Content uploaded by Michelle K Duffy
Author content
All content in this area was uploaded by Michelle K Duffy on Feb 15, 2023
Content may be subject to copyright.
RESEARCH ARTICLE
The balance between positive and negative affect in employee
well-being
David J. Yoon
1
| Joyce E. Bono
2
| Tao Yang
3
| KiYoung Lee
4
|
Theresa M. Glomb
5
| Michelle K. Duffy
5
1
Department of Management and Marketing,
Franklin P. Perdue School of Business,
Salisbury University, Salisbury, Maryland, USA
2
Department of Management, Warrington
College of Business Administration, University
of Florida, Gainesville, Florida, USA
3
Department of Management, Cameron
School of Business, University of North
Carolina Wilmington, Wilmington, North
Carolina, USA
4
Department of Management, School of
Business, Yonsei University, Seoul,
South Korea
5
Department of Work and Organizations,
Carlson School of Management, University of
Minnesota, Minneapolis, Minnesota, USA
Correspondence
David J. Yoon, Department of Management
and Marketing, Franklin P. Perdue School of
Business, Salisbury University, 1101 Camden
Avenue, Salisbury, MD 21801, USA.
Email: djyoon@salisbury.edu
Funding information
BK21 FOUR (Fostering Outstanding
Universities for Research), Grant/Award
Number: 336/19; Yonsei University's 2020-2
Future-Leading Research Initiative, Grant/
Award Number: 2020-22-0493; Signature
Research Cluster Program of 2021, Grant/
Award Number: 2021-22-0006
Summary
We examine the effects of the balance between positive and negative affect experi-
enced at work on well-being outcomes. An extensive literature on affect balance sug-
gests that it is not only positive affect (PA) and negative affect (NA) alone that affect
well-being; rather it is the balance between them that matters. We use experience
sampling methods and polynomial regression to test the notion that daily PA and NA
at work, along with their interactive and nonlinear effects, predict employee well-
being after work. In a sample of working adults, we find that affect balance—the
dynamic interplay between daily PA and NA—at work was differentially associated
with various indices of well-being: PA, NA, and the interaction between them
predicted physical and mental health. Affect balance at work also predicted life satis-
faction, but only for those low on trait affect balance. Detailed examination of the
joint effects of PA, NA, and the balance between them reveals that high PA at work
is most important for life satisfaction, whereas both low NA and high PA are impor-
tant for health. Low NA plays an especially important role in physical health.
KEYWORDS
affect, affect balance, health, life satisfaction, well-being
1|INTRODUCTION
Some of you say, “Joy is greater than sorrow,”and
others say, “Nay, sorrow is the greater.”But I say unto
you, they are inseparable. Together they come, and
when one sits alone with you at your board, remember
that the other is asleep upon your bed.—Khalil Gibran,
On Joy and Sorrow (Gibran, 1923/2001, p. 30)
Affect is important for work. Positive affect (PA) is beneficial for,
and negative affect (NA) is detrimental to, a wide variety of work
attitudes including job satisfaction, organizational commitment, and
turnover intentions (Thoresen et al., 2003). PA and NA also influence
work behaviors. PA is positively associated with task performance and
organizational citizenship behavior (OCB), whereas NA is negatively
associated with OCB and positively associated with counterproduc-
tive work behaviors (CWB), withdrawal behavior, and occupational
injury (Thoresen et al., 2003). Affective experiences are central to
employee well-being. A large body of research suggests that PA
improves and NA worsens healthy functioning of employees across
domains, including work life, social relationships, and mental
and physical health (e.g., Bono et al., 2013; Elfenbein, 2007;
Received: 22 June 2020 Revised: 31 October 2021 Accepted: 2 November 2021
DOI: 10.1002/job.2580
J Organ Behav. 2022;43:763–782. wileyonlinelibrary.com/journal/job © 2021 John Wiley & Sons, Ltd. 763
Hershcovis et al., 2007; Kaplan et al., 2009; Krantz & McCeney, 2002;
Lyubomirsky et al., 2005; Pressman & Cohen, 2005; Thoresen
et al., 2003).
Although the preponderance of research focuses on the indepen-
dent effects of either PA or NA, some research considers them in tan-
dem. Research on affect balance (Bradburn, 1969), positivity ratio
(Fredrickson & Losada, 2005), and emotional ambivalence (Fong &
Tiedens, 2002) considers the coexistence and combination of PA and
NA to better understand well-being (e.g., Bradburn, 1969; Schwartz
et al., 2002). Despite different labels and operationalizations, these
studies are similar in emphasizing that some combination of PA and
NA predict outcomes, presumably better than PA and NA alone,
though this assumption is not always tested explicitly.
There are three challenges in the existing literature on affect
balance. First, there is surprisingly little research on affect balance at
work (for an exception, see Fong & Tiedens, 2002). Although
workplace experiences impact the ebb and flow of PA and NA and
many studies link PA and NA to work outcomes (Connolly &
Viswesvaran, 2000; Kaplan et al., 2009), most research examining the
combined influence of PA and NA (i.e., affect balance) is in a nonwork
context (see Tov & Lee, 2015). This lack of attention to the work envi-
ronment renders an unclear picture of the extent to which the balance
of daily PA and NA at work is related to employee well-being.
Examination of affect balance at work could provide novel insights
into the role of affective experiences at work and their influence on
well-being.
Second, there is a paucity of compelling theory regarding how
and why the balance between PA and NA relates to well-being,
above and beyond main effects. There is a growing body of work on
affect balance, but much of it is grounded empirically
without a strong conceptual or theoretical base (Bradburn, 1969;
Derogatis, 1975; Fredrickson & Losada, 2005; Koydemir &
Schütz, 2012; Sanjuán, 2011). On the one hand, there are several
theoretical frameworks that explain the role that PA and NA
plays on well-being (e.g., affective events theory; Weiss &
Cropanzano, 1996), but these theories are built on the assumption
that PA and NA influence well-being independently. On the other
hand, research on affect balance typically considers only the balance
between PA and NA, ignoring their main effects and leaving open
the question of whether balance matters for well-being, once their
main effects are considered. Looking only at PA, or only at NA, or
only at a composite formed from them may mask any unique effects
of the dynamic interplay between positive and negative affective
states at work. For instance, when affect balance is conceptualized
as a ratio of PA to NA, it is not clear whether high PA or low NA is
most important. Furthermore, experiencing low PA and low NA
would produce the same ratio (balance) as would experiencing high
PA and high NA, but the effect of these two cases on employee
well-being may differ. Having a better conceptual basis for examining
affect balance would aid our understanding of how PA and NA
mutually influence each other throughout the workday and help
determine which combination of PA and NA levels the employee
should strive toward to maximize well-being. Better theory, however,
must also be accompanied by better methodology for testing main
and combined effects more precisely.
Third, it is unclear what role trait PA and NA, or the balance
between them, play in the experience of daily affect balance and its
outcomes. Trait affect is a stable and predictable individual difference
that serves as an “affective lens”through which individuals view and
respond to situations (Barsade & Gibson, 2007). Such traits are impor-
tant, in part, because they affect susceptibility to momentary experi-
ences of PA and NA (Larsen & Ketelaar, 1991). Momentary
experiences of PA and NA vary within individuals, but the degree of
variability may differ across individuals (Beal et al., 2005). Traits are
also important because there may be stable individual differences in
how levels of PA and NA are combined; people may vary in trait affect
balance, that is, the stable ratio of PA to NA. Thus, trait PA and NA
along with trait affect balance may influence both the experience of
momentary affect balance at work and its association with employee
well-being (see Ito & Cacioppo, 2005).
In this study, we aim to address each of these challenges. First,
we investigate the effects of affect balance at work on well-being
after work. We extend affect balance research by focusing on
employee well-being broadly construed (Ryan & Deci, 2001), including
employees' life satisfaction (Diener & Lucas, 1999) and their mental
and physical health (Ryff & Singer, 1998,2000), as these are the well-
being outcomes most proximal to, and thus most likely to be directly
influenced by, affective states at work (Pressman & Cohen, 2005;
Taylor, 1991). Mental and physical health are often used to assess
well-being in the social psychological and medical literature (Diener
et al., 1999; Ryff, 1989), but they receive less attention in organiza-
tional research even though—as healthcare costs continue to rise—
employee health is highly pertinent to work organizations.
Second, we advance affect balance research theoretically and
empirically. We integrate conservation of resources (COR;
Hobfoll, 1989) and broaden-and-build theories (Fredrickson, 1998,
2001) to explain why the interplay between daily PA and NA will dif-
ferentially predict well-being criteria. Furthermore, by empirically
mapping affective contours in a way that permits independence and
nonlinearity of PA and NA, we gain a better understanding of the vari-
ous theory-based strategies for improving well-being—by reducing
NA, increasing PA, or doing both (Fredrickson, 1998; Larsen &
Prizmic, 2008). Our use of polynomial regression and response surface
methods (Edwards, 2002) allows for a finer grained examination of
the unique roles of PA, NA, and affect balance in employee well-being,
thus allowing us to also extend theory both on affect balance and on
practical techniques for affect management.
Third, we integrate two theoretical perspectives—positivity offset
(Ito & Cacioppo, 2005) and negativity bias (Baumeister et al., 2001)—
to better understand the role of trait level affect (PA, NA, and affect
balance). We posit that trait affect places boundary conditions, such
that both the experience of daily affect balance at work and its associ-
ation with well-being depend on trait levels of affect balance. We
examine trait affect balance as a moderator of effects of daily affect
balance on employee well-being. By considering both trait and
momentary workplace affect balance, we advance understanding of
764 YOON ET AL.
the role that affective disposition plays in how individuals react to
daily affective states at work.
To examine the unique and interactive effects of PA, NA, and
affect balance on well-being—at both the daily and trait levels—we
used experience sampling data collected during a multi-phase study of
employee well-being. By collecting both momentary PA and NA at
work, and employees' trait levels of affect balance, we are able to
explore how daily affect balance at work spills over to influence
employee well-being in the evening, and how stable, trait level affect
balance places boundary conditions on those effects.
2|THEORY AND HYPOTHESES
2.1 |Daily affect balance and well-being
According to the Centers for Disease Control and Prevention (CDC) in
the United States, overall well-being is defined in its most simple
essence as “judging life positively and feeling good”(CDC, 2018).
Although it involves various elements, the three key indicators of
well-being include life satisfaction, emotional well-being, and physical
well-being (CDC, 2018). This is consistent with the definition of well-
being in the management literature that considers well-being as a
broader and more encompassing construct (Danna & Griffin, 1999).
According to Danna and Griffin (1999), well-being comprises
the “various life/non-work satisfactions enjoyed by individuals
(i.e., satisfaction and/or dissatisfaction with social life, family life, rec-
reation, spirituality, and so forth), work/job-related satisfactions
(i.e., satisfaction and/or dissatisfaction with pay, promotion opportuni-
ties, the job itself, co-workers, and so forth), and general health”
(p. 359). Moreover, general health comprises the “combination of such
mental/psychological indicators as affect, frustration, and anxiety and
such physical/physiological indicators as blood pressure, heart condi-
tion, and general physical health”(Danna & Griffin, 1999, p. 359). This
is also consistent with how Diener et al. (1999) conceptualized well-
being, which is a broad category of phenomena that includes life satis-
faction as well as satisfaction derived from a domain in life including
one's health. Thus, to capture the concept of well-being in the work-
place in this study, we focused on three daily indicators: life satisfac-
tion, mental health, and physical health. How employees attain well-
being includes the presence of PA and the absence of NA (Diener
et al., 1999; Diener & Seligman, 2002). This view is also reflected in a
long tradition of research on affect balance that operationalized affect
balance as either an excess-based (more PA than NA) or a ratio-based
(proportion of PA to NA) measure of PA and NA in predicting well-
being (e.g., Diener et al., 2010; Fredrickson & Losada, 2005).
The preponderance of PA over NA daily is important for well-
being (Pierce et al., 2017; Tov & Lee, 2015). In contrast to early work
on affect balance, where it was operationalized mostly as a single
ratio-based index (e.g., PA/NA), recent work (e.g., Pierce et al., 2017;
Tov & Lee, 2015) has moved away from combined measures. Rather,
affect balance has been reconceptualized more simply as the joint
effects of PA and NA since these two “cannot be studied in isolation”
(Pressman & Cohen, 2005, p. 961), especially at the daily level. This
move also reflects the organizational research view that daily emo-
tions are dynamic (Gooty et al., 2009), wherein PA and NA each have
independent effects on well-being but are also interdependent with
each other. When considering the interactive and nonlinear effects of
daily PA and NA, affect balance can be thought of as affective differ-
entiation (Pierce et al., 2017; Tov & Lee, 2015). Specifically, affective
differentiation is the discrepancy between PA and NA, which can take
on various forms: high PA and concurrently low NA (PA ascendancy
or positive discrepancy) versus high NA and concurrently low PA
(NA ascendancy or negative discrepancy).
Theoretical and empirical evidence supports the notion that posi-
tive discrepancy (preponderance of daily PA over NA) will be a signifi-
cant predictor of employee well-being. This is because having more
PA than NA will provide psychological resources for employees
throughout the day, helping them deal with the negative effects of
stressors. COR (Hobfoll, 1989) and broaden-and-build theories
(Fredrickson, 1998,2001) support this prediction. COR theory takes a
psychological and motivational approach, positing that individuals
have limited psychological resources that can be depleted and
replenished throughout the day (Hobfoll, 1989). Ebbs and flows of
resources in the form of PA and NA occur in response to experiences
(see Hobfoll, 1989). Net resources that remain after considering the
sum of resource-building positive experiences and resource-depleting
negative experiences determine individuals' well-being (Bono
et al., 2013), with an excess of daily PA over daily NA being important
for well-being (Pierce et al., 2017; Tov & Lee, 2015). Employees expe-
rience both PA and NA as they occur and co-occur throughout the
day (e.g., Miner et al., 2005) and they are capable of experiencing
them simultaneously with various levels of intensity (Fong &
Tiedens, 2002). Thus, the daily experience of PA may serve as a
buffer, both reducing resource losses and replenishing losses that
occur as a result of NA. This reduction in resource depletion supports
daily well-being. Moreover, because COR theory suggests that people
are motivated to conserve resources, they are only willing to expend
PA to counteract NA, when they have enough excess PA available.
Increased PA also influences well-being by suppressing stress and
its accompanying pathogenic influences. According to Pressman and
Cohen's (2005) review, increased PA changes behavior, affects physi-
ological processes, and improves social ties in ways that both improve
well-being directly and that mitigate the detrimental effects of NA. PA
is associated with improved health practices (e.g., relaxing at the end
of the day), improved regulation of both stress hormones (i.e., cortisol)
and endogenous opioids (e.g., endorphin), and improved social rela-
tionships, all of which are in turn associated with improved immune
and cardiac functioning and with reduced disease. Fredrickson's (1998,
2001) broaden-and-build theory further supports the role of PA in
breaking the link between NA and stress. Individuals with high PA are
able to improve and build up social, psychological, and physical
resources that allow them to more effectively recover from negative
events and NA. Thus, especially in stressful situations, PA has adapta-
tional consequences as it keeps NA from spiraling out of control. PA is
also associated with effective coping mechanisms, such as positive
YOON ET AL.765
reappraisal, problem-focused coping, and infusing ordinary
events with positive meaning (Folkman & Moskowitz, 2000;
Fredrickson, 2001). These coping mechanisms help individuals sustain
PA even in the presence of high NA and even in chronic stress situa-
tions (Folkman & Moskowitz, 2000). Tugade and Fredrickson (2004)
found that reinforcing PA via self-rewards was an effective way of
overcoming the detrimental effects of negative events and NA. In
summary, the preponderance of daily PA over daily NA broadens
individuals' physiological, psychological, and cognitive resources,
allowing them to better avoid the deleterious effects of NA
(Fredrickson, 2001).
Both COR and broaden-and-build theories suggest that an
excess of PA over NA is critical to well-being, as PA both helps
fight off the depleting effects of NA and replenishes NA-related
resource losses when they occur. Thus, we predict that daily affect
balance, in the form of positive discrepancy (an excess of daily PA
over NA), matters to well-being, even after considering their main
effects.
Hypothesis 1. Controlling for the main effects of daily
PA and NA, an excess of daily PA over NA (positive dis-
crepancy) will positively relate to well-being in terms of
(a) life satisfaction, (b) mental health, and (c) physical
health.
Although we expect a positive discrepancy of PA will broadly
contribute to well-being—across its three key indicators—we also
theorize that the form of the discrepancy may differ in predicting
life satisfaction versus health. Specifically, we expect that a positive
discrepancy driven by high PA will influence life satisfaction,
whereas a positive discrepancy driven by both high PA and low
NA will influence health. We turn to these issues in more detail
below.
2.2 |How to improve daily life satisfaction
We posit that increasing PA is the most influential strategy for achiev-
ing daily life satisfaction because it helps employees (1) better cope
with stressors and (2) seek out and achieve goals in life, which are
important for life satisfaction. First, PA helps employees better inter-
pret and cope with stressors, which is important for life satisfaction.
According to COR theory (Hobfoll, 1989), when individuals face
stressful situations throughout their workday, their resources are dra-
ined as they deal with stressors. Unless resources are replenished, loss
of resources may compound, resulting in suboptimal functioning in
work and life (Hobfoll, 1989). To conserve their resources, individuals
employ two cognitive strategies: (1) shifting the focus of their atten-
tion away from the stressor (e.g., reinterpret the stress as a challenge)
or (2) reevaluating the value of resources that are lost (Hobfoll, 1989).
By employing either of these psychological strategies, individuals can
maintain the net positive resources needed for optimal functioning
(Hobfoll, 1989), and increasing PA is critical for the employment of
these strategies (Bono et al., 2013; Cohn et al., 2009; Tugade &
Fredrickson, 2004).
First, increased PA provides physiological resources to cope with
stressors. According to neuropsychological theory (Ashby et al., 1999),
mild PA is associated with increased dopamine, which in turn
increases cognitive flexibility and creative problem-solving. More spe-
cifically, PA and its resultant dopamine boost form the basis of explo-
ration, taking chances, growing, developing, and trying new things.
This cognitive flexibility allows individuals to explore different ways to
“invest resources in order to enrich their resource pool”(p. 520), help-
ing them build a deeper reservoir of resources and preparing them for
future events where resources may be depleted (Hobfoll, 1989).
Broaden-and-build theory (Fredrickson, 1998,2001) also focuses on
the role of increased PA in expanding individuals' thought–action rep-
ertoire, which leads to broad-minded coping marked by generativity
and behavioral flexibility (Fredrickson, 2001). Individuals with higher
PA are better able to think of different ways to deal with a stressor
and step back from the stressor and view the situation more objec-
tively (Fredrickson, 2001). Thus, PA helps employees conserve and
build resources, altering their cognitive strategies in a way that helps
them become more resilient (Cohn et al., 2009) and bounce back from
negative emotional experiences (Tugade & Fredrickson, 2004). This
sense of having adequate resources to face life challenges is an
important precursor to life satisfaction (Cohn et al., 2009; Gable
et al., 2004). Hence, PA provides resources to better cope and
creatively solve problems, which is critical for life satisfaction
(Lyubomirsky et al., 2005).
Second, PA helps employees seek out and achieve higher goals,
which are important for achieving success and life satisfaction. PA is
also associated with reward-seeking systems (Ashby et al., 1999;
Carver & White, 1994). Striving for and attaining important goals in
life is critical to both job and life satisfaction (Judge et al., 2005).
Individuals with high PA tend to display a stronger likelihood of
pressing toward new goals, proactively seeking and meeting those
goals (Fredrickson, 1998,2001), and have higher levels of goal
achievement (Lyubomirsky et al., 2005). Thus, it is success, via goal
achievement, in various life domains, including work performance,
social relationships, and finances, that informs perceptions of life
satisfaction; goal attainment and life satisfaction are linked (Judge
et al., 2005).
Given the role of PA in developing psychological resources
needed for resiliency and goal attainment, which in turn predict life
satisfaction, positive daily affective discrepancy is important for daily
life satisfaction. We acknowledge that a positive discrepancy can be
achieved both by increasing PA and by decreasing NA, but theory
tends to suggest that the most important path to life satisfaction is via
increased PA, due to both its psychological (e.g., pressing toward
goals, increased flexibility, creativity, and resilience) and physiological
(e.g., dopamine) effects.
Hypothesis 2a. The most influential strategy for
achieving high daily life satisfaction is from increased
daily PA.
766 YOON ET AL.
2.3 |How to improve daily health
There are contrasting views on whether increasing PA or decreasing
NA is the most influential strategy for improving mental and physical
health. On the one hand, increasing PA is associated with better
health. Much of the logic linking PA to life satisfaction is relevant here
as well. PA improves health both physiologically and socially. It moti-
vates people to engage in mentally and physically healthy life prac-
tices (Pressman & Cohen, 2005). PA also aids the body in regulation
of the nervous system, as well as in the production and regulation of
cortisol and opioids; it helps reduce inflammatory processes and is
associated with cardiac health. It increases dopamine levels that
improve performance on cognitive tasks needed for a healthy and
resilient lifestyle (Ashby et al., 1999). Moreover, the increased resil-
iency that results from PA (Cohn et al., 2009) is also associated with
reduced healthcare utilization, including treatment for stress-related
illnesses (Stahl et al., 2015). This finding is consistent with
Fredrickson's (1998,2001) broaden-and-build theory which further
expands the positive role that PA plays on health by noting the effect
that it has on undoing the effects of NA.
On the other hand, decreasing NA is also an effective strategy
for improving mental and physical health. There are two theories
that support this assertion. First is the theory of negativity bias. Neg-
ativity bias refers to a “tendency for the negative motivational sys-
tem to respond more intensely than the positive motivational system
to comparable increases in input”(Ito & Cacioppo, 2005, p. 2). In
other words, people respond more immediately and put forth more
effort in response to negative events compared to positive events
and are three times more influenced by them than they are by posi-
tive events (Larsen & Prizmic, 2008). As such, Larsen and
Prizmic (2008) noted that the greatest gain in cognitive and behav-
ioral outcomes results from either reducing NA or processing it bet-
ter. In this theory, the most influential strategy for achieving health
is via reduction of NA due to its outsized effects on physical and
mental health.
A second theory that links NA to reductions in health is
Taylor's (1991) mobilization–minimization hypothesis. In this theory,
like in negativity bias theory, negative events demand more atten-
tion and vigilance from individuals—often resulting in an increase in
NA—than positive events. This is a functional approach for humans
as it allows them to better respond to threatening situations. Nega-
tive events elicit more causal attributional activity than positive
events and require more laborious, complex reasoning (Peeters &
Czapinski, 1990). They also elicit negative moods which are psycho-
logically and physically taxing in themselves and which require effort
to recover from (Taylor, 1991). This line of theoretical work sug-
gests that the most influential strategy for improving mental and
physical health is reducing the NA that results from negative
events.
Consistent with theoretical predictions, empirical work also finds
that NA, such as anxiety and anger, predicts negative mental and
physical health in the form of an increase in physical complaints, stress
(Watson, 1988), cardiovascular disease (e.g., Barefoot et al., 2000;
Kubzansky & Kawachi, 2000), asthma (Friedman & Booth-
Kewley, 1987), diabetes (Carnethon et al., 2003; Lustman et al., 1991),
and hypertension (Everson et al., 1998; Jonas et al., 1997). Richman
et al. (2005) reasoned that the link between NA and poor health is
attributed to both a direct (e.g., chronic sympathetic nervous system
activation and serotonergic dysregulation) and indirect pathway of NA
on health (e.g., engaging in adverse health behaviors such as smoking
and excessive alcohol consumption and in having distorted symptom
perception).
Due to these conflicting views of which is more important for
health—increasing PA or reducing NA—scholars suggest that the most
effective strategy for improving physical and mental health is to do
both. Since negative events and coping with NA deplete resources,
Hobfoll (1989) suggests two ways to maintain optimal functioning:
(1) find ways to reduce NA (e.g., reappraise the value of the resources
lost) or (2) find ways to gain resources (e.g., via increased PA).
Broaden-and-build perspective (Fredrickson, 2003) expands this idea
in the realm of health. Specifically, Fredrickson's (2003)“undoing
hypothesis”suggests that NA elicits increases in cardiovascular activ-
ity, which are necessary for the body in preparing to deal with the sit-
uation at hand. Prolonged states of heightened cardiovascular activity
result in poorer mental (increased stress) and physical health (heart
disease; Fredrickson, 2003). But since increased PA results in a
broader solution-seeking mindset, it may “loosen the hold that nega-
tive emotions gain on both mind and body, dismantle preparation for
specific action and undo the physiological effects of negative emo-
tions”(Fredrickson, 2003, p. 334). In other words, increased PA will
be accompanied by decreased NA, which, together, will result in
improved health.
In a similar vein, Larsen and Prizmic (2008) discussed dual
approaches where both increasing PA and decreasing NA will predict
health. Larsen and Prizmic (2008) advocated for speeding up adapta-
tion to negative events by finding meaning in them, engaging in down-
ward comparisons, expressing NA outwardly, and avoiding events that
elicit negative emotions. But they also suggested slowing adaptation
to positive events by savoring them, counting blessings, and sharing
positive experiences with others. Further supporting theory that both
PA and NA are important to health is research linking PA to reward-
seeking and NA to avoidance (Carver & White, 1994); positive mental
and physical health requires both active health-enhancing behaviors
(exercise, friendship; Watson, 1988; Pressman & Cohen, 2005) and
the avoidance of health-depleting behaviors, especially those that
cause stress and inflammation (for a review, see Richman et al., 2005).
In light of complementing theories with regard to daily affect and its
role in improving mental and physical health, we expect a positive dis-
crepancy (PA being higher than NA) to be important for mental and
physical health, but do not expect either PA or NA to have a dispro-
portionate influence on either outcome at higher levels. Thus, we
hypothesize below:
Hypothesis 2b and 2c. The most influential strategy
for achieving daily (b) mental health and (c) physical
health is from both higher daily PA and lower daily NA.
YOON ET AL.767
2.4 |Who gains the most from daily affect
balance?
Although it has been clearly established that PA and NA are states
that vary across time, a key issue emerging in recent research is the
notion that people have stable, trait-like dispositional properties
which influence how they experience and interpret affect-laden
events. It is in the context of this literature that research on affective
traits such as affect spin (Beal & Ghandour, 2011) and dialectical
thinking—“a cognitive style characterized by acceptance of inconsis-
tencies”(Hideg & van Kleef, 2017, p. 1196)—have emerged. In the
affect balance literature, there has been consideration of both daily
(Tov & Lee, 2015) and trait (Pierce et al., 2017) levels of affect and
their associations with well-being (Pierce et al., 2017), but these two
lines of inquiry have not been fully integrated. Previous work suggests
that people have stable affective baselines that shape how daily expe-
riences of affective differentiation and balance relate to well-being.
We posit that the ratio of stable PA and NA baselines can be concep-
tualized as trait affect balance and that differences between individ-
uals in stable tendencies to experience affect balance may exist. Two
theories, in conjunction, explain the source of trait affect balance.
2.4.1 | Positivity offset
According to Ito and Cacioppo (2005), positivity offset refers to the
“tendency for the positive motivation system to respond more than
the negative motivational system to comparably low levels of evalua-
tive input”(p. 2). Put differently, positivity offset refers to the notion
that individuals' affective states are not neutral; they are slightly
toward the positive when at rest (Larsen, 2009). This slightly positive
state is the point to which individuals return once they adapt to either
positive or negative events over time (hedonic treadmill theory;
Brickman & Campbell, 1971). Positivity offset provides a “higher inter-
cept value for the positive activation function”(Ito & Cacioppo, 2005,
p. 2) and results in an asymmetry in interpreting and responding to
neutral stimuli. The important part of this theory for our research is
that individuals differ in the extent to which they exhibit this positivity
offset (Ito & Cacioppo, 2005). Individuals who have a stronger positiv-
ity offset interpret and respond with more PA to neutral stimuli than
those with a weaker positivity offset (Ito & Cacioppo, 2005). Follow-
ing a negative event, positivity offset provides adaptive benefits as
individuals may experience more PA from neutral stimuli in their envi-
ronment (e.g., a song on the radio; Ito & Cacioppo, 2005).
2.4.2 | Negativity bias
Negativity bias refers to individuals' tendency to pay more attention
and be more affected by negative events emotions (Ito &
Cacioppo, 2005). Despite the human tendency to be in a slightly posi-
tive stable state (i.e., positivity offset), there is a second human ten-
dency to react more to negative events. This negativity bias results in
three types of asymmetries between PA and NA (Larsen, 2009). First,
individuals react more strongly in NA to negative events than they do
in PA to positive events of equivalent levels (reactivity asymmetry;
Larsen, 2002). Second, they adapt to NA more slowly than they do to
PA (duration asymmetry; Larsen, 2002). When individuals experience
events, they initially undergo strong and rapid emotional, physiologi-
cal, cognitive, and social responses. This mobilization is soon followed
by emotional, physiological, cognitive, and social responses to mini-
mize the impact of this event, and adaptation to positive events is typ-
ically faster than to negative events (Taylor, 1991). Third, they are
more strongly influenced by negative stimuli than they are by positive
stimuli (Larsen, 2009). Most important for our research is evidence
that people differ in their reactivity to negative events (Ito &
Cacioppo, 2005). Previous studies have found differences in negativ-
ity bias between individuals (Norris et al., 2011).
2.4.3 | Trait affect balance
In concert, stable individual differences between people in positivity
offset and negativity bias suggest individual differences in stable trait
levels of affect balance. Trait affect balance, then, refers to the
strength of an individual's positivity offset tendencies in comparison
to the strength of their negativity bias. As such, trait affect balance
represents the amount of affective resources (stable levels of excess
PA over NA) that an individual has at their disposal. Individuals with
high trait affect balance have a deeper reservoir of affective
resources, providing them with stabilizing force against daily fluctua-
tions in affect, and affect balance. Thus, we hypothesize that the
effect of positive discrepancy to be stronger for employees with lower
trait affect balance:
Hypothesis 3. The positive discrepancy between daily
PA and NA will more strongly predict daily (a) life satis-
faction, (b) mental health, and (c) physical health for
employees with low trait affect balance, as compared to
those with high trait affect balance.
3|METHOD
3.1 |Participants and procedures
Data for this study were drawn from archival data collected by two of
the authors as part of an ongoing, three-phase, multisite, multiyear
project focused on employee well-being. As each phase of this
research was completed, de-identified data were entered into a repos-
itory from which these data were drawn.
1
In all phases of the
research, participants completed a background survey and a series of
daily surveys using experience sampling methodology. The data
reported here are drawn from the first two phases of the project, both
of which included measurement of daily mood at work and some
aspect of daily well-being in the evening (life satisfaction measured in
768 YOON ET AL.
Phase I, mental and physical health measured in both Phase I and
Phase II). In Phase I, participants were drawn from a leadership devel-
opment program sponsored by a large public university, in which one
of the authors was involved. Managers from the local community
either signed up for the leadership development program voluntarily
in response to an advertisement or were sent by their organizations.
The program involved a 360-feedback process wherein managers
were asked to get feedback from coworkers, including supervisors,
peers, and direct reports. Participants in this research were not man-
agers enrolled in the development program. Instead, they were drawn
from the pool of individuals who reported on managers' behavior for
the feedback report. Our aim was to recruit workers in a broad variety
of jobs and industries; the leadership development program served as
our point of access. Thus, our Phase I participants were secondary to
the leadership development program and had no special connection
to the university where it was offered. Along with their survey for the
development program, they received an invitation to participate in a
“research study focused on links between employees' work experi-
ences and their experiences at home, after work.”The final sample in
Phase I (after excluding those who participated for the first day and
stopped and a few who were not able to access daily surveys) was
63 individuals, 76.2% female.
Phase II participants were drawn from employees of a large
ambulatory healthcare organization with multiple regional offices. This
research was designed in conjunction with organizational executives,
who were interested in improving employee well-being and reducing
stress. Employees were recruited directly in lunchrooms in their office
locations, as well as via email. They were invited to participate in a
research study “focused on whether work events influence employee
health.”This project involved a baseline data collection period of
7 days (from which these data were drawn) and an 8-day intervention
period with data not used here. Participants in Phase II worked in a
broad variety of healthcare jobs, but they all worked from the same
organization, in multiple offices. The final sample in Phase II (after
excluding those who participated for the first day and stopped and
those who enrolled but had scheduling difficulty) was 62; all except
one were female as is typical for these job types.
Procedures varied slightly across the phases. All background sur-
veys used to collect demographic and trait information were com-
pleted online, prior to the start of the daily survey period (trait PA
and NA were assessed 1–3 weeks prior to the daily measurement
period). In Phase I, participants received surveys by email three times
each day (morning and afternoon surveys to measure mood, evening
survey to measure well-being) for 10 consecutive workdays. In Phase
II, handheld devices were used for data collection during the work-
day (morning and afternoon surveys to measure mood) for seven
consecutive workdays. Evening well-being surveys were adminis-
tered via phone by research assistants. For participants in both
phases, daily surveys were restricted to just a few minutes. Compli-
ance was monitored and payment was prorated based on survey
completion.
All participants were employed at least 32 h a week. Across both
phases, 187 employees expressed interest in participation. Of those,
157 completed the background survey, but 24 dropped during the
first day of the daily surveys, and another eight were eliminated by
the researchers as they were on leave during the daily survey period.
Thus, our final sample comprised 125 participants (63 from Phase I
and 62 from Phase II) and 828 daily observations. Participants held
jobs such as accounting specialist, administrative assistant, lab techni-
cian, nurse, marketing coordinator, and receptionist, with healthcare
jobs overrepresented. Most were female (87.2%), consistent with the
demographics of these types of jobs. On average, they were
35.34 years old (SD =10.50) and earned an estimated hourly wage of
$19.35 (SD =$6.87).
3.2 |Measures
3.2.1 | Daily positive and negative affect
Because repeated surveys had to be kept short, a smaller set of
items was used to cover the full range of affect states. Previous
research has used brief measures for state PA and NA (e.g., Beal
et al., 2013; Bono et al., 2007; Lanaj et al., 2021; Lennard
et al., 2019). Daily PA and NA were measured with four items each,
drawn from Feldman Barrett and Russell's (1998) circumplex model,
to represent high and low levels of activation for PA and NA. The
items were “happy,”“excited,”“contented,”and “relaxed”for PA
(α=.81) and “nervous,”“angry,”“sad,”and “fatigued”for NA
(α=.75): 1 =strongly disagree to 5 =strongly agree. Participants
were asked to report on their momentary moods at work with
slightly different question stems across the phases. In Phase I, par-
ticipants were asked to “describe your general mood at work this
morning (afternoon).”In Phase II, participants indicated “how you
are feeling right now”in the morning and afternoon surveys. We
computed daily PA and NA by averaging the PA and NA affect
scores for each day.
3.2.2 | Daily well-being
Following the literature and consistent with Danna and Griffin's (1999)
definitions, we measured three aspects of daily well-being. In Phase I
only, we measured life satisfaction in the evening survey using the
item “Please rate how satisfied you feel with your life right now”
(1 =very unsatisfied to 5 =very satisfied), adapted from Diener
et al. (1985). In both phases, we measured mental and physical health
complaints in the evening survey based on Goldberg (1972): “To what
extent did you experience the following symptoms since you left
work”(Phase I via email), and “Since you left work today, to what
extent did you experience the following”(Phase II via evening phone
survey). The items were “difficulty concentrating,”“difficulty making
decisions,”and “felt stressed”for mental health complaints, and
“upset stomach,”“neck or back pain,”“headaches,”and “painful or
tense muscles”for physical health complaints. Response options were
not at all (coded 1), slight,moderate,a great deal, and severely (coded 5).
YOON ET AL.769
We reverse-scored the items and averaged them to form a score for
daily mental and physical health, respectively; high scores indicated
better health.
3.2.3 | Trait affect balance
Trait PA and NA were measured using the 20-item Positive and Nega-
tive Affect Schedule (PANAS; Watson et al., 1988), which includes
10 items each for positive (e.g., “enthusiastic”and “active”) and nega-
tive (e.g., “upset”and “distressed”) affect. Participants indicated the
extent to which they “generally feel this way”on a 5-point scale
(1 =very slightly or not at all to 5 =very much). Scores across items
were averaged to form a single score for trait PA (α=.89) and NA
(α=.79). Trait affect balance was computed as trait PA divided by
trait NA.
3.3 |Analytical strategies
3.3.1 | Data structure
Our data represent days nested within individuals, with 125 partici-
pants and 828 daily observations. To test Hypotheses 1and 2 at the
daily level (Level 1), we person mean-centered the data for daily
affect, such that daily values for PA and NA represent deviations
(increases or decreases) from each person's mean. This allows us to
examine the extent to which one unit of change in daily PA or NA rel-
ative to a person's mean results in an increase or decrease in life satis-
faction or health, for that person. Hypothesis 3is a cross-level
moderation hypothesis with trait affect balance at Level 2 (grand
mean-centered) moderating the association between daily affect bal-
ance and well-being at Level 1. We used Mplus 7.11 (Muthén &
Muthén, 2012) with robust maximum likelihood estimation to account
for nonindependence and non-normality of observations (Muthén &
Muthén, 2012).
3.3.2 | Daily affect balance and well-being
In examining the effects of daily affect balance on well-being, we used
polynomial regression (Cacioppo & Berntson, 1994; Edwards, 2002).
This approach allows for examination of the unique effects of PA and
NA as well as their interplay; it recognizes the independent and inter-
related nature of PA and NA and examines both linear and nonlinear
associations between affect and well-being. This operationalization of
affect balance allows us to sort out main effects of daily PA and NA
from the interactions between them.
We used a two-step procedure. In Step 1, we entered daily PA
and NA; in Step 2, we entered a block of nonlinear and interaction
terms (PA
2
,NA
2
, and PA NA). Unique variance in well-being is
attributed to daily affect balance if the incremental variance (ΔR
2
within
) in Step 2 is significant (Edwards, 2002). We also plotted
response surfaces to interpret and display the joint role of daily PA
and NA. We reported R
2within
as the proportion of within-level vari-
ance explained by predictors (e.g., Hofmann et al., 2000; Kreft & de
Leeuw, 1998) and used the Satorra–Bentler scaled chi-square differ-
ence (TRd; Satorra & Bentler, 2010) to determine the significance of
ΔR
2within
.
3.3.3 | Cross-level moderation
Our goal in the cross-level analyses is to test the notion that daily
affect balance may differentially influence well-being for people
who are high or low on trait affect balance. We conducted a series
of moderated multilevel polynomial regressions, extending Edwards
and Rothbard's (1999) principle of moderated polynomial regression
to the multilevel context. In Step 1, we allowed random slopes of
within-person predictors (i.e., PA, NA, PA
2
,PANA, and NA
2
) and
regressed random slopes of PA and NA, as well as the random
intercept, on trait affect balance. In Step 2, we regressed random
slopes of PA
2
,PANA, and NA
2
on trait affect balance. Significant
incremental variance in Step 2 indicates that trait affect balance
moderated the relationship between daily affect balance and well-
being. The equations for the cross-level analyses are noted in
Appendix A.
We reported R
2total
as the proportion of total variance explained
by predictors (Snijders & Bosker, 1999). We tested the significance of
the incremental variance in Step 2 (ΔR
2total
) using the Satorra–Bentler
scaled chi-square difference (TRd; Satorra & Bentler, 2010). When sig-
nificant moderating effects were found, we plotted response surfaces
at high and low levels (1 SD above and below the mean) of trait affect
balance.
4|RESULTS
4.1 |Confirmatory factor analyses
We performed confirmatory factor analyses (CFA) to examine the dis-
tinctiveness of daily measures using clustered CFA to account for the
nested data structure. In Phase I, we ran a five-factor model with daily
PA, NA, life satisfaction, mental health, and physical health. For the
single-item life satisfaction, we set its factor loading to one and error
variance to zero (for an example, see Bono et al., 2017). The five-
factor model had an acceptable fit (CFI =.91, RMSEA =.07,
SRMR =.06) and showed a better fit than three alternative models
that combine (1) PA and NA (CFI =.86, RMSEA =.08, SRMR =.07;
Satorra & Bentler's, 2010 scaled Δχ
2
=60.53, Δdf =4, p< .01),
(2) mental and physical health (CFI =.87, RMSEA =.08, SRMR =.06;
scaled Δχ
2
=14,319.14, Δdf =4, p< .01), and (3) NA and mental
health (CFI =.83, RMSEA =.09, SRMR =.07; scaled Δχ
2
=116.65,
Δdf =4, p< .01). In Phase II, a four-factor model with daily PA, NA,
mental health, and physical health (life satisfaction was not measured
in this phase) yielded an acceptable fit (CFI =.90, RMSEA =.06,
770 YOON ET AL.
SRMR =.06) and showed a better fit than three alternative models
that combine (1) PA and NA (CFI =.75, RMSEA =.09, SRMR =.09;
scaled Δχ
2
=251.47, Δdf =3, p< .01), (2) mental and physical health
(CFI =.85, RMSEA =.07, SRMR =.06; scaled Δχ
2
=88.13,
Δdf =3), and (3) NA and mental health (CFI =.67, RMSEA =.11,
SRMR =.11; scaled Δχ
2
=55.93, Δdf =3).
2
Overall, results support
the distinctiveness of the daily measures.
4.2 |Daily affect balance and well-being
We report the zero-order correlations between daily PA, NA, and
evening well-being, as well as trait affect balance, in Table 1.As
expected, the within-person correlations between our three well-
being variables are small to moderate (ranging from .22 to .32,
ps < .01), providing support for our attempt to capture distinct
aspects of well-being.
Hypotheses 1and 2 address a set of interrelated questions.
Hypothesis 1examines whether affect balance (i.e., excess of daily PA
over NA) matters for well-being, once we control for the main effects
of PA and NA. Hypothesis 2 seeks to determine the most influential
strategy for achieving well-being: increased PA, decreased NA, or
both. Two conditions are needed to support Hypothesis 1. First, the
block of polynomial terms (PA
2
,PANA, NA
2
) must explain signifi-
cant incremental variance in well-being, indicating that the interplay
between PA and NA adds to the prediction of well-being, even after
controlling for PA and NA. If this block is not significant, then we can-
not proceed with testing Hypotheses 1and 2 by examining the
response surface (Edwards, 2002). Second, provided that the first con-
dition is met, Hypothesis 1would be supported if the slope along the
PA =–NA line was significant and positive. Support for Hypothesis 1
is needed as a condition to proceed with testing Hypothesis
2. Hypothesis 2a would be supported if PA
2
was significant in
predicting life satisfaction. Hypotheses 2c and 2c would be supported
if the curvature along the PA =–NA line was significant and positive
for mental and physical health.
4.2.1 | Life satisfaction
Results of our polynomial regression for life satisfaction in Table 2
reveal a nonsignificant incremental effect for the block of polynomial
terms (ΔR
2
=.03, p=.12). Thus, we do not find support for Hypothe-
sis 1a; daily affect balance does not predict daily life satisfaction, even
though the pattern of results is consistent with our expectation—that
excess PA over NA matters for well-being—as the slope of the line
where PA =–NA, each expressed as a deviation from the person's
mean, is positive and significant (b
1
b
2
=.28, p< .01; Table 2). Thus,
Hypotheses 1a and 2a are not supported.
4.2.2 | Health
In contrast to our results for life satisfaction, the block of polynomial
terms for health (Table 2; Step 2) explained additional unique variance
for both mental (ΔR
2
=.01, p< .05) and physical health (ΔR
2
=.02,
p< .05), allowing for further testing of Hypotheses 1b and 1c and
Hypotheses 2c and 2c.
For mental health, we observed a significant slope (b
1
b
2
=.31,
p< .01) along the PA =–NA line where deviations of PA and NA from
a person's mean are in opposite directions. Figure 1shows that opti-
mal daily mental health occurs when PA is higher and NA is lower
than a person's mean (represented by the area in the upper right cor-
ner of the response surface). Thus, Hypothesis 1b is supported; even
after controlling for the main effects of PA and NA, mental health
improved on days when there was greater excess of PA over
NA. Moreover, we find a significant positive curvature (b
3
b
4
+b
5
=.10, p< .05; Table 2) along the PA =–NA line, indicating better
mental health as PA reaches its highest level while NA reaches its low-
est. Thus, we find support for Hypothesis 2b; the most influential
strategy for employees to achieve daily mental health is to both
increase PA and decrease NA.
For physical health, there is a significant slope (b
1
b
2
=.16,
p< .01) along the PA =–NA line where deviations of PA and NA from
TABLE 1 Descriptive statistics and zero-order correlations
MSD12345678
1. Trait PA 3.72 0.63
2. Trait NA 1.71 0.44 .33
**
3. Trait affect balance 2.36 0.86 .71
**
.84
**
4. Daily PA 3.26 0.62 .37
**
.28
**
.40
**
.61
**
.17
**
.24
**
.12
**
5. Daily NA 2.19 0.67 .28
**
.43
**
.44
**
.60
**
.18
**
.22
**
.15
**
6. Daily life satisfaction 3.74 0.96 .30
*
.33
**
.31
*
.66
**
.55
**
.26
**
.22
**
7. Daily mental health 4.52 0.65 .24
**
.43
**
.40
**
.44
**
.49
**
.51
**
.32
**
8. Daily physical health 4.56 0.58 .26
**
.34
**
.33
**
.42
**
.42
**
.57
**
.77
**
Notes: Between-person correlations are below the diagonal, and within-person correlations are above the diagonal. Trait affect balance is the ratio of trait
PA over trait NA. For correlations involving daily life satisfaction, Level 1 N=439, Level 2 N=63. For correlations involving daily mental and physical
health, Level 1 N=828, Level 2 N=125.
*
p< .05.
**
p< .01 (two-tailed tests).
YOON ET AL.771
a person's mean are in opposite directions. Figure 2shows that—
similar to our results for mental health—the optimal point of physical
health occurs when a person experiences greater PA and less NA than
the person's normal levels. Thus, Hypothesis 1c is supported. More-
over, we find a significant positive curvature (b
3
b
4
+b
5
=.07,
p< .01; Table 2) along the PA =–NA line, suggesting better physical
health as PA reaches its highest level while NA reaches its lowest.
Thus, Hypothesis 2c is supported; the most influential strategy for
employees to achieve daily physical health is to both increase PA and
decrease NA.
The significant positive curvatures along the PA =–NA line
also suggest that it is better for individuals' mental and physical
FIGURE 1 Response surface
for daytime affect balance
predicting employee mental
health
TABLE 2 Daytime affect balance and evening employee well-being: A polynomial approach
Life satisfaction Mental health Physical health
Step 1 Step 2 Step 1 Step 2 Step 1 Step 2
Within-level effects
PA .11 .13 .15
**
.16
**
.03 .04
NA .16 .15 .13
*
.15
**
.11
*
.13
**
PA
2
.31
*
.11 .14
PA NA .75 .21 .34
NA
2
.49 .01 .14
R
2within
.04
**
.07
**
.07
**
.08
**
.02
**
.04
**
ΔR
2within
.03 .01
*
.02
*
TRd (df ) 10.80 (2) 5.77 (3) 39.09 (2) 9.65 (3) 16.23 (2) 7.92 (3)
Shape along PA =NA line
Slope (b
1
+b
2
).01 .01 .09
Curvature (b
3
+b
4
+b
5
) 1.56 .32 .62
*
Shape along PA =–NA line
Slope (b
1
b
2
) .28
**
.31
**
.16
**
Curvature (b
3
b
4
+b
5
) .05 .10
*
.07
**
Notes: Estimates are unstandardized coefficients. For life satisfaction, Level 1 N=439, Level 2 N=63. For mental and physical health, Level 1 N=828,
Level 2 N=125. PA and NA were person mean-centered before computing squared and interaction terms. Within-level residual variance and between-
level mean and variance were estimated but were omitted from the table for brevity. Satorra–Bentler scaled chi-square statistic (TRd; Satorra &
Bentler, 2010) indicates significance of incremental variance (ΔR
2within
) explained in Step 2.
*p< .05. **p< .01 (two-tailed tests).
772 YOON ET AL.
health if they have dramatically higher levels of NA than their
levels of PA compared to if they have equally moderate levels of
NA and PA. On the basis of Taylor's (1991) theory, having very
high levels of NA paired with very low levels of PA may send a
stronger alarm signal to individuals to minimize the effects of NA,
which may entail thinking of positive events and not ruminating on
NA-inducing events. These minimization tactics may be more
urgently and sharply triggered when individuals experience very
high levels of NA and very low levels of PA concurrently, as this
makes NA more salient, leading to stronger mobilization (see
Taylor, 1991).
4.3 |Cross-level moderation
Hypothesis 3suggests that the effects of daily affect balance on daily
well-being differ based on levels of trait affect balance. To test this
hypothesis, we ran cross-level polynomial regressions with trait
PA/NA as the moderator (see Tables 3and 4) of the daily association
between affect balance and well-being.
For life satisfaction, we found a significant moderating effect of
trait affect balance (ΔR
2total
=.006, p< .05; Table 3). In Table 4,we
present the association between daily affect and well-being for those
1SD higher and lower than the mean on trait affect balance. We
FIGURE 2 Response surface
for daytime affect balance
predicting employee physical
health
TABLE 3 Daytime affect balance and evening employee well-being: A polynomial approach with moderating effects of trait affect balance
(PA/NA)
Life satisfaction Mental health Physical health
Step 1 Step 2 Step 1 Step 2 Step 1 Step 2
Between-level effects
Trait affect balance àrandom intercept .29 .29 .25
**
.26
*
.19
**
.19
**
Trait affect balance àrandom slope of PA .19
*
.17
**
.06 .07 .02 .01
Trait affect balance àrandom slope of NA .17 .12 .10 .09 .04 .03
Trait affect balance àrandom slope of PA
2
.07 .17 .09
Trait affect balance àrandom slope of PA NA .53 .23 .09
Trait affect balance àrandom slope of NA
2
.54
*
.10 .07
R
2total
.095
**
.101
**
.172
**
.174
*
.119
**
.119
ΔR
2total
.006
*
.002 .000
TRd (df ) 40.69(13) 9.05(3) 96.29(13) 0.25(3) 41.16(13) 0.45(3)
Notes: Estimates are unstandardized coefficients. For life satisfaction, Level 1 N=439, Level 2 N=63. For mental and physical health, Level 1 N=828,
Level 2 N=125. Between-level intercepts and residual variances and within-level residual variance are estimated but are omitted from the table for
brevity.
*p< .05. **p< .01 (two-tailed tests).
YOON ET AL.773
found a positive slope along the line of positive discrepancy (PA =–
NA) when trait affect balance was low (b
1
b
2
=.35, p< .01) but not
when it was high (b
1
b
2
=.27, ns). Response surface plots in
Figure 3show that for individuals who have low levels of trait affect
balance, an increase in positive discrepancy (PA higher than NA) is
associated with higher life satisfaction (Figure 3a), but for those who
have high levels of trait affect balance, an increase in positive discrep-
ancy on a daily basis is not related to life satisfaction (Figure 3b). Thus,
Hypothesis 3a is supported.
For mental and physical health, we did not find significant moderat-
ing effects of trait affect balance. Incremental variances explained by
the addition of the moderated polynomial terms were not significant
(mental health: ΔR
2total
=.002, p=.97; physical health: ΔR
2total
=.000,
p=.93; Table 3). Thus, Hypotheses 3b and 3c are not supported.
Considered as a whole, our results show that for life satisfaction,
but not for mental and physical health, those with high levels of trait
affect balance are less responsive to experiences of daily fluctuations
in affect balance. Hence, our results indicate that when it comes to
sustaining strong mental and physical health, traits play less of a role
and daily affective experiences matter for everyone.
4.4 |Post hoc analysis
We ran three sets of post hoc analyses.
3
First, our use of a ratio index
of trait affect balance (i.e., PA/NA) means that we cannot determine
whether it is trait PA alone, trait NA alone, or the interaction between
them that moderates the daily association between affect balance and
life satisfaction. When possible, the best operationalization of affect
balance is to include the polynomial terms (PA
2
,NA
2
, and PA NA),
after controlling for PA and NA, as we did at Level 1, testing Hypothe-
ses 1and 2. But incorporating the cross-level interactions between all
these terms makes interpretation impossible. Thus, we followed pre-
cedents of Diener and Suh (1999), who suggest that single indices,
such as PA/NA, can be useful when comparing people broadly. Yet,
taking this ratio approach to trait affect balance means that we cannot
know whether it is high PA or low NA that led to the moderating
effects we found for life satisfaction. Therefore, we ran supplemental
analyses examining the moderating effect of trait PA and trait NA
alone, in lieu of trait affect balance (PA/NA). Results reveal that nei-
ther trait PA nor trait NA alone was a significant moderator of the
daily association between affect balance and life satisfaction,
suggesting that the moderating effects found are not due to trait PA
or trait NA alone, but rather due to their joint, combined effect.
Second, given the associations of age (e.g., Steptoe et al., 2015),
gender (e.g., Fujita et al., 1991), and job tenure (e.g., Warr, 1992) with
aspects of well-being, we added age, gender, and job tenure as control
variables in our cross-level models. No changes in results were
observed, and the only control variable that was significant was gen-
der in predicting physical health (γ=.21, p< .05). This finding sug-
gests that women in our sample have worse physical health than men
do. Based on this result and because gender is sometimes associated
with the experience of negative emotions, we ran additional post hoc
analyses examining gender. As a first step, we examined the main
effects of gender (Level 2) on other Level 2 variables of trait PA, trait
NA, and trait affect balance (PA/NA), as well as on Level 1 variables of
daily PA, NA, and well-being. We found no significant effects at the
trait level, nor was gender associated with daily NA. But we did find a
significant association between gender and daily PA, such that women
experienced less PA daily than men did (γ=.28, p< .01). We found
no significant main effects of gender on life satisfaction or mental
health, but as expected based on our analyses of the control variables,
women reported worse physical health than men (γ=.24, p< .01).
Given the main effects of gender on both daily PA and daily physical
health, we reestimated our cross-level moderation model for physical
health (testing Hypothesis 3c) in two ways. First, in addition to trait
affect balance, we added gender as a Level 2 moderator. There was
no change in the results, nor was the gender term significant. Next,
we examined gender alone as a moderator (removing trait affect bal-
ance at Level 2 from the model). Again, there was no significant
change in the results. Together, these findings reveal significant differ-
ences between men and women in the experience of daily PA (where
women report lower PA) and physical health (where women report
lower physical health), but gender had no effect on the association
between affect balance and health at the trait or daily levels.
Finally, we included phase of the research as a control variable in
our cross-level model (Table 3). Substantive results remain unchanged
for mental and physical health (life satisfaction was measured in Phase I
only). Of note, however, phase significantly predicted mental health
(γ=.19, p< .05), suggesting Phase I participants (drawn from a wide
variety of jobs) had worse mental health than those in Phase II (drawn
from ambulatory healthcare); no differences were found for physical
health. Based on this result, we conducted additional post hoc analyses
TABLE 4 Features of response surfaces for daytime affect
balance predicting evening employee well-being at low and high levels
of trait affect balance (PA/NA)
Life satisfaction
Low AB High AB
b
0
: Intercept 3.33
**
3.85
**
b
1
: PA .37
**
.06
b
2
: NA .01 .21
b
3
:PA
2
.48
**
.35
b
4
:PANA .29 1.27
b
5
:NA
2
.04 1.03
*
Shape along PA =NA line
Slope (b
1
+b
2
) .38
*
.14
Curvature (b
3
+b
4
+b
5
) .80 2.64
Shape along PA =–NA line
Slope (b
1
b
2
) .35
**
.27
Curvature (b
3
b
4
+b
5
) .23 .11
Note: Estimates are unstandardized coefficients.
Abbreviation: AB, trait affect balance (trait PA/NA).
*p< .05. **p< .01 (two-tailed tests).
774 YOON ET AL.
at Level 1 and 2, as we did for gender. There was no significant change
in the results. Although there were mean differences in mental health
between the samples, we found the same association between affect
balance and health at both trait and daily levels in both samples.
5|DISCUSSION
Our goal in this research was to further understand if, why, and when
the balance between daily positive and negative affect at work mat-
ters for employee well-being. Taken as a whole, our results suggest
that affect balance matters; the dynamic interplay between PA and
NA on a daily basis adds to the prediction of both mental and physical
health, even after controlling for the main effects of PA and NA. Daily
affect balance was also a predictor of life satisfaction, but only for
people who had low levels of trait affect balance.
5.1 |Theoretical implications
Decades of research have been devoted to a search for the optimal
balance between PA and NA. This was the original aim of the affect
FIGURE 3 Response surfaces
for daytime affect balance
predicting employee life
satisfaction at low and high levels
of trait affect balance (PA/NA).
(a) Low trait PA/NA. (b) High trait
PA/NA
YOON ET AL.775
balance literature (Bradburn, 1969) and was the topic of numerous
publications and debates in the early 21st century. Contrary to previ-
ous studies (e.g., Fredrickson & Losada, 2005), our results suggest that
the benefit of workplace affect balance (excess PA over NA) on well-
being is not limited at the highest levels. Our results support the
notion that the interaction between PA and NA matters, but they do
not support a ceiling of an excess of PA over NA; we observed no
declines in well-being when PA was very high, nor when NA was
very low, nor—mostly importantly—when the excess of PA over NA
was at extreme levels. Indeed, the clearest conclusion to be drawn
from our analysis is that people flourish when they have an excess of
PA over NA, with more PA being better, a finding consistent with the
COR theory (Hobfoll, 1989), and challenging those who suggest a ceil-
ing on the beneficial effects of excess PA (see Fredrickson &
Losada, 2005).
Moreover, the beneficial effects of excess PA over NA are found
even when controlling for the main effects of PA and NA. Our use of
polynomial regression allows for the separation of the main effects of
daily PA and NA and the balance between them. This is an important
contribution to the literature because, across people and days, there is
a tendency toward a positivity offset, whereby people generally expe-
rience higher levels of PA than NA. Our methodology allows us to
control for this positivity offset and to more clearly establish the
incremental effects of excess PA. Responding to Gooty et al. (2009),
we test our hypotheses using experience sampling methods to cap-
ture the dynamic interaction of daily PA and NA.
We contribute to the COR theory (Hobfoll, 1989) by examining
the role that affect balance plays in several broad categories of well-
being. Most workplace studies applying the COR framework focus on
how stressors lead to negative employee behavior such as CWB, with
NA being an important mechanism through which stressors deplete
employees' resources, resulting in CWB (Bruk-Lee & Spector, 2006;
Fida et al., 2015; Fox et al., 2001). However, an important, yet
underexamined, aspect of the COR theory (Hobfoll, 1989) is that
employees are not passive agents, watching their resources being
depleted without taking any action. A major tenet of the COR theory
(Hobfoll, 1989) is that employees also seek to gain resources through-
out a workday to build their reservoir of resources. Yet, despite the
rich tradition of examining PA and NA on work attitudes and out-
comes (e.g., Thoresen et al., 2003), there has been less theoretical
emphasis in COR research that considers the reality that employees
experience both PA and NA throughout the day, with PA continuously
serving to undo the resource-depleting effects of NA on well-being
per the broaden-and-build explanation (Fredrickson, 1998,2001). We
contribute to the COR theory by examining the dynamic effects of
both daily PA and NA—particularly, the “undoing”effects of PA over
NA (Fredrickson, 2003)—on broad categories of employee well-being
that extends beyond work: life satisfaction, mental health, and physi-
cal health. Recent studies (Cho & Tay, 2016; Sirgy et al., 2020) show
that employees' sense of low life satisfaction—an indicator of well-
being—can spill over from work to life and vice versa to potentially
affect employees' work behavior. Hence, it is important to consider
how affect balance experienced at work can affect a wider array of
well-being outcomes after work such as the three criteria in this study
since these well-being outcomes may directly affect employees'
performance.
We also contribute to theory by examining various affective strat-
egies people can utilize to achieve well-being, another area where
there is considerable debate in the literature. On the one hand, our
results are consistent with the existing literature in that they show
that for well-being, in general, both increasing PA and decreasing NA
are important. On the other hand, our use of polynomial regression,
and especially the response surface analysis, suggests the need for
more research on the differential effects of increasing PA and
decreasing NA. Well-being outcomes that are appetitive in nature,
involving approach goals and striving for and attaining success, such
as life satisfaction, may depend more on achieving affect balance via
increases in PA (Cohn et al., 2009; Kuppens et al., 2008)—at least for
people who do not have high levels of trait affect balance. In contrast,
when well-being is operationalized in an avoidance-oriented way, as
the absence of health complaints, then reductions in NA may be most
important (Mayne, 1999; Pressman et al., 2013).
Although we have only begun to scratch the surface of this line
of inquiry, our results suggest promise for future research focused on
well-being achieved via promotion or prevention strategies. Not only
might daily PA and NA relate differently to various aspects of well-
being (e.g., life satisfaction vs. health), but increases in PA may be
more important for people with a promotion regulatory focus, who
are “concerned with advancement, growth, and accomplishment”
(Crowe & Higgins, 1997, p. 117), and decreases in NA may be more
important for those with a prevention focus, whose primary concerns
are “security, safety, and responsibility”(Crowe & Higgins, 1997,
p. 117). Our results demonstrate the importance of individual differ-
ences in trait affect balance, such that those with a shallow well of
excess PA are more responsive to daily variations in affect balance, at
least for life satisfaction. But other individual differences
(e.g., promotion and prevention focus) may also affect reactions to
daily affect balance for health outcomes.
5.2 |Methodological implications
Despite the importance of our methodology, which allows us to sepa-
rate the main effects of PA and NA from the effects of their relative
standing (excess PA over NA), we used a ratio measure of trait affect
balance for our moderation analyses. Indeed, there are reasons not to
completely abandon the use of ratio measures (PA/NA) of affect bal-
ance. Our cross-level results point out the usefulness of having a sin-
gle index of affect balance. Even if some incremental predictive
validity is lost by combining PA and NA, doing so can be useful in
some situations such as what we faced here, where a single index of
affect balance is needed to compare people. Along similar lines—when
the goal is to assess well-being at the national level, as in the World
Values Survey (Diener & Suh, 1999)—a composite index is helpful,
even though such measures conflate the level of PA and NA with
excess. Thus, single index measures of affect balance (such as PA/NA)
776 YOON ET AL.
can be helpful at the person or national level, but to fully understand
whether it is increased PA or decreased NA that is the most influential
strategy toward achieving well-being, it is critical to separately assess
the effects of PA, NA, and affect balance.
5.3 |Practical implications
Our findings are of practical significance. Employee well-being is a key
driver of work engagement and performance (Purcell, 2020), whereas
poor health is increasingly costly for organizations. According to the
Integrated Benefits Institute (2020), health problems cost US
employers $575 billion and 1.5 billion days of lost productivity per
year. Better mental and physical health associated with higher daily
PA and lower daily NA not only cuts down employers' healthcare
expenses but also boosts engagement and productivity. These bene-
fits extend to employees with low trait affect balance, as they have
greater life satisfaction when experiencing higher daily PA and lower
daily NA.
One of our goals in this research was to better understand the
various strategies to achieving affect balance that were most effec-
tive (increased PA, decreased NA, or both) so that managers and
organizations can provide interventions that could focus on areas of
the greatest potential gains in well-being for their employees. Our
results suggest that although increasing PA and reducing NA is
especially important for employees' mental and physical health,
increasing PA may be more important for employees' life satisfac-
tion. For managers and organizations, these results offer guidance
on how to manage their employees' affective experiences through-
out the day. For PA, this may take a two-pronged approach from
managers and the organization. Managers may introduce more daily
positive experiences for their employees by encouraging them to
make positive connections with others (Diener & Seligman, 2002).
Managers may foster these positive connections for employees by
providing them with emotional support during stressful times during
the day or by building a team where its members willingly provide
task assistance to each other (Colbert et al., 2016). Organizations
can introduce mindfulness training for employees to help them bet-
ter capture PA from positive experiences by reflecting on good
things that employees experience throughout the day or the week
(Sawyer et al., 2021) and by sharing them with others (Gable
et al., 2004).
For NA, this may take on a similar two-pronged approach as in
the case of PA. Managers may focus on minimizing employees' expo-
sure to daily negative events by making sure employees have breaks
after dealing with rude customers or by postponing criticism of
employees to days where they have more positive interactions with
customers. Organizations can introduce mindfulness training to mini-
mize the magnitude of reactions to daily negative events (Good
et al., 2016). Since negative interactions with customers and
coworkers contribute a large part to employees' NA throughout the
day and the day after (Tremmel & Sonnentag, 2018), mindfulness
training may encourage employees to take customers' and coworkers'
perspectives, which may help them minimize the magnitude of reac-
tions to daily negative events (Sawyer et al., 2021) and help them with
effective coping (Folkman & Moskowitz, 2000; Larsen &
Prizmic, 2008). If organizations are aiming to improve employee health
and reduce healthcare costs, our data suggest that decreasing NA and
increasing PA are effective. But what is most important is an excess
of PA over NA, which—no matter how it is achieved—can be expected
to improve well-being.
5.4 |Future research
5.4.1 | Sequencing of affect balance
People experience PA and NA in various sequences throughout the
workday. Hence, it is important to learn more about patterns of
changes in PA and NA over the course of the workday and their
effects on well-being. For instance, one person may maintain rela-
tively higher PA than NA consistently over the course of a day, while
another person may feel more PA than NA at first and then experi-
ence diminishing excess PA as the day wears on, or the opposite. In
the same way that the research on affective spin looks at movement
across the affective circumplex over time, future research might
examine the effects of variability in excess PA over time. Is it better to
have a small but stable level of excess PA? Or would well-being be
improved with extreme levels of excess at discrete points during the
day, even if that meant pockets of time with less excess PA? In work
settings, the affective shift model (Bledow et al., 2011)—which exam-
ines the sequential experience of PA and NA—has been shown to be
important for work engagement. In a similar way, the sequencing of
affect balance states may be informative. This research would also
dovetail nicely with recent reviews of the role of short-term variability
in emotions and psychological well-being (Houben et al., 2015) as well
as with studies that link affect spin and workplace well-being (Beal
et al., 2013).
5.4.2 | Affect balance and other work outcomes
We focused on employee well-being because it has been the focus of
the broader affect balance literature, but there are other important
and understudied work outcomes that may be related to affect bal-
ance as well. There has been growing attention to the role of affective
states in work outcomes such as performance, citizenship behavior,
and creativity (George & Zhou, 2007; Kaplan et al., 2009). For exam-
ple, the dual process model of creativity (de Dreu et al., 2008;
George & Zhou, 2007) suggests that both positive and negative
moods foster creativity because positive mood generates unusual,
novel thoughts and negative mood facilitates deep analysis of each
thought to evaluate its usefulness (Forgas, 2013). This research sug-
gests the possibility that low ratios of PA to NA, where both are high,
might be especially useful for creativity. Although our results strongly
suggest that there is no optimal maximum level of excess PA, further
YOON ET AL.777
consideration of the balance of positive and negative mood for spe-
cific work-related outcomes is warranted.
5.4.3 | Affect balance and gender
Past work in affect shows that females are more susceptible to
experiencing NA (Fujita et al., 1991) and emotional contagion
(Magen & Konasewich, 2011). Our data also reveal some gender dif-
ferences in that women in our sample experienced less PA and poorer
physical health (more symptoms) than men did. This is an important
issue that should be explored further in future research, as women
may need to especially focus on achieving a high ratio of PA to NA via
increased PA. This is especially true since the effects we found of
affect and affect balance on well-being did not differ for men and
women.
5.5 |Limitations
Although we have measures of affect balance at work and well-being
after work, and analyzed deviations from individuals' unique means
while assuming causality from affect balance to well-being, it is possi-
ble that well-being and affect balance have mutual influences on each
other. Indeed, both resource-based theories that serve as the basis for
our research suggest that people will invest their resources when they
have an excess of them. Thus, it is entirely plausible that when
employees experience an excess of PA, they invest those resources
into goal pursuit, goal achievement, and pleasurable activities, which
lead to increases in well-being, as suggested by our theorizing. But it
is also plausible that when employees come to work feeling good
about themselves and their lives, they then experience a more posi-
tive affect balance over the course of the workday (Rothbard &
Wilk, 2011). Thus, a spiraling cycle of accumulating affect balance and
well-being is likely, though it remains untested in our research.
Furthermore, there were range limitations in our data. The maxi-
mum daily ratio was 5.00, and the minimum was 0.20. Although this
range represents the daily experiences of our participants at work, the
maximum ratio that could be achieved was limited by the scales used.
Although affective experiences and moods are not isomorphic, esti-
mating affect balance as the interplay between positive and negative
work events—which can be counted without measurement
constraints—might be useful. Limited by the length of surveys, we
used a reduced set of items to cover high and low activations of state
PA and NA. Deploying a larger set of items might allow us to better
capture the entire range of affective states.
Lastly, whereas our sample size for mental and physical health is
comparable to the recommended sample size for experience sampling
studies (Gabriel et al., 2019), we recognize that our sample size for life
satisfaction is relatively small, as it was assessed only in Phase I. Thus,
we had less power to detect significant effects for life satisfaction.
This concern is alleviated to some degree since we found both signifi-
cant Level 1 effect and cross-level moderation for life satisfaction.
6|CONCLUSION
Our data clearly reveal daily benefits associated with a positive bal-
ance between PA and NA on health and happiness, even after con-
trolling for their main effects. We highlight the importance of
affect-inducing daily experiences and events at work for well-being,
as the right balance of PA and NA at work is beneficial for every-
one, and it is especially useful for people who have low trait affect
balance. Perhaps most importantly, our results suggest that a search
for the ideal ratio of PA to NA may not productive. Both our
between- and within-person results dispel the notion that there is a
ceiling on the effectiveness of a positive affect balance. More PA is
better for well-being, and people flourish when they have an excess
of PA over NA.
ACKNOWLEDGMENTS
We recognize and appreciate the staff at our respective universities
that supported us as faculty members while we conducted this
research. The fourth author received financial support from the
“BK21 FOUR (Fostering Outstanding Universities for Research)”
in 2021, and Yonsei University's 2020-2 Future-Leading Research
Initiative (2020-22-0493), and Signature Research Cluster Program of
2021 (2021-22-0006).
DATA AVAILABILITY STATEMENT
Research data are not shared.
ORCID
David J. Yoon https://orcid.org/0000-0001-6900-3531
ENDNOTES
1
Bono et al. (2013) and Koopmann et al. (2016) use data drawn from the
data repository to examine substantively different research questions.
2
Comparison of the four-factor model with the latter two alternative
models yielded negative values of scaled Δχ
2
(which may occur in certain
instances; Satorra & Bentler, 2010), preventing us from computing p-
values. Nonetheless, the latter two alternative models had smaller CFIs
and larger RMSEA and/or SRMR than the four-factor model, suggesting
poorer fit.
3
Full results for the post hoc analyses are available from the authors.
REFERENCES
Ashby, F. G., Isen, A. M., & Turken, A. U. (1999). A neuropsychological the-
ory of positive affect and its influence on cognition. Psychological
Review,106, 529–550. https://doi.org/10.1037/0033-295X.106.
3.529
Barefoot, J. C., Brummett, B. H., Helms, M. J., Mark, D. B., Siegler, I. C., &
Williams, R. B. (2000). Depressive symptoms and survival of patients
with coronary artery disease. Psychosomatic Medicine,62, 790–795.
https://doi.org/10.1097/00006842-200011000-00008
Barsade, S. G., & Gibson, D. E. (2007). Why does affect matter in organiza-
tions? Academy of Management Perspectives,21,36–59. https://doi.
org/10.5465/amp.2007.24286163
Baumeister, R. F., Bratslavsky, E., Finkenauer, C., & Vohs, K. D. (2001). Bad
is stronger than good. Review of General Psychology,5, 323–370.
https://doi.org/10.1037/1089-2680.5.4.323
778 YOON ET AL.
Beal, D. J., & Ghandour, L. (2011). Stability, change, and the stability of
change in daily workplace affect. Journal of Organizational Behavior,32,
526–546. https://doi.org/10.1002/job.713
Beal, D. J., Trougakos, J. P., Weiss, H. M., & Dalal, R. S. (2013). Affect spin
and the emotion regulation process at work. Journal of Applied Psychol-
ogy,98, 593–605. https://doi.org/10.1037/a0032559
Beal, D. J., Weiss, H. M., Barros, E., & MacDermid, S. M. (2005). An epi-
sodic process model of affective influences on performance. Journal of
Applied Psychology,90, 1054–1068. https://doi.org/10.1037/0021-
9010.90.6.1054
Bledow, R., Schmitt, A., Frese, M., & Kühnel, J. (2011). The affective shift
model of work engagement. Journal of Applied Psychology,96, 1246–
1257. https://doi.org/10.1037/a0024532
Bono, J. E., Braddy, P. W., Liu, Y., Gilbert, E. K., Fleenor, J. W.,
Quast, L. N., & Center, B. A. (2017). Dropped on the way to the top:
Gender and managerial derailment. Personnel Psychology,70, 729–768.
https://doi.org/10.1111/peps.12184
Bono, J. E., Foldes, H. J., Vinson, G., & Muros, J. P. (2007). Workplace emo-
tions: The role of supervision and leadership. Journal of Applied Psy-
chology,92, 1357–1367. https://doi.org/10.1037/0021-9010.92.5.
1357
Bono, J. E., Glomb, T. M., Shen, W., Kim, E., & Koch, A. J. (2013). Building
positive resources: Effects of positive events and positive reflection
on work stress and health. Academy of Management Journal,56,
1601–1627. https://doi.org/10.5465/amj.2011.0272
Bradburn, N. M. (1969). The structure of psychological well-being. Aldine
Publishing Company.
Brickman, P., & Campbell, D. T. (1971). Hedonic relativism and planning
the good society. In M. H. Apley (Ed.), Adaptation level theory: A sympo-
sium (pp. 287–302). Academic Press.
Bruk-Lee, V., & Spector, P. E. (2006). The social stressors-
counterproductive work behaviors link: Are conflicts with supervisors
and coworkers the same? Journal of Occupational Health and
Psychology,11, 145–156. https://doi.org/10.1037/1076-8998.11.
2.145
Cacioppo, J. T., & Berntson, G. G. (1994). Relationship between attitudes
and evaluative space: A critical review, with emphasis on the separabil-
ity of positive and negative substrates. Psychological Bulletin,115,
401–423. https://doi.org/10.1037/0033-2909.115.3.401
Carnethon, M. R., Kinder, L. S., Fair, J. M., Stafford, R. S., & Fortmann, S. P.
(2003). Symptoms of depression as a risk factor for incident diabetes:
Findings from the national health and nutrition examination epidemio-
logic follow-up study 1971-1992. American Journal of Epidemiology,
158, 416–423. https://doi.org/10.1093/aje/kwg172
Carver, C. S., & White, T. L. (1994). Behavioral inhibition, behavioral activa-
tion, and affective responses to impending reward and punishment:
The BIS/BAS scales. Journal of Personality and Social Psychology,67,
319–333. https://doi.org/10.1037/0022-3514.67.2.319
Centers for Disease Control and Prevention. (2018). Well-being concepts.
https://www.cdc.gov/hrqol/wellbeing.htm
Cho, E., & Tay, L. (2016). Domain satisfaction as a mediator of the relation-
ship between work–family spillover and subjective well-being: A longi-
tudinal study. Journal of Business and Psychology,31, 445–457.
https://doi.org/10.1007/s10869-015-9423-8
Cohn, M. A., Fredrickson, B. L., Brown, S. L., Mikels, J. A., & Conway, A. M.
(2009). Happiness unpacked: Positive emotions increase life satisfac-
tion by building resilience. Emotion,9, 316–368. https://doi.org/10.
1037/a0015952
Colbert, A. E., Bono, J. E., & Purvanova, R. K. (2016). Flourishing via work-
place relationships: Moving beyond instrumental support. Academy of
Management Journal,59, 1199–1223. https://doi.org/10.5465/amj.
2014.0506
Connolly, J. J., & Viswesvaran, C. (2000). The role of affectivity in job satis-
faction: A menta-analysis. Personality and Individual Differences,29,
265–281. https://doi.org/10.1016/S0191-8869(99)00192-0
Crowe, E., & Higgins, E. T. (1997). Regulatory focus and strategic inclina-
tions: Promotion and prevention in decision-making. Organizational
Behavior and Human Decision Processes,69, 117–132. https://doi.org/
10.1006/obhd.1996.2675
Danna, K., & Griffin, R. W. (1999). Health and well-being in the workplace:
A review and synthesis of the literature. Journal of Management,25,
357–384. https://doi.org/10.1016/S0149-2063(99)00006-9
de Dreu, C. K. W., Baas, M., & Nijstad, B. A. (2008). Hedonic tone and acti-
vation level in the mood-creativity link: Toward a dual pathway to cre-
ativity model. Journal of Personality and Social Psychology,94, 739–
756. https://doi.org/10.1037/0022-3514.94.5.739
Derogatis, L. R. (1975). The affect balance scale. Clinical Psychometric
Research.
Diener, E., Emmons, R. A., Larsen, R. J., & Griffin, S. (1985). The satisfaction
with life scale. Journal of Personality Assessment,49,71–75. https://
doi.org/10.1207/s15327752jpa4901_13
Diener, E., & Lucas, R. E. (1999). Personality and subjective well-being. In
D. Kahneman, E. Diener, & N. Schwarz (Eds.), Well-being: Foundations
of hedonic psychology (pp. 213–229). Russell Sage Foundation.
Diener, E., & Seligman, M. E. P. (2002). Very happy people. Psychological
Science,13,81–84. https://doi.org/10.1111/1467-9280.00415
Diener, E., & Suh, E. M. (1999). National differences in subjective well-
being. In D. Kahneman, E. Diener, & N. Schwarz (Eds.), Well-being: The
foundations of hedonic psychology (pp. 434–451). Russell Sage
Foundation.
Diener, E., Suh, E. M., Lucas, R. E., & Smith, H. L. (1999). Subjective well-
being: Three decades of progress. Psychological Bulletin,125, 276–
302. https://doi.org/10.1037/0033-2909.125.2.276
Diener, E., Wirtz, D., Tov, W., Kim-Prieto, C., Choi, D., Oishi, S., & Biswas-
Diener, R. (2010). New well-being measures: Short scales to assess
flourishing and positive and negative feelings. Social Indicators
Research,97, 143–156. https://doi.org/10.1007/s11205-009-9493-y
Edwards, J. R. (2002). Alternatives to difference scores: Polynomial regres-
sion analysis and response surface methodology. In F. Drasgow & N.
Schmitt (Eds.), Measuring and analyzing behavior in organizations:
Advances in measurement and data analysis (pp. 350–400). Jossey-
Bass.
Edwards, J. R., & Rothbard, N. P. (1999). Work and family stress and well-
being: An examination of person–environment fit in the work and fam-
ily domains. Organizational Behavior and Human Decision Processes,77,
85–129. https://doi.org/10.1006/obhd.1998.2813
Elfenbein, H. A. (2007). Emotion in organizations: A review and theoretical
integration. Academy of Management Annals,1, 315–386. https://doi.
org/10.5465/078559812
Everson, S. A., Goldberg, D. E., Kaplan, G. A., Julkunen, J., & Salonen, J. T.
(1998). Anger expression and incident hypertension. Psychosomatic
Medicine,60, 730–735. https://doi.org/10.1097/00006842-
199811000-00014
Feldman Barrett, L., & Russell, J. A. (1998). Independence and bipolarity in
the structure of current affect. Journal of Personality and Social Psychol-
ogy,74, 967–984. https://doi.org/10.1037/0022-3514.74.4.967
Fida, R., Paciello, M., Tramontano, C., Fontaine, R. G., Barbaranelli, C., &
Farnese, M. L. (2015). An integrative approach to understanding
counterproductive work behavior: The roles of stressors, negative
emotions, and moral disengagement. Journal of Business Ethics,130,
131–144. https://doi.org/10.1007/s10551-014-2209-5
Folkman, S., & Moskowitz, J. T. (2000). Positive affect and the other side
of coping. American Psychologist,55, 647–654. https://doi.org/10.
1037/0003-066X.55.6.647
Fong, C. T., & Tiedens, L. Z. (2002). Dueling experiences and dual ambiva-
lence: Emotional and motivational ambivalence of women in high sta-
tus positions. Motivation and Emotion,26, 105–121. https://doi.org/
10.1023/A:1015198209285
Forgas, J. P. (2013). Don't worry, be sad!: On the cognitive, motivational,
and interpersonal benefits of negative mood. Current Directions in
YOON ET AL.779
Psychological Science,22, 225–232. https://doi.org/10.1177/
0963721412474458
Fox, S., Spector, P. E., & Miles, D. (2001). Counterproductive work behav-
ior (CWB) in response to job stressors and organizational justice: Some
mediator and moderator tests for autonomy and emotions. Journal of
Vocational Behavior,59, 291–309. https://doi.org/10.1006/jvbe.2001.
1803
Fredrickson, B. L. (1998). What good are positive emotions? Review of Gen-
eral Psychology,2, 300–319. https://doi.org/10.1037/1089-2680.2.
3.300
Fredrickson, B. L. (2001). The role of positive emotions in positive psychol-
ogy: The broaden-and-build theory of positive emotions. American
Psychologist,56, 218–226. https://doi.org/10.1037/0003-066X.56.
3.218
Fredrickson, B. L. (2003). The value of positive emotions: The emerging
science of positive psychology is coming to understand why it's good
to feel good. American Scientist,91, 330–335. https://doi.org/10.
1511/2003.4.330
Fredrickson, B. L., & Losada, M. F. (2005). Positive affect and the complex
dynamics of human flourishing. American Psychologist,60, 678–686.
https://doi.org/10.1037/0003-066X.60.7.678
Friedman, H. S., & Booth-Kewley, S. (1987). The “disease-prone personal-
ity”: A meta-analytic view of the construct. American Psychologist,42,
539–555. https://doi.org/10.1037/0003-066X.42.6.539
Fujita, F., Diener, E., & Sandvik, E. (1991). Gender differences in negative
affect and well-being: The case for emotional intensity. Journal of Per-
sonality and Social Psychology,61, 427–434. https://doi.org/10.1037/
0022-3514.61.3.427
Gable, S. L., Reis, H. T., Impett, E. A., & Asher, E. R. (2004). What do you
do when things go right? The intrapersonal and interpersonal benefits
of sharing positive events. Journal of Personality and Social Psychology,
87, 228–245. https://doi.org/10.1037/0022-3514.87.2.22
Gabriel, A. S., Podsakoff, N. P., Beal, D. J., Scott, B. A., Sonnentag, S.,
Trougakos, J. P., & Butts, M. M. (2019). Experience sampling methods:
A discussion of critical trends and considerations for scholarly
advancement. Organizational Research Methods,22, 969–1006.
https://doi.org/10.1177/1094428118802626
George, J. M., & Zhou, J. (2007). Dual tuning in a supportive context: Joint
contributions of positive mood, negative mood, and supervisory
behaviors to employee creativity. Academy of Management Journal,50,
605–622. https://doi.org/10.5465/AMJ.2007.25525934
Gibran, K. (2001). The prophet. Alfred A. Knopf, Inc. (Original work publi-
shed 1923)
Goldberg, D. P. (1972). The detection of psychiatric illness by questionnaire.
Oxford University Press.
Good, D. J., Lyddy, C. J., Glomb, T. M., Bono, J. E., Brown, K. W.,
Duffy, M. K., Baer, R. A., Brewer, J. A., & Lazar, S. W. (2016). Contem-
plating mindfulness at work: An integrative review. Journal of Manage-
ment,42, 114–142. https://doi.org/10.1177/0149206315617003
Gooty, J., Gavin, M., & Ashkanasy, N. M. (2009). Emotions research in OB:
The challenges that lie ahead. Journal of Organizational Behavior,30,
833–838. https://doi.org/10.1002/job.619
Hershcovis, M. S., Turner, N., Barling, J., Arnold, K. A., Dupré, K. E.,
Inness, M., LeBlanc, M. M., & Sivanathan, N. (2007). Predicting work-
place aggression: A meta-analysis. Journal of Applied Psychology,92,
228–238. https://doi.org/10.1037/0021-9010.92.1.228
Hideg, I., & van Kleef, G. A. (2017). When expressions of fake emotions
elicit negative reactions: The role of observers' dialectical thinking.
Journal of Organizational Behavior,38, 1196–1212. https://doi.org/10.
1002/job.2196
Hobfoll, S. E. (1989). Conservation of resources: A new attempt at concep-
tualizing stress. American Psychologist,44, 513–524. https://doi.org/
10.1037//0003-066x.44.3.513
Hofmann, D. A., Griffin, M. A., & Gavin, M. B. (2000). The application of
hierarchical linear modeling to organizational research. In K. J. Klein &
S. W. J. Kozlowski (Eds.), Multilevel theory, research, and methods in
organizations: Foundations, extensions, and new directions (pp. 467–
511). Jossey-Bass.
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. https://doi.org/
10.1037/a0038822
Integrated Benefits Institute. (2020). Poor health costs US Employers
$575 billion and 1.5 billion days of lost productivity per Integrated
Benefits Institute. https://www.ibiweb.org/poor-health-costs-us-
employers-575-billion/
Ito, T. A., & Cacioppo, J. T. (2005). Variations on a human universal: Indi-
vidual differences in positivity offset and negativity bias. Cognition and
Emotion,19,1–26. https://doi.org/10.1080/02699930441000120
Jonas, B. S., Franks, P., & Ingram, D. D. (1997). Are symptoms of anxiety
and depression risk factors for hypertension? Longitudinal evidence
from the National Health and nutrition examination survey I epidemio-
logic follow-up study. Archives of Family Medicine,6,43–49. https://
doi.org/10.1001/archfami.6.1.43
Judge, T. A., Bono, J. E., Erez, A., & Locke, E. A. (2005). Core self-
evaluations and job and life satisfaction: The role of self-concordance
and goal attainment. Journal of Applied Psychology,90, 257–268.
https://doi.org/10.1037/0021-9010.90.2.257
Kaplan, S., Bradley, J. C., Luchman, J. N., & Haynes, D. (2009). On the role
of positive and negative affectivity in job performance: A meta-
analytic investigation. Journal of Applied Psychology,94, 162–176.
https://doi.org/10.1037/a0013115
Koopmann, J., Lanaj, K., Bono, J., & Campana, K. (2016). Daily shifts in reg-
ulatory focus: The influence of work events and implications for
employee well-being. Journal of Organizational Behavior,37, 1293–
1316. https://doi.org/10.1002/job.2105
Koydemir, S., & Schütz, A. (2012). Emotional intelligence predicts compo-
nents of subjective well-being beyond personality: A two-country
study using self- and informant reports. Journal of Positive Psychology,
7, 107–118. https://doi.org/10.1080/17439760.2011.647050
Krantz, D. S., & McCeney, M. K. (2002). Effects of psychological and social
factors on organic disease: A critical assessment of research on coro-
nary heart disease. Annual Review of Psychology,53, 341–369. https://
doi.org/10.1146/annurev.psych.53.100901.135208
Kreft, I., & de Leeuw, J. (1998). Introducing multilevel modeling. Sage
Publications.
Kubzansky, L. D., & Kawachi, I. (2000). Going to the heart of the matter:
Do negative emotions cause coronary heart disease? Journal of Psycho-
somatic Research,48, 323–337. https://doi.org/10.1016/S0022-3999
(99)00091-4
Kuppens, P., Realo, A., & Diener, E. (2008). The role of positive and nega-
tive emotions in life satisfaction judgment across nations. Journal of
Personality and Social Psychology,95,66–75. https://doi.org/10.1037/
0022-3514.95.1.66
Lanaj, K., Gabriel, A. S., & Chawla, N. (2021). The self-sacrificial nature of
leader identity: Understanding the costs and benefits at work and
home. Journal of Applied Psychology,106, 345–363. https://doi.org/10.
1037/apl0000505
Larsen, R. J. (2002). Differential contributions of positive and negative
affect to subjective well-being. In J. A. da Silva, E. H. Matsushima, &
N. P. Riberio-Filho (Eds.), Annual meeting of the International Society for
Psychophysics (Vol. 18) (pp. 186–190). Editora Legis Summa Ltda.
Larsen, R. J. (2009). The contributions of positive and negative affect to
emotional well-being. Psychological Topics,18, 247–266.
Larsen, R. J., & Ketelaar, T. (1991). Personality and susceptibility to positive
and negative emotional states. Journal of Personality and Social Psychol-
ogy,61, 132–140. https://doi.org/10.1037/0022-3514.61.1.132
Larsen, R. J., & Prizmic, Z. (2008). Regulation of emotional well-being:
Overcoming the hedonic treadmill. In M. Eid & R. J. Larsen (Eds.), The
science of subjective well-being (pp. 258–289). Guilford Press.
780 YOON ET AL.
Lennard, A. C., Scott, B. A., & Johnson, R. E. (2019). Turning frowns (and
smiles) upside down: A multilevel examination of surface acting posi-
tive and negative emotions on well-being. Journal of Applied Psychol-
ogy,104, 1164–1180. https://doi.org/10.1037/apl0000400
Lustman, P. J., Frank, B. L., & McGill, J. B. (1991). Relationship of personal-
ity characteristics to glucose regulation in adults with diabetes. Psycho-
somatic Medicine,53, 305–312. https://doi.org/10.1097/00006842-
199105000-00004
Lyubomirsky, S., King, L., & Diener, E. (2005). The benefits of frequent pos-
itive affect: Does happiness lead to success? Psychological Bulletin,
131, 803–855. https://doi.org/10.1037/0033-2909.131.6.803
Magen, E., & Konasewich, P. A. (2011). Women support providers are
more susceptible than men to emotional contagion following brief
supportive interactions. Psychology of Women Quarterly,35, 611–616.
https://doi.org/10.1177/0361684311423912
Mayne, T. J. (1999). Negative affect and health: The importance of being
earnest. Cognition and Emotion,13, 601–635. https://doi.org/10.
1080/026999399379203
Miner, A. G., Glomb, T. M., & Hulin, C. (2005). Experience sampling mood
and its correlates at work. Journal of Occupational and Organizational
Psychology,78,171–193. https://doi.org/10.1348/096317905X40105
Muthén, L. K., & Muthén, B. O. (2012). Mplus user's guide (7th ed.).
Muthén & Muthén.
Norris, C. J., Larsen, J. T., Crawford, L. E., & Cacioppo, J. T. (2011). Better
(or worse) for some than others: Individual differences in the positivity
offset and negativity bias. Journal of Research in Personality,45, 100–
111. https://doi.org/10.1016/j.jrp.2010.12.001
Peeters, G., & Czapinski, J. (1990). Positive-negative asymmetry in evalua-
tions: The distinction between affective and informational negativity
effects. European Review of Social Psychology,1,33–60. https://doi.
org/10.1080/14792779108401856
Pierce, J., Zhdanova, L., & Lucas, T. (2017). Positive and negative affectiv-
ity, stress, and well-being in African-Americans: Initial demonstration
of a polynomial regression and response surface methodology
approach. Psychology & Health,33, 445–464. https://doi.org/10.
1080/08870446.2017.1368510
Pressman, S. D., & Cohen, S. (2005). Does positive affect influence health?
Psychological Bulletin,131, 925–971. https://doi.org/10.1037/0033-
2909.131.6.925
Pressman, S. D., Gallagher, M. W., & Lopez, S. J. (2013). Is the emotion-
health connection a “first-world problem”?Psychological Science,24,
544–549. https://doi.org/10.1177/0956797612457382
Purcell, J. (2020). Employee well-being: A new perspective on ROI. Forbes.
https://www.forbes.com/sites/jimpurcell/2020/02/11/employee-
wellbeing-a-new-perspective-on-roi/?sh=577568df574d
Richman, L. S., Kubzansky, L., Maselko, J., Kawachi, I., Choo, P., & Bauer, M.
(2005). Positive emotion and health: Going beyond the negative. Health
Psychology,24,422–429. https://doi.org/10.1037/0278-6133.24.4.422
Rothbard, N. P., & Wilk, S. L. (2011). Waking up on the right or wrong side
of the bed: Start-of-workday mood, work events, employee affect, and
performance. Academy of Management Journal,54, 959–980. https://
doi.org/10.5465/amj.2007.0056
Ryan, R. M., & Deci, E. L. (2001). On happiness and human potentials: A
review of research on hedonic and eudaimonic well-being. Annual
Review of Psychology,52, 141–166. https://doi.org/10.1146/annurev.
psych.52.1.141
Ryff, C. D. (1989). Happiness is everything, or is it? Explorations on the
meaning of psychological well-being. Journal of Personality and Social Psy-
chology,57,1069–1081. https://doi.org/10.1037/0022-3514.57.6.1069
Ryff, C. D., & Singer, B. (1998). The contours of positive human
health. Psychological Inquiry,9,1–28. https://doi.org/10.1207/
s15327965pli0901_1
Ryff, C. D., & Singer, B. (2000). Interpersonal flourishing: A positive health
agenda for the new millennium. Personality and Social Psychology
Review,4,30–44. https://doi.org/10.1207/S15327957PSPR0401_4
Sanjuán, P. (2011). Affect balance as mediating variable between effective
psychological functioning and satisfaction with life. Journal of Happi-
ness Studies,12, 373–384. https://doi.org/10.1007/s10902-010-
9199-5
Satorra, A., & Bentler, P. M. (2010). Ensuring positiveness of the scaled dif-
ference chi-square test statistic. Psychometrika,75, 243–248. https://
doi.org/10.1007/s11336-009-9135-y
Sawyer, K. B., Thoroughgood, C. N., Stillwell, E. E., Duffy, M. K.,
Scott, K. L., & Adair, E. A. (2021). Being present and thankful: A multi-
study investigation of mindfulness, gratitude, and employee helping
behavior. Journal of Applied Psychology.https://doi.org/10.1037/
apl0000903
Schwartz, R. M., Reynolds, C. F. III, Thase, M. E., Frank, E., Fasiczka, A. L., &
Haaga, D. A. (2002). Optimal and normal affect balance in psychother-
apy of major depression: Evaluation of the balanced states of mind
model. Behavioural and Cognitive Psychotherapy,30, 439–450. https://
doi.org/10.1017/S1352465802004058
Sirgy, M. J., Lee, D. J., Park, S., Joshanloo, M., & Kim, M. (2020). Work-
family spillover and subjective well-being: The moderating role of cop-
ing strategies. Journal of Happiness Studies,21, 2909–2929. https://
doi.org/10.1007/s10902-019-00205-8
Snijders, T. A. B., & Bosker, R. J. (1999). Multilevel analysis: An introduction
to basic and advanced multilevel modeling. Sage Publications.
Stahl, J. E., Dossett, M. L., LaJoie, A. S., Denninger, J. W., Mehta, D. H.,
Goldman, R., Fricchione, G. L., & Benson, H. (2015). Relaxation
response and resiliency training and its effect on healthcare resource
utilization. PLoS ONE,10, e0140212. https://doi.org/10.1371/journal.
pone.0140212
Steptoe, A., Deaton, A., & Stone, A. A. (2015). Subjective wellbeing, health,
and ageing. The Lancet,385, 640–648. https://doi.org/10.1016/
S0140-6736(13)61489-0
Taylor, S. E. (1991). Asymmetrical effects of positive and negative events:
The mobilization-minimization hypothesis. Psychological Bulletin,110,
67–85. https://doi.org/10.1037/0033-2909.110.1.67
Thoresen, C. J., Kaplan, S. A., Barsky, A. P., Warren, C. R., & de
Chermont, K. (2003). The affective underpinnings of job perceptions
and attitudes: A meta-analytic review and integration. Psychological
Bulletin,129, 914–945. https://doi.org/10.1037/0033-2909.129.
6.914
Tov, W., & Lee, H. W. (2015). A closer look at the hedonics of everyday
meaning and satisfaction. Journal of Personality and Social Psychology,
111, 585–609. https://doi.org/10.1037/pspp0000081
Tremmel, S., & Sonnentag, S. (2018). A sorrow halved? A daily diary study
on talking about experienced workplace incivility and next-morning
negative affect. Journal of Occupational Health Psychology,23, 568–
583. https://doi.org/10.1037/ocp0000100
Tugade, M. M., & Fredrickson, B. L. (2004). Resilient individuals use posi-
tive emotions to bounce back from negative emotional experiences.
Journal of Personality and Social Psychology,86, 320–333. https://doi.
org/10.1037/0022-3514.86.2.320
Warr, P. (1992). Age and occupational well-being. Psychology and Aging,7,
37–45. https://doi.org/10.1037/0882-7974.7.1.37
Watson, D. (1988). Intraindividual and interindividual analyses of positive
and negative affect: Their relation to health complaints, perceived
stress, and daily activities. Journal of Personality and Social Psychology,
54, 1020–1030. https://doi.org/10.1037/0022-3514.54.6.1020
Watson, D., Clark, L. A., & Tellegen, A. (1988). Development and validation
of brief measures of positive and negative affect: The PANAS scales.
Journal of Personality and Social Psychology,54, 1063–1070. https://
doi.org/10.1037/0022-3514.54.6.1063
Weiss, H. M., & Cropanzano, R. (1996). Affective events theory: A theoret-
ical discussion of the structure, causes and consequences of affective
experiences at work. In B. M. Staw & L. L. Cummings (Eds.), Research in
organizational behavior: An annual series of analytical essays and critical
reviews (Vol. 18) (pp. 1–74). JAI Press.
YOON ET AL.781
AUTHOR BIOGRAPHIES
David J. Yoon is an Assistant Professor of Management at the
Franklin P. Perdue School of Business, Salisbury University. He
received his Ph.D. in human resources and industrial relations
from the University of Minnesota and a master's in human
resource management from Rutgers University. His areas of
research include mistreatment in the workplace, affect and emo-
tions, and power and leadership.
Joyce E. Bono is the W. A. McGriff III Professor of Management
at the Warrington College of Business Administration, University
of Florida. She received her Ph.D. in organizational behavior from
the University of Iowa and an M.S. in administration from the Uni-
versity of Notre Dame. Her research interests include leadership,
personality, motivation, the advancement of women leaders, and
the effects of positive events, resources, and relationships on
employee flourishing at work, and in life.
Tao Yang is an Assistant Professor of Management at the
Cameron School of Business, University of North Carolina
Wilmington. He received his Ph.D. in business administration from
the University of Minnesota. His research focuses on employee
moods and emotions, mindfulness, work–nonwork interface, and
the use of interventions to improve well-being.
KiYoung Lee is an Associate Professor at the School of Business,
Yonsei University, Seoul, Korea. He received his Ph.D. in business
administration from the University of Minnesota. His research
focuses on affect and emotions, workplace mistreatment, and
moral cognition and behavior.
Theresa M. Glomb is the Toro Company-David M. Lilly Chair of
Organizational Behavior in the Carlson School of Management,
University of Minnesota. She received her Ph.D. in social, organi-
zational, and individual differences psychology from the Univer-
sity of Illinois. Her research focuses on the well-being of workers,
defined broadly to include psychological, physiological, affective,
and familial effects, and the use of micro-interventions to pro-
mote flourishing.
Michelle K. Duffy is the Vernon Heath Chair in the Department
of Work and Organizations in the Carlson School of Management,
University of Minnesota. She received her Ph.D. in organizational
behavior and human resource management from the University of
Arkansas and a master's in psychology from Xavier University.
Her research focuses on the ways in which employee emotions
and affect influence organizational outcomes, the antecedents
and consequences of antisocial behavior at work, and the role of
interventions in improving employee well-being.
How to cite this article: Yoon, D. J., Bono, J. E., Yang, T., Lee,
K., Glomb, T. M., & Duffy, M. K. (2022). The balance between
positive and negative affect in employee well-being. Journal of
Organizational Behavior,43(4), 763–782. https://doi.org/10.
1002/job.2580
AP PE N D I X A : EQUATIONS FOR CROSS-LEVEL MODERATION
The equations used for the cross-level moderation (Table 3, Step 2)
wherein trait PA/NA moderated the relationship between daily affect
predictors and well-being are as below.
Within-person equation:
Yij ¼β0j þβ1j PAij þβ2j NAij þβ3j PA2
ij þβ4j PAij NAij þβ5j NA2
ij þεij
Between-person equations:
β0j ¼γ00 þγ01
PA
NA
j
þμ0j
β1j ¼γ10 þγ11
PA
NA
j
þμ1j
β2j ¼γ20 þγ21
PA
NA
j
þμ2j
β3j ¼γ130 þγ31
PA
NA
j
þμ3j
β4j ¼γ40 þγ41
PA
NA
j
þμ4j
β5j ¼γ50 þγ51
PA
NA
j
þμ5j
In the within-person equation, Yij refers to life satisfaction, mental
health, or physical health on day ifor person j;PAij and NAij refer to
positive and negative affect on day i, centered at person j's means;
and εij refers to residual in the outcome variable on day i. In the
between-person equations, PA
NA
jrefers to trait affect balance
(i.e., PA/NA) for person j, centered at the grand mean.
782 YOON ET AL.
