Affect
calls. Across the 14 days, participants spent the most time interact-
Schedule (Watson et al., 1988) to assess affect. The negative affect
ing with others in person (M = 311.19 min, SD = 253.50), fol-
items were distressed, upset, guilty, scared, hostile, irritable,
lowed by messaging (M = 89.14 min, SD = 109.47) and calls (M =
ashamed, nervous, jittery, and afraid. The between-person reliability
49.24 min, SD = 68.60). A one-way repeated-measures analysis of
was Rkf = .98, and the within-person reliability was Rc = .82 (Cran-
variance, with modality (in person, messaging, calls) as the
ford et al., 2006).1 The positive affect items were interested, excited,
within-subject factor, was conducted to examine whether time
strong, enthusiastic, proud, alert, inspired, determined, attentive, and
spent interacting with others significantly differed across modal-
active. The between-person reliability was Rkf =.99, and the within-
ities. Results showed a significant difference across modalities,
person reliability was Rc = .88 (Cranford et al., 2006). Participants
F(1.28, 476.47) = 285.76, p , .001. Pairwise comparisons with a
reported each emotion they experienced that day (consistent with
Bonferroni adjustment showed that participants interacted with
Shahar & Herr, 2011) on a scale from 1 (very slightly or not at all)
others significantly more in person compared to via messaging,
to 5 (extremely).
p , .001, 95% confidence interval (CI) [188.97, 255.13], or calls,
Event Type, Pleasantness, Controllability, and Stressfulness p , .001, 95% CI [228.49, 295.41]. Participants also interacted
with others significantly more via messaging than calls, p , .001,
Participants were asked to report on the most significant event 95% CI [25.60, 54.20].
that occurred each day (consistent with Genet & Siemer, 2012)
and to indicate the type, pleasantness, controllability, and stressful-
Procedure
ness of the event (Kircanski et al., 2015). Participants selected the
one category that best described the event: close friendship, social This study received approval from the University of British Co-
life, romantic relationship, family relationship, education, work, lumbia (UBC) Behavioural Research Ethics Board. After obtain-
finances, health of self, health of family, health of friends, recrea- ing consent, participants completed the baseline questionnaire,
tion, and other. Next, they rated the pleasantness, controllability, which assessed demographic characteristics and other measures as
and stressfulness of the event using items that were derived from part of the larger study. Participants who also consented to com-
past research (e.g., Kircanski et al., 2015). Participants were asked plete the daily diaries were sent the diaries through email at 5 p.m.
to report the pleasantness of the most significant event they experi- each day for 14 days after the completion of the baseline question-
enced that day on a scale from 1 (very unpleasant) to 5 (very naire. Participants did not receive remuneration for their
pleasant). Pleasantness ratings of 1 and 2 were considered nega- participation.
tive events in our analysis for research Aim 1. In addition, partici-
pants reported the controllability of the event on a scale from 1 1
All within- and between-person reliabilities were calculated using the
(very uncontrollable) to 5 (very controllable) and the stressfulness psych R package (Revelle, 2021).
1052 BATTAGLINI, RNIC, JAMESON, JOPLING, AND LEMOULT

Data Analysis analyses, participants (n = 235) were included if they reported a


negative event; a total of 474 negative events were reported.2 To
As described in the online supplemental materials, the data were assess context sensitivity, we conducted MLMs predicting affect
cleaned prior to conducting analyses based on our preregistration as a function of the degree to which the IER strategy was imple-
and study inclusion criteria. All analyses were conducted using R mented, the controllability of the event, and their interaction at
Version 4.4 (R Core Team, 2021) and the lme4 (Bates et al., 2015) Level 1. A separate MLM was conducted for each IER strategy
and lmerTest (Kuznetsova et al., 2017) R packages. Using random (i.e., coreappraisal, codistraction, coproblem-solving,3 corumina-
intercepts and slopes, MLMs were conducted to test research aims tion). Contrary to expectations, perceived controllability of the
with a nested structure to the data, in which days (Level 1) were
Content may be shared at no cost, but any requests to reuse this content in part or whole must go through the American Psychological Association.

negative event did not significantly moderate the association of


nested within people (Level 2). All continuous variables at Level 1 IER strategy use (i.e., coreappraisal, codistraction, coproblem-
were person-mean centered to examine within-person deviations solving, corumination) with negative affect at occasion t, jbsj #
from the participant’s average, and Level 2 variables were grand- .48, ps $ .270, or with positive affect at occasion t, jbsj # .33,
mean centered to parse out between-person effects (Bolger & Lau- ps $ .501 (see Supplemental Table S4).
renceau, 2013). Regression analyses were used for research aims
without a nested structure to the data (Blanke et al., 2020). The
This document is copyrighted by the American Psychological Association or one of its allied publishers.

Research Aim 2: Repertoire (Bonanno & Burton, 2013)


analyses were conducted independently for negative affect and pos-
itive affect (see Blanke et al., 2020; Thompson et al., 2012). Details Repertoire was indexed based on both categorical and temporal
regarding the computation of predictors can be found in Table 1. variability, as is consistent with theory and empirical research
(Blanke et al., 2020; Bonanno & Burton, 2013).4 A higher number of
Transparency and Openness strategies implemented indicated greater categorical variability. Par-
ticipants reported using an average of two IER strategies per day
In line with transparency and openness, the data, code, and preregis- (M = 1.77, SD = 1.41). MLMs were conducted with categorical vari-
tered hypotheses, design, and analytic plan for this study can be found ability predicting negative and positive affect. In line with our predic-
on the Open Science Framework (Battaglini, 2022): https://osf.io/ tion, categorical variability predicted less negative affect, b = 1.12,
yfz9r/?view_only=6db19703141f4c398dd125e65d94650b. Above and p , .001, 95% CI [1.68, .49]. This suggests that on days when
in the online supplemental materials, we report how we determined participants employed more IER strategies than their average, they
our sample size and a detailed description of our data inclusion/ reported less negative affect. In contrast, categorical variability did
exclusion criteria and data cleaning procedure, and we explicitly not significantly predict positive affect, b = .03, p = .942, [.63, .71].
describe the measures used for this study. There were no study To assess temporal variability, within- and between-strategy
manipulations to report. variability were analyzed based on past research (Blanke et al.,
2020).5 To assess within-strategy variability, negative and positive
Results affect (averaged across the 14 days) were regressed on within-
strategy variability. Because the average amount of strategy use
tends to be related to variability, the average amount of IER strat-
Preliminary Analyses
egy use across the 14 days was entered as a covariate to separate
To calculate the intraclass correlation coefficients (ICC), uncon- this effect from variability (consistent with Blanke et al., 2020). In
ditional models were conducted for negative and positive affect. contrast to our expectations, within-strategy variability did not sig-
The ICCs for research Aims 2 and 3 were conducted separately nificantly predict negative affect, b = .20, p = .845, 95% CI
from Aims 1 and 4 as the latter aims were analyzed with subsets of [2.23, 2.24], or positive affect, b = 1.96, p = .108, [4.19,
the total sample based on events reported. Specifically, only partici- .36].
pants who reported negative events could be included in Aim 1,
and only participants who reported negative, neutral, or positive 2
Even though the sample (n = 235) was included in these analyses, only
events could be included in Aim 4. The ICCs for negative affect participants who had data for variables in the model could be analyzed.
ranged from .53 to .58, indicating that 53% to 58% of the variability Whereas the majority of the models contained all 235 participants, the
models examining corumination consisted of n = 234 participants.
in negative affect was due to between-person variance. The ICCs 3
This model was near singular. To account for singularity, in an
for positive affect were approximated to .53, indicating that 53% of exploratory analysis, we conducted a fixed slope model instead of a random
the variability in positive affect was due to between-person var- slope model. This was supported by a likelihood ratio test that showed that
iance. Participants completed an average number of 7.29 (SD = random slopes did not significantly improve model fit, v2(2)= 0.01, p =
3.99) diaries across the 14 days. Between-person and within-person .994. Results were consistent with the original analyses reported, whereby
perceived controllability of the negative event did not significantly
correlations of the main study variables can be found in the online moderate the association of coproblem-solving with negative affect at
supplemental materials (see Supplemental Table S3). occasion t, b = 0.09, p = .824, 95% CI [0.76, 0.63].
4
Although the full sample (n = 384) was eligible for inclusion in these
Research Aim 1: Context Sensitivity (Bonanno & analyses, only participants who had data for assessing categorical
variability could be analyzed (n = 383).
Burton, 2013) 5
Although the entire sample (n = 384) was eligible for inclusion in the
Consistent with past research that investigated context sensitiv- within-strategy variability analyses, only participants who had data for
computing within-strategy variability could be analyzed (n = 378).
ity (Haines et al., 2016; Troy et al., 2013), we tested whether the Additionally, for between-strategy variability analyses, only participants
controllability of a negative event would moderate the association who had data for computing between-strategy variability could be analyzed
between IER strategy use and affect at occasion t. For these (n = 383).
This document is copyrighted by the American Psychological Association or one of its allied publishers.
Content may be shared at no cost, but any requests to reuse this content in part or whole must go through the American Psychological Association.

Table 1
Flexibility Constructs, Definitions, Research Aims, Hypotheses, and Computed Predictors
Flexibility construct Definition Research aim Hypothesis Computed predictors
Context sensitivity Appropriate alignment of emotion 1 We predicted that the perceived controll- Context sensitivity was indexed as the
regulation strategy with context ability of negative events would mod- interaction of the controllability of the
erate the association between IER event and the degree to which the IER
strategy use and affect. strategy was implemented.
Repertoire 2 We hypothesized that higher categorical
Categorical variability The number of strategies and temporal variability in IER strat- Categorical variability was computed by
implemented each day egy use would predict lower negative recoding IER strategies, whereby 1
affect and higher positive affect com- (no use of the strategy) was recoded
pared to lower categorical and tempo- as 0, indicating that the strategy was
ral variability. not implemented that day, and values
25 were recoded as 1, indicating
that the strategy was implemented
that day. Categorical variability was
quantified based on the total number
of IER strategies participants used
each day (Eldesouky & English,
2018).
Temporal variability
Within-strategy variability Fluctuations in the implementation Within-strategy variability was
of each strategy across the computed as the standard deviation
14-days of each IER strategy used across the
14 days, averaged across IER strat-
egies for each participant (Blanke
et al., 2020).
Between-strategy Fluctuations in the use of different Between-strategy variability was
variability IER strategies at one time point measured as the standard deviation
of IER strategies used each day
(Blanke et al., 2020).
Feedback When an individual incorporates 3 We hypothesized that at higher levels of Day-to-day changes in IER strategy use
internal feedback (i.e., their negative affect the day prior, greater were indexed by the average of the
affective state) and then change in IER strategy use from one absolute value of the difference in the
changes their IER strategy day to the next would predict reduced amount each IER strategy was used
use accordingly negative affect. We also hypothesized from one day to the next (i.e., from
that at lower levels of positive affect time t-1 to t).
INTERPERSONAL EMOTION REGULATION FLEXIBILITY

the day prior, greater change in IER


strategy use from one day to the next
would predict higher positive affect.
IER-environmental covariation 4 We predicted that higher covariation of
IER (within-strategy variabili- Joint fluctuation of the IER variability and environmental Within-strategy variability was
ty)-environmental covariation implementation of each strategy change would predict lower negative computed as the standard deviation
(number of events experienced) across the 14 days and the affect and higher positive affect. of each IER strategy used across the
number of distinct event types 14 days, averaged across IER strat-
experienced egies for each participant (Blanke
et al., 2020). Environmental vari-
ability was indexed based on the
number of distinct event types
reported by each participant across
the 14 days.
(table continues)
1053
1054 BATTAGLINI, RNIC, JAMESON, JOPLING, AND LEMOULT

To assess between-strategy variability, MLMs were conducted

codes indicated 0 = no change and 1 =


et al., 2020). Environmental variabili-
ty was indexed as the change in event
measured as the standard deviation of
IER strategies used each day (Blanke
to test the association between between-strategy variability and

type (from t-1 to t), where dummy


affect. At Level 1, between-strategy variability and the average
Between-strategy variability was

amount of IER strategy use each day were used to predict affect.
Computed predictors

The average amount of IER strategy use each day was included as
a covariate at Level 1 to separate its effects from between-strategy

change in event type.


variability (as recommended by Blanke et al., 2020). In line with
our hypothesis, between-strategy variability significantly predicted
less negative affect, b = 3.31, p , .001, 95% CI [4.79, 1.74].
Content may be shared at no cost, but any requests to reuse this content in part or whole must go through the American Psychological Association.

This suggests that on days when participants had greater between-


strategy variability than usual, meaning greater fluctuations in the
use of multiple IER strategies, they reported less negative affect.
Contrary to our prediction, between-strategy variability did not
significantly predict positive affect, b = .80, p = .408, [1.14,
This document is copyrighted by the American Psychological Association or one of its allied publishers.

2.73] (see Supplemental Table S5).

Research Aim 3: Feedback (Bonanno & Burton, 2013)


Hypothesis

Consistent with the theoretical formulation of feedback


(Bonanno & Burton, 2013),6 we predicted that the association
between day-to-day changes in IER use and affect would be mod-
erated by affect the previous day (Bonanno & Burton, 2013). To
test this hypothesis, we conducted MLMs predicting affect at time
t as a function of change in IER strategy use (from t-1 to t), affect
at t-1, and their interaction at Level 1.7 In line with our hypothesis,
negative affect at t-1 significantly moderated the association
measured as the standard deviation

indexed as the difference in event

between change in IER strategy use (from t-1 to t) and negative


of IER strategies used each day
Between-strategy variability was

Environmental variability was

affect at time t, b = .11, p = .017, 95% CI [.20, .02]. Thus,


stressfulness (from t-1 to t).

the association between change in IER strategy use (from t-1 to t)


Research aim

and less negative affect at time t was greater at higher levels of


(Blanke et al., 2020).

negative affect at t-1. Furthermore, the simple slopes analysis


showed that this association was significant at all levels of t-1 neg-
ative affect, b = 4.10, p , .001, [5.56, 2.64], for 1 standard
deviation above the mean, b = 3.61, p , .001, [5.02, 2.20],
at the mean, and b = 3.12, p , .001, [4.59, 1.65], for 1 stand-
ard deviation below the mean (Figure 1a).
Additionally, positive affect at t-1 significantly moderated the
different IER strategies each day

different IER strategies each day


and difference in event stressful-

association between change in IER strategy use (from t-1 to t) and


and change in event type from

ness from one day to the next

positive affect at time t, b = .14, p = .012, 95% CI [.24, .03].


Joint fluctuation of the use of

Joint fluctuation of the use of

Thus, the association between change in IER strategy use (from t-


1 to t) and more positive affect at time t was greater at lower levels
Definition

one day to the next

of positive affect at t-1. The simple slopes analysis was not signifi-
cant, jbsj # .93, ps $ .337. However, the Johnson-Neyman test
for the regions of significance (Hayes & Matthes, 2009; Johnson
& Neyman, 1936) showed that when t-1 positive affect was below
Note. IER = interpersonal emotion regulation.

17.94 (3.24 standard deviations below the mean), greater change


in IER strategy use (from t-1 to t) predicted higher positive affect
variability)-environmental cova-

variability)-environmental cova-
riation (change in event stress-

6
riation (change in event type

Even though the entire sample (n = 384) was included in these


analyses, only participants who had data for computing the lagged
Flexibility construct

predictors could be analyzed (n = 328).


fulness from t-1 to t)

7
IER (between-strategy

IER (between-strategy

As an exploratory analysis, the average amount of IER strategy use


Table 1 (continued)

each day was included at Level 1. The direction of moderation effects


remained the same for both negative and positive affect, though they were
from t-1 to t)

only marginally significant for negative affect. Negative affect at t-1


moderated the association between change in IER strategy use (from t-1 to
t) and negative affect at time t, b = 0.09, p = .067, 95% CI [0.18, 0.00].
Positive affect at t-1 moderated the association between change in IER
strategy use (from t-1 to t) and positive affect at time t, b = 0.14, p = .012,
95% CI [0.24, 0.02].
INTERPERSONAL EMOTION REGULATION FLEXIBILITY 1055

Figure 1
Research Aim 3: Feedback Associated With Negative and Positive Affect
Content may be shared at no cost, but any requests to reuse this content in part or whole must go through the American Psychological Association.
This document is copyrighted by the American Psychological Association or one of its allied publishers.

Note. Panel a: The association between change in interpersonal emotion regulation strategy use (from t-1 to t) and less negative affect was stronger at
higher levels of t-1 negative affect but was significant at all observed levels of t-1 negative affect. Panel b: Positive affect (at t-1) significantly moderated
the association between change in interpersonal emotion regulation strategy use (from t-1 to t) and positive affect. Because the simple slopes analysis was
not significant, the Johnson-Neyman test for the regions of significance is presented to elucidate the interaction. Results showed that when t-1 positive
affect was below 17.94, greater change in interpersonal emotion regulation strategy use (from t-1 to t) predicted higher positive affect. Additionally,
when t-1 positive affect was above 25.12, participants who reported a greater change in interpersonal emotion regulation strategy use (from t-1 to t)
reported less positive affect. However, this range is beyond the range of observable data. Negative affect (t-1), positive affect (t-1), and change in inter-
personal emotion regulation strategy use were person-mean centered.

at time t. Additionally, when t-1 positive affect was above 25.12 of within-strategy IER variability and environmental variability sig-
(4.54 standard deviations above the mean), greater change in IER nificantly predicted higher negative affect, b = 2.91, p = .001, 95%
strategy use (from t-1 to t) predicted less positive affect, although CI [.77, 5.23], and did not significantly predict positive affect, b =
this range was beyond the range of observable data (Figure 1b; .38, p = .719, [1.90, 2.58].
Supplemental Table S6). In terms of the covariation of between-strategy variability and
environmental variability, an MLM was conducted with between-
Research Aim 4: Covariation of Interpersonal Emotion strategy variability regressed onto the change in event type (from
Regulation Strategy Variability and Environmental t-1 to t).9 The residuals from this model were saved and used in
Variability (Aldao et al., 2015)
8
Although n = 334 participants were eligible for inclusion in these
In accordance with the theory articulated by Aldao et al. (2015), analyses, only participants that had within-strategy variability data could be
we hypothesized that greater covariation of IER strategy variabili- analyzed for the within-strategy variability analyses (n = 251, which is
ty and environmental variability would predict lower negative appropriate for unbiased estimates using regression analyses; Austin &
Steyerberg, 2015). For between-strategy variability, only participants with
affect and higher positive affect. The analyses were conducted in data for the lagged predictors could be analyzed (n = 156, which is
line with suggestions made by Aldao et al. (2015). Participants sufficient for accurate estimates when conducting MLM; Maas & Hox,
(n = 334) who reported the occurrence of events were included in 2005).
9
these analyses, with a total of 1,090 (negative, neutral, or positive; This model was near singular. To account for singularity, in an
see Supplemental Table S2) events reported. Within-strategy vari- exploratory analysis, we extracted the residuals from a fixed slope model
instead of a random slope model. This was supported by a Likelihood ratio
ability was regressed onto the number of distinct event types, and test (LRT) which showed that random slopes did not significantly improve
residuals were saved. Positive residuals indicate greater covaria- model fit v2(2) = 0.08, p = .961. The residuals from the fixed slope model
tion of IER variability and environmental variability, whereas nega- were then used to predict affect. Results were consistent with the original
tive residuals indicate lower covariation of IER and environmental analyses reported, whereby the covariation of between-strategy IER
variability and environmental variability (change in event type from t-1 to
variability (Aldao et al., 2015). The residuals were included as the t) did not significantly predict negative affect, b = 2.85, p = .231, 95% CI
predictor in regression models predicting negative and positive [7.59, 2.50], or positive affect, b = 2.16, p = .450, 95% CI [7.66,
affect, respectively.8 Contrary to expectations, greater covariation 3.62].
1056 BATTAGLINI, RNIC, JAMESON, JOPLING, AND LEMOULT

MLMs to predict negative and positive affect, respectively. The immediate process unfolding in the laboratory (Birk & Bonanno,
residuals were person-mean centered at Level 1 and grand-mean 2016), the present study examined feedback as it unfolds day-
centered at Level 2. Contrary to our prediction, the covariation of to-day. It is unclear whether feedback has the same meaning across
between-strategy IER variability and environmental variability these different timescales. It is possible our metric of feedback pro-
(change in event type from t-1 to t) did not significantly predict vides a slower, top-down feedback process whereby an individual
negative affect, b = 2.92, p = .219, 95% CI [7.26, 1.58], or pos- assesses internal affective information and modifies the degree to
itive affect, b = 2.08, p = .467, [8.18, 3.01]. which IER strategies are implemented from one day to the next.
Furthermore, we conducted an MLM with between-strategy var- Overall, these findings suggest that there are benefits to implement-
Content may be shared at no cost, but any requests to reuse this content in part or whole must go through the American Psychological Association.

iability regressed onto the difference in event stressfulness (from ing a breadth of IER strategies (repertoire) and adjusting IER strat-
t-1 to t).10 The residuals from this model were used as predictors egy use (feedback), which is akin to findings from the extant
for MLMs predicting negative and positive affect, respectively. intrapersonal emotion regulation flexibility research (e.g., South-
The residuals were person-mean centered at Level 1 and grand- ward & Cheavens, 2020). It is interesting to consider whether fac-
mean centered at Level 2. Contrary to our hypothesis, greater tors related to one’s social network might facilitate or thwart IER
covariation of between-strategy IER variability and environmental flexibility. A large social network, for example, may offer more
This document is copyrighted by the American Psychological Association or one of its allied publishers.

variability (indexed based on the difference in event stressfulness) opportunities to implement a broader range of IER strategies,
did not significantly predict negative affect, b = 3.09, p = .222, which may promote greater categorical variability, between-strat-
95% CI [8.46, 2.02], or positive affect, b = 2.14, p = .478, egy variability, and feedback. Indeed, the social support literature
[8.01, 4.17] (see Supplemental Table S7). shows that greater social support offers a myriad of beneficial out-
comes (e.g., reduced mortality risk; Shor et al., 2013). Alterna-
Additional Analyses That Were Not Preregistered tively, it may be that social networks high in closeness or
attunement could facilitate greater IER flexibility because IER may
We conducted several follow-up nonpreregistered analyses in
be encouraged. Interestingly, physical distancing requirements dur-
order to test the robustness of our findings. We examined whether
ing the COVID-19 pandemic may have restricted the size of one’s
any demographic characteristics of the sample were associated
social network and altered the modality in which some people
with the number of daily diaries completed, and we found that sig-
communicated (e.g., digital instead of in-person). Digital commu-
nificant effects remained when relevant demographic characteris-
nication can limit nonverbal behaviors that facilitate interpersonal
tics were included as covariates. Similarly, we conducted
intimacy (Andersen et al., 2006). Although research by Battaglini
exploratory analyses to account for the differences in the number
and colleagues (2021) documented that individuals engage in IER
of daily diaries completed and found the pattern of significant
across in-person and digital modalities, future work should con-
results remained the same. Given the number of analyses con-
tinue to explore the influence of communication modality on asso-
ducted, we used the Benjamini Hochberg approach to account for
ciations between IER and affect. Additionally, future research
the false discovery rate from conducting multiple tests (Benjamini
may seek to determine which characteristics of one’s support net-
& Hochberg, 1995). Notably, all significant effects remained.
work (e.g., size, closeness) enhance IER flexibility, even beyond
Finally, to verify the null findings that we reported above, we con-
the context of COVID-19.
ducted a Bayesian analysis known as the Bayes factor, and all
The remaining components of IER flexibility showed mixed
Bayes factor results supported the null hypotheses. Additional
findings regarding their association with affect. First, contrary to
details on all supplemental analyses can be found in the online
our expectations based on Aldao et al. (2015), the covariation of
supplemental materials.
IER within-strategy variability and environmental variability,
defined as the number of distinct event types, prospectively pre-
Discussion dicted higher negative affect and did not predict positive affect.
This study was the first to reconceptualize the adaptiveness of Moreover, the covariation of IER between-strategy variability and
IER based on one’s ability to flexibly implement IER strategies environmental variability did not significantly predict affect. There
given contextual demands. IER flexibility has notable empirical are three possible explanations for these unexpected findings. One
and practical contributions. The interaction of IER strategies with possibility is that, unlike intrapersonal emotion regulation, there
contextual factors is largely absent in the IER literature to date, may be other factors contributing to the adaptiveness of the cova-
and investigating IER flexibility fills this gap. This empirical shift riation of IER and environmental variability. For instance, IER-
may also inform recommendations regarding the use of IER in environmental covariation may be more adaptive when IER
daily life. Whereas the current literature has evaluated the adap- involves more self-disclosure about the environmental change,
tiveness of particular strategies based on their intransient out- perhaps resulting in a richer discussion and the use of IER strat-
comes, findings from the current study suggest there are affective egies to directly regulate the emotions elicited by the environmen-
benefits to flexibly accounting for internal and environmental fac- tal change. Consistent with this proposition, researchers have
tors when implementing IER strategies. Thus, IER flexibility has
the potential to provide a more thorough understanding of IER 10
The pre-registration for the analysis of the covariation of between-
adaptiveness in daily life. strategy variability and environmental variability in predicting affect did
As expected, based on Bonanno and Burton (2013), both reper- not correctly indicate the variables that were included at Level 2 and grand-
mean centered. Indeed, only the predictors of the model can be grand-mean
toire (i.e., greater categorical variability and between-strategy vari- centered at Level 2; therefore, change in event type, difference in event
ability) and feedback predicted adaptive affective outcomes. stressfulness, and the residuals were grand-mean centered at Level 2 in
Interestingly, whereas past research measured feedback as an their respective models, as described above.
INTERPERSONAL EMOTION REGULATION FLEXIBILITY 1057

documented that certain aspects of a dyadic interaction influence past research has documented that context controllability influences
the adaptiveness of corumination (e.g., talking about the problem the adaptiveness of cognitive reappraisal (i.e., cognitive reappraisal is
may be more adaptive than focusing on negative affect; Rose associated with maladaptive outcomes in more controllable contexts
et al., 2014). A second possibility for the unexpected findings is and adaptive outcomes in less controllable contexts; Haines et al.,
that there are multiple ways of defining environmental variability 2016; Troy et al., 2013), but it has yet to be examined with other
(Aldao et al., 2015). In the current study, we defined environmen- strategies. Interestingly, these findings may not translate to coreap-
tal variability as the number of distinct event types, day-to-day praisal. It is possible that there are differences in the process and out-
change in event type, and day-to-day change in event stressfulness. comes of intrapersonal emotion regulation versus IER. Alternatively,
Content may be shared at no cost, but any requests to reuse this content in part or whole must go through the American Psychological Association.

These may not be the facets of environmental variability that are the day-to-day controllability of events may have been diluted by the
most relevant to IER flexibility. Thus, additional research is COVID-19 pandemic, which may be perceived as an uncontrollable
needed to investigate meaningful markers of environmental context (Usher et al., 2020). Future research is needed to investigate
change in the context of IER flexibility. Third, participants whether context sensitivity is unique to intrapersonal emotion regula-
reported on the most significant event they experienced that day. tion strategies (e.g., cognitive reappraisal; Haines et al., 2016; Troy
Thus, we indexed variability of events across all diaries, rather et al., 2013) or whether it applies to IER in particular contexts only,
than within one single day, which may partially explain this unex- including post-COVID-19 pandemic.
This document is copyrighted by the American Psychological Association or one of its allied publishers.

pected finding for research Aim 4. Furthermore, the lockdown Overall, findings from the current study suggest that not all compo-
(closing of schools and nonessential businesses) and physical dis- nents of intrapersonal emotion regulation flexibility predicted the
tancing requirements necessitated by the COVID-19 pandemic adaptiveness of IER flexibility in day-to-day life. This may be due, in
may have restricted the range of possible events individuals expe- part, to the fundamental differences between intrapersonal emotion
rienced. To expand variability in examining IER-environmental regulation and IER. For example, given the dyadic nature of IER, the
covariation, future research may consider investigating these ques- type of relationship (e.g., friend vs. romantic partner) may play a role
tions post-COVID-19 pandemic. in predicting the adaptiveness of IER strategies in day-to-day life
Also contrary to our expectations, repertoire showed mixed find- (White & Shih, 2012). In addition, the quality of the relationship may
ings regarding its association with affect. Specifically, when reper- impact the effectiveness of IER. Although this has yet to be investi-
toire was defined as within-strategy variability (Bonanno & Burton, gated with respect to most IER strategies (e.g., coreappraisal, codis-
2013), it did not significantly predict negative or positive affect, traction, coproblem-solving), research shows that corumination has
which is consistent with past research (Blanke et al., 2020). In con- maladaptive consequences in low-quality relationships, but not in
trast, when repertoire was defined as categorical variability or high-quality relationships (Guassi Moreira et al., 2016). Moreover,
between-strategy variability, it predicted less negative affect. Taken individuals may prefer to implement some IER strategies over others,
together, these findings suggest that the use of a greater number of which could result in a mismatch between the strategy preferred by
IER strategies (categorical variability) and the variable use of IER the IER receiver and provider. Finally, unlike intrapersonal emotion
across strategies (between-strategy variability) is more adaptive than regulation, IER is more likely to be a verbal process, which can affect
the variable use of individual strategies (within-strategy variability). the way information is processed and its emotional effects (Stanton &
These findings underscore the importance of having a broad range Low, 2012). These are interesting directions for future research that
of IER strategies accessible rather than a limited range (e.g., using would allow us to better understand the bounds of adaptive IER
predominantly one IER strategy). In addition, it is important to note flexibility.
that although greater categorical and between-strategy variability
predicted less negative affect, it did not predict positive affect. Inter- Limitations and Future Directions
estingly, this is consistent with previous work documenting that neg-
ative and positive affect are orthogonal in nature (Kercher, 1992; There are multiple strengths and limitations of this study. First,
Watson et al., 1988). This may indicate affective differentiation although there are strengths of the daily diary design (e.g., daily
regarding the adaptiveness of repertoire, which may have implica- assessments mitigate participant burden), future research could
tions for individuals’ regulatory goals (Sheppes et al., 2014). For examine the adaptiveness of IER flexibility on a shorter timescale
instance, greater categorical variability and between-strategy vari- using ecological momentary assessment (EMA). EMA would
ability may be effective for down-regulating negative affect, but not facilitate the temporal ordering of event occurrence, affect, and
for up-regulating positive affect. Future research is required to inves- subsequent event-specific emotion regulation that occur in close
tigate the association between regulatory goal attainment (e.g., proximity. EMA would also facilitate the assessment of multiple
down-regulating or up-regulating affect; Sheppes et al., 2014) and events (of varying valence) each day. Second, the present study
the implementation of IER flexibility. assessed the adaptiveness of IER flexibility based exclusively on
Finally, in terms of context sensitivity (Bonanno & Burton, 2013), its short-term affective consequences. Long-term longitudinal
the perceived controllability of negative events did not moderate the designs are needed to elucidate the long-term effects of IER flexi-
association of coreappraisal, codistraction, coproblem-solving, or bility (e.g., on psychopathology symptoms), which is another indi-
corumination with negative or positive affect. One possible explana- cator of emotion regulation adaptiveness (Aldao et al., 2010).
tion for these findings is that the association between IER strategy Third, although this study investigated a breadth of IER strategies,
use and affect may be moderated by other contexts, such as perceived future research on IER flexibility should incorporate additional
agency in one’s ability to regulate their own emotions. The imple- IER strategies as the IER literature expands to elucidate these con-
mentation of IER (reducing one’s burden of regulating in silo) may structs. Seminal to reaching this goal is the establishment of vali-
predict adaptive affective outcomes in contexts where one experien- dated measures of state IER. In the current study, state IER items
ces low agency in their ability to regulate independently. To date, were derived from intrapersonal emotion regulation measures or
1058 BATTAGLINI, RNIC, JAMESON, JOPLING, AND LEMOULT

trait IER questionnaires. It is noteworthy that some IER strategies M. L. Patterson (Eds.), The Sage handbook of nonverbal communication
(e.g., coreappraisal) are assessed in the past tense, whereas others (pp. 259–278). Sage, Inc. https://doi.org/10.4135/9781412976152.n14
(e.g., codistraction) are assessed based on the act of using the strat- Austin, P. C., & Steyerberg, E. W. (2015). The number of subjects per
egy. Future research might examine whether differences in the variable required in linear regression analyses. Journal of Clinical Epi-
demiology, 68(6), 627–636. https://doi.org/10.1016/j.jclinepi.2014.12
phrasing of items influences the findings reported here. Fourth, a
.014
considerable strength of this study is the assessment of IER with
Bates, M., Mächler, M., Bolker, B., & Walker, S. (2015). Fitting linear
others generally, without restricting participants to report on one mixed-effects models using lme4. Journal of Statistical Software, 67(1),
type of relationship (e.g., with a romantic partner). This allows a 1–48. https://doi.org/10.18637/jss.v067.i01
Content may be shared at no cost, but any requests to reuse this content in part or whole must go through the American Psychological Association.

broader and more thorough assessment of IER as it occurs with Battaglini, A. (2022, January 12). COVID-19 study – Interpersonal emo-
others more generally. Future research may explore whether there tion regulation flexibility, daily diary, negative and positive affect. osf
are different affective outcomes of IER flexibility across different .io/yfz9r
types of relationships. Battaglini, A. M., Rnic, K., Tracy, A., Jopling, E., & LeMoult, J. (2021).
Taken together, this study informs a notable future direction for Co-rumination across in-person and digital communication: Associa-
the field of emotion regulation. This is the only study to date to find tions with affect and relationship closeness in adolescents. Journal of
Adolescence, 89(1), 161–169. https://doi.org/10.1016/j.adolescence
that the flexible implementation of IER strategies is important for
This document is copyrighted by the American Psychological Association or one of its allied publishers.

.2021.04.011
mitigating negative affect and/or enhancing positive affect. In addi-
Benjamini, Y., & Hochberg, Y. (1995). Controlling the false discovery
tion, we found that some components of IER flexibility, but not rate: A practical and powerful approach to multiple testing. Journal of
others, were associated with adaptive affective outcomes. As such, the Royal Statistical Society B: Methodological, 57(1), 289–300. https://
additional research is needed to better understand the components doi.org/10.1111/j.2517-6161.1995.tb02031.x
that determine the day-to-day adaptiveness of IER flexibility, spe- Birk, J. L., & Bonanno, G. A. (2016). When to throw the switch: The adap-
cifically. Additionally, a strength of this current study is examining tiveness of modifying emotion regulation strategies based on affective
IER flexibility in adults during the COVID-19 pandemic when and physiological feedback. Emotion, 16(5), 657–670. https://doi.org/10
emotion regulation may be necessary to manage the consequences .1037/emo0000157
of the pandemic (Restubog et al., 2020; Wen et al., 2021). How- Blanke, E. S., Brose, A., Kalokerinos, E. K., Erbas, Y., Riediger, M., &
Kuppens, P. (2020). Mix it to fix it: Emotion regulation variability in daily
ever, future research should extend the generalizability of our find-
life. Emotion, 20(3), 473–485. https://doi.org/10.1037/emo0000566
ings by examining our research aims in other samples (e.g., Bolger, N., & Laurenceau, J. P. (2013). Methodology in the social scien-
adolescents) and contexts (i.e., beyond the COVID-19 pandemic). ces. Intensive longitudinal methods: An introduction to diary and expe-
rience sampling research. Guilford Press.
Conclusion Bonanno, G. A., & Burton, C. L. (2013). Regulatory flexibility: An indi-
vidual differences perspective on coping and emotion regulation. Per-
The current study presents a shift in the way IER adaptability is spectives on Psychological Science, 8(6), 591–612. https://doi.org/10
defined. Based on the findings of the current study, we argue that .1177/1745691613504116
there is reason to consider flexibility when assessing IER adaptive- Chen, S., & Bonanno, G. A. (2021). Components of emotion regulation
ness. Applying the two core theoretical models of emotion regula- flexibility: Linking latent profiles to depressive and anxious symptoms.
tion flexibility (Aldao et al., 2015; Bonanno & Burton, 2013) to the Clinical Psychological Science, 9(2), 236–251. https://doi.org/10.1177/
assessment of IER flexibility, this study found that greater IER strat- 2167702620956972
egy repertoire (i.e., categorical, between-strategy variability) and Christensen, K. A., Seager van Dyk, I., Nelson, S. V., & Vasey, M. W.
(2020). Using multilevel modeling to characterize interpersonal emotion
feedback predicted affective benefits (i.e., less negative affect and/or
regulation strategies and psychopathology in female friends. Personality
more positive affect). However, IER flexibility defined by greater
and Individual Differences, 165, Article 110156. https://doi.org/10
IER-environmental variability, repertoire (i.e., within-strategy vari- .1016/j.paid.2020.110156
ability), or context sensitivity was not associated with adaptive Connor-Smith, J. K., Compas, B. E., Wadsworth, M. E., Thomsen, A. H.,
affective outcomes. Findings present a novel view of IER flexibility & Saltzman, H. (2000). Responses to stress in adolescence: Measure-
and provide the first evidence of flexibility components determining ment of coping and involuntary stress responses. Journal of Consulting
day-to-day IER adaptiveness. As such, the current study marks the and Clinical Psychology, 68(6), 976–992. https://doi.org/10.1037/0022-
beginning of an emerging area of investigation that has important 006X.68.6.976
implications for emotion regulation theory and research, as well as Cranford, J. A., Shrout, P. E., Iida, M., Rafaeli, E., Yip, T., & Bolger, N.
implications for the adaptiveness of social support in the context of (2006). A procedure for evaluating sensitivity to within-person change:
Can mood measures in diary studies detect change reliably? Personality
emotion regulation.
and Social Psychology Bulletin, 32(7), 917–929. https://doi.org/10
.1177/0146167206287721
Eisenberg, N., Cumberland, A., & Spinrad, T. L. (1998). Parental social-
References
ization of emotion. Psychological Inquiry, 9(4), 241–273. https://doi
Aldao, A., Nolen-Hoeksema, S., & Schweizer, S. (2010). Emotion-regulation .org/10.1207/s15327965pli0904_1
strategies across psychopathology: A meta-analytic review. Clinical Psy- Eldesouky, L., & English, T. (2018). Another year older, another year
chology Review, 30(2), 217–237. https://doi.org/10.1016/j.cpr.2009.11.004 wiser? Emotion regulation strategy selection and flexibility across adult-
Aldao, A., Sheppes, G., & Gross, J. J. (2015). Emotion regulation flexibil- hood. Psychology and Aging, 33(4), 572–585. https://doi.org/10.1037/
ity. Cognitive Therapy and Research, 39(3), 263–278. https://doi.org/10 pag0000251
.1007/s10608-014-9662-4 Gelb, Y. (2013). Co-distraction and co-rumination in the friendships of
Andersen, P. A., Guerrero, L. K., & Jones, S. M. (2006). Nonverbal behav- undergraduate college students [Doctoral dissertation, Alliant Interna-
ior in intimate interactions and intimate relationships. In V. Manusov & tional University]. https://doi.org/10.1037/e586212013-001
INTERPERSONAL EMOTION REGULATION FLEXIBILITY 1059

Genet, J. J., & Siemer, M. (2012). Rumination moderates the effects of Morey, R. D., & Rouder, J. N. (2011). Bayes factor approaches for testing
daily events on negative mood: Results from a diary study. Emotion, interval null hypotheses. Psychological Methods, 16(4), 406–419.
12(6), 1329–1339. https://doi.org/10.1037/a0028070 https://doi.org/10.1037/a0024377
Gross, J. J. (1998). The emerging field of emotion regulation: An integra- Nicolai, K. A., Laney, T., & Mezulis, A. H. (2013). Different stressors, dif-
tive review. Review of General Psychology, 2(3), 271–299. https://doi ferent strategies, different outcomes: How domain-specific stress
.org/10.1037/1089-2680.2.3.271 responses differentially predict depressive symptoms among adoles-
Gross, J. J., & John, O. P. (2003). Individual differences in two emotion cents. Journal of Youth and Adolescence, 42(8), 1183–1193. https://doi
regulation processes: Implications for affect, relationships, and well- .org/10.1007/s10964-012-9866-4
being. Journal of Personality and Social Psychology, 85(2), 348–362. Niven, K., Totterdell, P., & Holman, D. (2009). A classification of con-
Content may be shared at no cost, but any requests to reuse this content in part or whole must go through the American Psychological Association.

https://doi.org/10.1037/0022-3514.85.2.348 trolled interpersonal affect regulation strategies. Emotion, 9(4),


Guassi Moreira, J. F., Miernicki, M. E., & Telzer, E. H. (2016). Relation- 498–509. https://doi.org/10.1037/a0015962
ship quality buffers association between co-rumination and depressive R Core Team. (2021). R: A language and environment for statistical
symptoms among first year college students. Journal of Youth and Ado- computing [Computer software]. https://www.R-project.org/
lescence, 45(3), 484–493. https://doi.org/10.1007/s10964-015-0396-8 Restubog, S. L. D., Ocampo, A. C. G., & Wang, L. (2020). Taking control
Haines, S. J., Gleeson, J., Kuppens, P., Hollenstein, T., Ciarrochi, J., amidst the chaos: Emotion regulation during the COVID-19 pandemic.
Labuschagne, I., Grace, C., & Koval, P. (2016). The wisdom to know Journal of Vocational Behavior, 119, Article 103440. https://doi.org/10
This document is copyrighted by the American Psychological Association or one of its allied publishers.

the difference: Strategy-situation fit in emotion regulation in daily life is .1016/j.jvb.2020.103440


associated with well-being. Psychological Science, 27(12), 1651–1659. Revelle, W. (2021). psych: Procedures for psychological, psychometric,
https://doi.org/10.1177/0956797616669086 and personality research [Computer software]. Northwestern University.
Hayes, A. F., & Matthes, J. (2009). Computational procedures for probing Rose, A. J. (2002). Co-rumination in the friendships of girls and boys.
interactions in OLS and logistic regression: SPSS and SAS implementa- Child Development, 73(6), 1830–1843. https://doi.org/10.1111/1467-
tions. Behavior Research Methods, 41(3), 924–936. https://doi.org/10 8624.00509
.3758/BRM.41.3.924 Rose, A. J., Schwartz-Mette, R. A., Glick, G. C., Smith, R. L., & Luebbe,
Horn, A. B., & Maercker, A. (2016). Intra- and interpersonal emotion regu- A. M. (2014). An observational study of co-rumination in adolescent
lation and adjustment symptoms in couples: The role of co-brooding and friendships. Developmental Psychology, 50(9), 2199–2209. https://doi
co-reappraisal. BMC Psychology, 4(1), Article 51. https://doi.org/10 .org/10.1037/a0037465
.1186/s40359-016-0159-7 Seo, E., & Patall, E. A. (2021). Feeling proud today may lead people to
Jeffreys, H. (1961). The theory of probability (3rd ed.). Oxford University coast tomorrow: Daily intraindividual associations between emotion and
Press. effort in academic goal striving. Emotion, 21(4), 892–897. https://doi
Johnson, P., & Neyman, J. (1936). Tests of certain linear hypotheses and .org/10.1037/emo0000752
their applications to some educational problems. Statistical Research Shahar, B., & Herr, N. R. (2011). Depressive symptoms predict inflexibly
Memoirs, 1, 57–93. high levels of experiential avoidance in response to daily negative affect:
Kashdan, T. B., & Steger, M. F. (2006). Expanding the topography of A daily diary study. Behaviour Research and Therapy, 49(10), 676–681.
social anxiety. An experience-sampling assessment of positive emotions, https://doi.org/10.1016/j.brat.2011.07.006
positive events, and emotion suppression. Psychological Science, 17(2), Sheppes, G., Scheibe, S., Suri, G., Radu, P., Blechert, J., & Gross, J. J.
120–128. https://doi.org/10.1111/j.1467-9280.2006.01674.x (2014). Emotion regulation choice: A conceptual framework and sup-
Kercher, K. (1992). Assessing subjective well-being in the old-old: The porting evidence. Journal of Experimental Psychology: General, 143(1),
PANAS as a measure of orthogonal dimensions of positive and negative 163–181. https://doi.org/10.1037/a0030831
affect. Research on Aging, 14(2), 131–168. https://doi.org/10.1177/ Shor, E., Roelfs, D. J., & Yogev, T. (2013). The strength of family ties: A
0164027592142001 meta-analysis and meta-regression of self-reported social support and
Kircanski, K., Thompson, R. J., Sorenson, J., Sherdell, L., & Gotlib, I. H. mortality. Social Networks, 35(4), 626–638. https://doi.org/10.1016/j
(2015). Rumination and worry in daily life: Examining the naturalistic .socnet.2013.08.004
validity of theoretical constructs. Clinical Psychological Science, 3(6), Southward, M. W., & Cheavens, J. S. (2020). Quality or quantity? A multi-
926–939. https://doi.org/10.1177/2167702614566603 study analysis of emotion regulation skills deficits associated with bor-
Kuznetsova, A., Brockhoff, P. B., & Christensen, R. H. B. (2017). lmerT- derline personality disorder. Personality Disorders: Theory, Research,
est package: Tests in linear mixed effects models. Journal of Statistical and Treatment, 11(1), 24–35. https://doi.org/10.1037/per0000357
Software, 82(13), 1–26. https://doi.org/10.18637/jss.v082.i13 Stanton, A. L., & Low, C. A. (2012). Expressing emotions in stressful con-
Lakens, D. (2021). Sample size justification. PsyArXiv. https://doi.org/10 texts: Benefits, moderators, and mechanisms. Current Directions in Psy-
.31234/osf.io/9d3yf chological Science, 21(2), 124–128. https://doi.org/10.1177/0963721
Lazarus, R. S. (1993). Coping theory and research: Past, present, and 411434978
future. Psychosomatic Medicine, 55(3), 234–247. https://doi.org/10 Stone, L. B., Mennies, R. J., Waller, J. M., Ladouceur, C. D., Forbes,
.1097/00006842-199305000-00002 E. E., Ryan, N. D., Dahl, R. E., & Silk, J. S. (2019). Help me feel better!
Lazarus, R. S., & Folkman, S. (1984). Stress, appraisal, and coping. Ecological momentary assessment of anxious youths’ emotion regula-
Springer. tion with parents and peers. Journal of Abnormal Child Psychology,
Liu, D. Y., Strube, M. J., & Thompson, R. J. (2021). Interpersonal emotion 47(2), 313–324. https://doi.org/10.1007/s10802-018-0454-2
regulation: An experience sampling study. Affective Science, 2, 273–288. Thompson, R. J., Mata, J., Jaeggi, S. M., Buschkuehl, M., Jonides, J., &
https://doi.org/10.1007/s42761-021-00044-y Gotlib, I. H. (2012). The everyday emotional experience of adults with
Lougheed, J. P., & Hollenstein, T. (2012). A limited repertoire of emotion major depressive disorder: Examining emotional instability, inertia, and
regulation strategies is associated with internalizing problems in adoles- reactivity. Journal of Abnormal Psychology, 121(4), 819–829. https://
cence. Social Development, 21(4), 704–721. https://doi.org/10.1111/j doi.org/10.1037/a0027978
.1467-9507.2012.00663.x Troy, A. S., Shallcross, A. J., & Mauss, I. B. (2013). A person-by-situation
Maas, C. J., & Hox, J. J. (2005). Sufficient sample sizes for multilevel approach to emotion regulation: Cognitive reappraisal can either help or
modeling. Methodology, 1(3), 86–92. https://doi.org/10.1027/1614-2241 hurt, depending on the context. Psychological Science, 24(12),
.1.3.86 2505–2514. https://doi.org/10.1177/0956797613496434
1060 BATTAGLINI, RNIC, JAMESON, JOPLING, AND LEMOULT

Usher, K., Jackson, D., Durkin, J., Gyamfi, N., & Bhullar, N. (2020). Pan- Wen, J. H., Klaiber, P., DeLongis, A., Slavish, D. C., & Sin, N. L. (2021).
demic-related behaviours and psychological outcomes; A rapid literature Day-to-day associations between nightly sleep and next-day well-being
review to explain COVID-19 behaviours. International Journal of Men- amid the COVID-19 pandemic in North America. Sleep Health, 7(6),
tal Health Nursing, 29(6), 1018–1034. https://doi.org/10.1111/inm 666–674. https://doi.org/10.1016/j.sleh.2021.09.007
.12790 Wenzel, M., Blanke, E. S., Rowland, Z., & Kubiak, T. (2021). Emotion
Valentiner, D. P., Holahan, C. J., & Moos, R. H. (1994). Social support, regulation dynamics in daily life: Adaptive strategy use may be vari-
appraisals of event controllability, and coping: An integrative model. able without being unstable and predictable without being autoregres-
Journal of Personality and Social Psychology, 66(6), 1094–1102. sive. Emotion. Advance online publication. https://doi.org/10.1037/
https://doi.org/10.1037/0022-3514.66.6.1094 emo0000967
Content may be shared at no cost, but any requests to reuse this content in part or whole must go through the American Psychological Association.

van der Stouwe, E. C. D., Groenewold, N. A., Bos, E. H., de Jonge, P., White, M. E., & Shih, J. H. (2012). A daily diary study of co-rumination,
Wichers, M., & Booij, S. H. (2019). How to assess negative affective stressful life events, and depressed mood in late adolescents. Journal of
reactivity to daily life stress in depressed and nondepressed individuals? Clinical Child and Adolescent Psychology, 41(5), 598–610. https://doi
Psychiatry Research, 279, 259–266. https://doi.org/10.1016/j.psychres .org/10.1080/15374416.2012.706518
.2019.03.040 Zaki, J., & Williams, W. C. (2013). Interpersonal emotion regulation. Emo-
Waller, J. M., Silk, J. S., Stone, L. B., & Dahl, R. E. (2014). Co-rumination tion, 13(5), 803–810. https://doi.org/10.1037/a0033839
and co-problem solving in the daily lives of adolescents with major Zelic, K. J., Ciesla, J. A., Dickson, K. S., Hruska, L. C., & Ciesla, S. N.
This document is copyrighted by the American Psychological Association or one of its allied publishers.

depressive disorder. Journal of the American Academy of Child & Ado- (2017). An experimental investigation of co-rumination, problem solv-
lescent Psychiatry, 53(8), 869–878. https://doi.org/10.1016/j.jaac.2014 ing, and distraction. Behavior Therapy, 48(3), 403–412. https://doi.org/
.05.004 10.1016/j.beth.2016.11.013
Watson, D., Clark, L. A., & Tellegen, A. (1988). Development and valida-
tion of brief measures of positive and negative affect: The PANAS scales. Received July 20, 2021
Journal of Personality and Social Psychology, 54(6), 1063–1070. https:// Revision received April 27, 2022
doi.org/10.1037/0022-3514.54.6.1063 Accepted May 12, 2022 n

You might also like