Retrieve 6

Psychology and Aging
© 2018 American Psychological Association 2019, Vol. 34, No. 1, 106 –123
0882-7974/19/$12.00 http://dx.doi.org/10.1037/pag0000305
Age-Related Differences in Levels and Dynamics of Workplace Affect
Susanne Scheibe Dannii Y. Yeung

University of Groningen City University of Hong Kong
Friederike Doerwald
University of Groningen
Affective experiences at work are a key contributing factor to long-term job-related well-being and
This article is intended solely for the personal use of the individual user and is not to be disseminated broadly.
effectiveness, yet may systematically change as workers get older. Given the central role of affect in work
This document is copyrighted by the American Psychological Association or one of its allied publishers.
settings, it is important to obtain a thorough understanding of older workers’ strengths and vulnerabilities
in affective functioning. This paper’s goal was to comprehensively study age differences in mean levels
and dynamics of affect (affect stability, occurrence of positive and negative daily work events, and
affective reactivity) and to link these with perceptions of global occupational well-being and effective-
ness. In three diary studies, employees of different occupational and cultural backgrounds (Hong Kong
Chinese managerial employees, German hospital employees, German office workers) reported daily
affective work events and affect across multiple workdays. Higher age was associated with more positive
and less negative affect (Study 1), more frequent positive daily work events (Study 2 and 3), and lower
variability of negative affect (Study 1 and 2). Age was unrelated to frequency of negative work events
and positive event reactivity (all studies). There were mixed age differences in negative event reactivity
(lower reactivity in Study 1 and on subtypes of affect in Study 2, but higher worry in Study 3). Several
of the indicators of affective experience emerged as mediators of positive age differences in work
engagement and self-rated task performance. Overall, with one exception, results point at maintenance
or improvement of workers’ affective functioning with age. Through improved affective functioning,
older workers likely contribute to organizational effectiveness.
Keywords: aging workforce, affect dynamics, affective work events, job attitudes, task performance
Affective experiences are an inevitable part of daily work life. at work, change as workers get older. Although there is a substan-
Moods and emotions experienced at work influence employees’ tial knowledge base on affective development across adulthood
cognitions, attitudes, and behaviors (Brief & Weiss, 2002). Work- from general-population samples, it remains open whether findings
place affect has for instance been found to predict health and generalize to worker samples (Scheibe & Zacher, 2013). Not only
absenteeism (Bono, Glomb, Shen, Kim, & Koch, 2013), job sat- are there substantial differences in age range—worker samples
isfaction (Dimotakis, Scott, & Koopman, 2011; Pelled & Xin, typically do not include adults over the age of 65—the affective
1999), affective commitment to the organization (C. D. Fisher, events people encounter, their motivational orientation, and their
2002), job performance (Kaplan, Bradley, Luchman, & Haynes, affect-regulatory options also may vary substantially between
2009; Spector & Fox, 2002), and team cohesion and effectiveness work and nonwork contexts (Scheibe, Wisse, & Schulz, 2017).
(Barsade & Knight, 2015). In light of the aging workforce, an Moreover, few studies to date have linked age differences in
important question is how different aspects of workplace affect, workplace affect to job-related attitudes and behaviors.
including mean levels and variability of daily affective experiences With the present project, we aim to provide a comprehensive
account of age differences in the affective lives of workers. Spe-
cifically, we adopt an intraindividual perspective to illuminate how
This article was published Online First October 8, 2018. aging shapes dynamics of workplace affect at the daily level. With
Susanne Scheibe, Department of Psychology, University of Groningen; affect dynamics we refer to variations and covariations of affective
Dannii Y. Yeung, Department of Social and Behavioural Sciences, City events and experiences over time, which can be examined when-
University of Hong Kong; Friederike Doerwald, Department of Psychol- ever such events and experiences are measured repeatedly in the
ogy, University of Groningen. same individuals in their natural environments. Pulling together
The first study was supported by the General Research Fund from the data from three diary studies of workers from different occupa-
Research Grants Council of the Hong Kong SAR (CityU153411). Susanne tional sectors (health care, administration, and managerial) and
Scheibe is supported by the Netherlands Organization for Scientific Re-
cultural backgrounds (Hong Kong and Germany), we examine age
search Vidi Grant (452-16-014). We thank Kristin Kammer for her assis-
tance in data collection of Study 3. differences in mean levels and dynamics of workplace affect and
Correspondence concerning this article should be addressed to Susanne their links with job attitudes and self-rated task performance.
Scheibe, Department of Psychology, University Groningen, Grote Kruis- Across the three studies, participants reported on daily work events
straat 2/1, 9712 TS Groningen, the Netherlands. E-mail: [email protected] and affective states for up to 15 workdays. Besides aggregates of
106
AGE AND DYNAMICS OF WORKPLACE AFFECT 107
mean levels of affect across sampling occasions, we derived three vulnerability integration theory (SAVI; Charles, 2010) holds that
indicators of affect dynamics: (a) affect stability (i.e., fluctuations changes in emotional well-being can be linked not only to differ-
around mean levels of affect over time), (b) the frequency of ences in remaining time in life, but also enhanced expertise in
positive and negative daily work events, and (c) affective reactivity regulating emotions accumulated over time. A recent systematic
to daily events (during or surrounding the event). Figure 1 provides review indeed yielded age-related advantages in several aspects of
an overview of our conceptual model, specifying the different emotion regulation (Doerwald, Scheibe, Zacher, & Van Yperen,
indicators of workplace affect that we considered. 2016). Although existing theories also specify boundary conditions
Our aim was twofold: first, to examine age differences in mean for positive age-related trends in affective functioning (Charles,
levels and dynamic aspects of workplace affect. A fine-grained 2010), overall the theories converge in predicting higher average
look at age differences in affective experience, as adopted here, is levels of affective well-being in older age.
common in the life span literature but is a novel contribution to Consistent with this assumption, cross-sectional studies that
research on work and aging which so far is largely limited to chart levels of affect in general-population samples find that age is
global, one-time assessments of affect. Second, we aimed to test associated with higher levels of affective well-being, which is
whether age has indirect effects via affective experiences on global mostly driven by reductions in mean levels of negative affect and
perceptions of job satisfaction, work engagement, and effective- less often by increases in positive affect (Charles, 2010; Morgan &
ness in the job. The consideration of positive work outcomes Scheibe, 2014). Longitudinal studies that follow participants for 10
represents an important extension of theories and research on up to 23 years show that positive age trends in affect do not
emotional development across adulthood. Knowledge of these typically continue in late adulthood; rather, affective well-being
issues will further our understanding of the practical implications
tends to improve until the mid to late 60s and decrease thereafter
of adult emotional development and the impact of the aging
(Sliwinski & Scott, 2014). However, this turn in affect trajectories
workforce on individual and organizational effectiveness.
in the mid to late 60s should not affect the working population as
most people retire before the downward trend in affective well-
Age Differences in Levels and Stability of being would set in. Indeed, several cross-sectional studies focusing
Workplace Affect on worker samples yielded higher levels of positive affect and/or
Theories of well-being across adulthood suggest that aging is lower levels of negative affect with age (Dahling & Perez, 2010;
marked by improvements in affective well-being at least into Lin, Wu, Chen, & Chen, 2014; Wegge, van Dick, Fisher, West, &
young-old age. Socioemotional selectivity theory (SST, Dawson, 2006; Yeung, Wong, & Lok, 2011), although other
Carstensen, 2006) holds that an increasing perception of time studies find age to be unrelated to affect (Bindl, Parker, Totterdell,
limitation leads older adults to place higher priority on experienc- & Hagger-Johnson, 2012; Lee & Allen, 2002; Tavares, 2016;
ing emotional meaning and well-being in the present, relative to Yeung & Fung, 2012). Positive age trends in affective well-being
acquiring knowledge and building resources. Through cognitive at work are consistent with meta-analytic findings on the relation-
and behavioral mechanisms, such as the well-documented positiv- ship between age and work motives, indicating that older workers
ity effect in attention and memory (Reed, Chan, & Mikels, 2014) emphasize intrinsic motives (such as accomplishment, job enjoy-
or a stronger tendency to avoid interpersonal conflicts in older ment, and skill utilization) more than younger workers do (Kooij,
adults (Birditt, Fingerman, & Almeida, 2005), these motivational de Lange, Jansen, Kanfer, & Dikkers, 2011). Thus, although the
shifts can benefit well-being. In an extension of SST, strength and evidence on worker age and affect is somewhat mixed, it is notable
Levels and dynamics

of workplace affect
Overall affective well-being
- Mean levels of affect [S1,2]
- Affect stability [S1,2]
+ +
Positive work outcomes
Affective event occurrence
+ + -- Task performance [S1]
Worker age - High # positive events [S1,2,3]
-- Job satisfaction [S1,2,3]
- Low # negative events [S1,2,3]
? ? -- Work engagement [S2,3]
Affective event reactivity

- During the event (intensity) [S1,3]
- Surrounding the event (frequency) [S2]
Figure 1. Conceptual model. Older workers, as compared with younger workers, are expected to show higher
affective well-being (Hypothesis 1) and higher stability of affect (Hypothesis 2), and to report more positive and
fewer negative work events (Hypothesis 3). The relation between worker age and affective event reactivity—
either during the event or surrounding the event—is unclear. Age is predicted to have indirect relationships with
positive work outcomes via affect indicators (Hypothesis 4). The figure also indicates which variable was
measured in which study (S ⫽ Study; S1 ⫽ Hong Kong managers; S2 ⫽ German health care workers; S3 ⫽
German office workers]. # ⫽ number.
108 SCHEIBE, YEUNG, AND DOERWALD
that negative age trends of affect are rarely observed. For the nights revealed that young and middle-aged adults reported more
present studies, we therefore predicted that worker age is associ- frequent everyday stressors than older adults (Almeida & Horn,
ated with higher mean levels of positive affect and lower mean 2004). Using the same dataset, Neupert, Almeida, and Charles
levels of negative affect (Hypothesis 1). (2007) found middle-aged adults to encounter fewer interpersonal
Affect stability represents the degree to which affect fluctuates arguments than young adults, but comparable numbers of negative
around mean levels of affect over time. It is regarded as an work events (hassles at work other than interpersonal conflicts).
individual difference characteristic in the sense that some people Other experience-sampling and qualitative studies similarly re-
vary more in their affective experiences than others (Eid & Diener, ported negative age trends in the number of unpleasant events
1999). Fluctuations may occur in response to various internal or (Sliwinski, Smyth, Hofer, & Stawski, 2006; Stawski, Sliwinski,
external influences, such as work stressors, and reflect how people Almeida, & Smyth, 2008).
deal with changes in their environment and regulate their emotions Despite the fact that the self-reported occurrence of positive
(Beal & Ghandour, 2011; Kuppens, Oravecz, & Tuerlinckx, 2010). daily events is about twice as high as for negative events (Aldwin,
A common operationalization is the intraindividual standard devi- Jeong, Igarashi, & Spiro, 2014; Charles et al., 2010; Gunaydin,
ation of affect across multiple measurement occasions. There is Selcuk, & Ong, 2016), very few studies have focused on age
solid evidence that affect variability is negatively linked to indi- differences in positive daily events. In one study of older women
cators of psychological health and well-being, with this relation- (in their 60s to 90s), positive daily events were reported less
ship being stronger for negative than positive affect variability frequently with age (Charles et al., 2010). A longitudinal study
(Houben, Van Den Noortgate, & Kuppens, 2015). In other words, with older men (in their 50s to 90s) across 16 years showed
more stable affective experiences appear to be a correlate of increases in occurrence of uplifts until the early 60s and decreases
mental health. thereafter (Aldwin et al., 2014). Unfortunately, given their age
Past studies with general-populations samples consistently show range these studies provide limited information on what would be
that affect stability is higher in older adults compared with young expected for a working sample. Nevertheless, as positive events
adults (Brose, Scheibe, & Schmiedek, 2013; Röcke, Li, & Smith, arise from attention and appraisal processes, just like negative
2009). This can partly be attributed to changes in life contexts. In events (Lazarus & Folkman, 1984), and given predictions of an
particular, younger and older adults, who are matched on stressor enhanced motivation to focus on the positive aspects and selection
occurrence, severity, and impact, were found to be more similar in into positive environments with age (Carstensen, 2006), one may
affective variability across 100 days than the unmatched groups of expect age-related increases in occurrence of positive daily work
younger and older adults (Brose et al., 2013). However, even when events. For the present study, we therefore predicted that older
accounting for life context characteristics, age differences in neg- workers report (a) lower occurrence of negative daily work events
ative affect variability were not fully eliminated (Brose et al., and (b) higher occurrence of positive daily work events (Hypoth-
2013). Another study also reported reduced variability for negative esis 3).
affect, though age differences in positive affect variability were not Along with the assessment of positive and negative daily events
confirmed (Grühn, Lumley, Diehl, & Labouvie-Vief, 2013). To the at work, the present studies also examine whether older workers
best of our knowledge, no study has yet assessed age differences show similar or different affective reactivity to these events. Af-
in affect variability in the work setting, rendering this an important fective reactivity is usually operationalized as the degree to which
area of inquiry. Based on developmental theories and prior find- affect experienced during, or close to, discrete events deviates
ings of older adults’ higher affect stability in general-population from baseline levels. According to the SAVI model, time is an
samples, we predicted that worker age is associated with higher important dimension to consider when studying age differences in
affect stability across measurement occasions (Hypothesis 2). affective reactivity to daily events (Charles, 2010). Age advan-
tages are mostly expected before and after negative events but not
Age Differences in Occurrence of and Reactivity to necessarily in the midst of negative events. Before and after
events, affective arousal is relatively low and older adults should
Daily Work Events
be able to effectively use cognitive– behavioral emotion regulation
Theoretically, more positive and stable workplace affect with strategies such as positive reappraisal or situation selection to
age could result from a lower exposure to negative daily work maintain well-being. In the midst of negative events when affec-
events and/or more exposure to positive events, or from effectively tive arousal is high, in contrast, such strategies are often difficult
regulating emotions vis-à-vis everyday work events (Scheibe & to use and older adults’ lower physiological flexibility likely
Zacher, 2013). Studying the occurrence of and affective reactivity prolongs recovery (Charles, 2010). In studies of daily stress, it is
to daily events experienced at work thus promises unique insight therefore important to consider how affective reactivity is mea-
into the impact of aging on the affective lives of workers and the sured: As the immediate response to an event (e.g., retrospectively
resulting work outcomes. The present studies make use of the diary rated as affect intensity or stressfulness), or as affective experi-
design to assess daily experiences at work, which provide infor- ences surrounding the event (e.g., affect frequency during the day
mation about the appraisal of daily discrete events as positive or that the event occurred).
negative and the covariation with affect. Keeping this distinction in mind, the available evidence on age
There is evidence from several general-population studies that differences in reactivity to daily stressors is mixed (see also
older adults report fewer distressing events in their everyday life Sliwinski & Scott, 2014). A limited number of studies that exam-
than younger adults do (see Riediger & Rauers, 2014, for review). ined immediate reactivity (operationalized as perceived stressful-
A large-scale investigation of U.S. adults ranging in age from 24 ness) found reduced affective reactivity to daily stressors in older
to 74 years reporting on daily experiences across eight consecutive as compared with younger adults (Birditt et al., 2005; Charles &
Almeida, 2007). Studies that examined affective experience sur- work predicted daily job satisfaction. Rothbard and Wilk (2011)
rounding daily negative events found either reduced reactivity with found that in a sample of call center employees, within-person
age (Brose et al., 2013; Piazza, Charles, & Almeida, 2007), no age positive and negative affect subsequent to calls was significantly
differences (Röcke et al., 2009; Stawski et al., 2008), or higher associated with daily objective task performance. Collectively,
affective reactivity with age for complex daily stressors (Wrzus, these studies suggest that affect dynamics at work predict work
Müller, Wagner, Lindenberger, & Riediger, 2013). Neupert et al. outcomes. Age differences in affect dynamics therefore likely
(2007) distinguished different types of stressors and observed trickle down to impact job-related attitudes and behaviors. We
reduced affective reactivity (measured as frequency of distress predicted that worker age is indirectly associated with positive job
on stressor days relative to nonstressor days) with age to interper- attitudes and higher sense of effectiveness and work engagement
sonal stressors, but no age differences in affective reactivity to via (a) mean levels of affect, (b) affect stability, (c) occurrence of
work stressors. One of the few available longitudinal studies found daily work events, and (d) affective reactivity to daily work events
an age-related increase in affective reactivity (also measured as (Hypothesis 4).
frequency of distress on stressor relative to nonstressor days) over
a time interval of 10 years (Sliwinski, Almeida, Smyth, & Stawski,

2009). The Present Studies

Again, there is a dearth of theory and research on age differ-
ences in affective reactivity to positive daily events. We are aware The present investigation comprises three diary studies with
of only one study on this issue: relative to younger adults, older samples from diverse occupational and cultural backgrounds and
adults reported larger reductions in negative affect on positive different methods of assessing daily affect and events. The samples
event days compared with nonevent days, though the event-related include Hong Kong Chinese managerial employees (Study 1),
increase in positive affect was comparable across ages (Gunaydin German employees of one large hospital operator (Study 2), and
et al., 2016). Importantly, whereas older adults tend to be more German administrative and office workers with heterogeneous
motivated than younger adults to maintain positive affect, they do occupational backgrounds (Study 3). The number of daily surveys
not report a heightened motivation to increase positive affect in was 15, 10, and three, respectively, for Studies 1 to 3. Whereas in
their daily life (Riediger, Schmiedek, Wagner, & Lindenberger, Studies 1 and 2, we obtained one affect and event measurement per
2009). Given the mixed prior evidence, we did not specify directed workday, in Study 3, participants reported their affect in response
hypotheses on worker age and affective reactivity, but posed the to up to 6 positive and 6 negative work events per workday.
following research questions: What is the relationship between Regarding affective reactivity, Studies 1 and 3 assessed affect
worker age and affective reactivity to daily positive and negative intensity during specific events, whereas Study 2 assessed affect
work events, and does it matter whether affect is assessed during frequency on the day of the event. Job attitudes and self-rated task
the event (as affect intensity) or on the day surrounding the event performance were assessed in a baseline survey. Due to design
(as affect frequency)? differences, Hypotheses 1 and 2 could be tested in the first two
studies only, while the remaining hypotheses and our research
Age, Affect Dynamics, and Work Outcomes questions on affective reactivity could be tested in all three studies.
The diverse samples and methods help determine robustness of
Affective daily events and their emotional sequelae also affect findings. Past studies showed that German and Chinese workers
work outcomes. Based on perspectives from affective events the- vary in the rank order of various work values (e.g., Chinese
ory (Weiss & Cropanzano, 1996) and the episodic process model workers rank achievement and advancement as more important,
of performance (Beal, Weiss, Barros, & MacDermid, 2005), one and meaningful work and positive social relationships as less
can speculate that improvements in the affective lives of workers
important than German workers; Elizur, Borg, Hunt, & Beck,
help explain previously identified age advantages in job attitudes
1991) and in the importance of work in life (i.e., German employ-
(Ng & Feldman, 2010). Both of these theories position moods and
ees perceive work as more important than Hong Kong Chinese
emotions experienced at work as key factors driving work-related
employees; World Values Survey Association, 2010-2014). This
outcomes. Affect fluctuations are thought to lead to fluctuations in
suggests that affective responses to daily work events, as well as
task performance, with negative emotions pulling cognitive re-
affect-work outcome relationships, could be less pronounced in
sources off-task and positive emotions enhancing task attention
(Beal et al., 2005). Although these processes operate on a momen- Hong Kong Chinese than German participants. The nature of
tary level, effects accumulate over time, such that employees with affective work events can further vary depending on occupational
higher mean levels of affective well-being display more positive characteristics. Managerial and health care workers on average
job attitudes (Spector, 1997) and show higher job performance face higher emotional demands (such as establishing and main-
(Kaplan et al., 2009). taining relationships or influencing others) than office workers,
In addition, there is accumulating evidence on intraperson as- whereas office work entails higher demands of interacting with
sociations between workplace affect, job attitudes, and task per- computers and analyzing data or information (OⴱNET Occupa-
formance. Bledow and colleagues (2011), for instance, found that tional Network; United States Department of Labor/Employment
negative affect and negative events experienced in the morning of and Training Administration, 2018). Hence, the frequency and
a working day predicted higher work engagement in the afternoon nature of affective work events, and their affective ramifications,
if workers managed to turn negative affect into positive affect may differ between samples. Our data therefore represent an
during the day. Dimotakis et al. (2011) report that positive and important step toward establishing generalizability across cultures
negative affect resulting from daily interpersonal interactions at and occupations.
Study 1 (Hong Kong Chinese Managers) days were 654 and 679, respectively. Accordingly, the total num-
ber of event reports is 1333, with an average of 4.51 positive and
4.82 negative event days per participant.
Method
Positive and negative affective responses to work events.
Participants and procedure. One hundred fifty Chinese After the participants gave ratings to the most salient event hap-
managerial employees aged between 25 and 62 years (M ⫽ 42.45, pened on the day, they were asked to report their affective re-
SD ⫽ 8.97) took part in this diary study that spanned a 3-week sponses to this event. With reference to prior studies on emotions
period. Among them, 54% were female and 60.7% were in a at work (Fisher, 2002; Yeung & Fung, 2012), this study measured
supervisory position. Average organizational tenure was 11.31 six positive emotions (happy, joyful, enthusiastic, excited, calm,
years (SD ⫽ 9.94). and relaxed), and five negative emotions (angry, sad, anxious,
Participants met the research assistant individually to learn worried, and irritable) experienced during the event. Participants
about the study details and procedures. They were first asked to rated these affective terms on a 5-point Likert scale (1 ⫽ not at all
sign the paper consent form and confidentiality of responses was to 5 ⫽ extremely). A confirmatory factor analysis (CFA) was
assured. Participants then completed a baseline online survey conducted using MPlus 7 (Muthén & Muthén, 1998 –2015). The
which assessed demographic information, job satisfaction, and results of CFA revealed that the goodness-of-fit of the two-factor
self-reported task performance. Each participant was then assigned model containing positive and negative affect (␹2 ⫽ 575, df ⫽ 85;
a unique login ID and password and was instructed to complete a CFI ⫽ .94; RMSEA ⫽ .07, ␻ ⫽ .50) was better than that of a
10-min online survey each evening for 15 consecutive workdays. single-factor model (␹2 ⫽ 1870, df ⫽ 90; CFI ⫽ .79; RMSEA ⫽
There were 37 participants who completed 16 surveys, whereas the .12, ␻ ⫽ .59) and a four-factor model distinguishing affect by
rest completed 15 surveys (compliance rate 100%). Most of the valence and arousal (␹2 ⫽ 2814, df ⫽ 83; CFI ⫽ .67; RMSEA ⫽
participants (93.4%) submitted the daily reports on the same eve- .16, ␻ ⫽ .50). Therefore, positive and negative affect scores were
ning. For the participants who did not submit the report on a computed for each diary report.
particular day, a reminder was sent, and they were allowed to Person-level variables. In the baseline survey, we recorded
complete it before 11 a.m. on the next day. In each diary, the demographic information, including age (in years), gender (0 ⫽
participants were asked to report the events happened on the day men, 1 ⫽ women), organizational tenure (in years), and job posi-
and their affective responses to these events. We also recorded the tion (0 ⫽ officer, 1 ⫽ manager). We also assessed two global work
day of the week (0 ⫽ Monday, 1 ⫽ Tuesday, 3 ⫽ Wednesday, outcomes. Kunin’s (1955) single-item faces scale was used to
etc.) and time of the day (1 ⫽ morning; 2 ⫽ evening) and added measure the participants’ overall job satisfaction. The participants
them as covariates in the analyses. Each participant received were instructed to choose one of the seven faces that best repre-
HKD500 (⬃USD63) after completing the study. The study was sents their feeling about their job in general, with higher scores
approved by the Research Ethics Committee at City University of indicating higher job satisfaction. Self-reported task performance
Hong Kong. was assessed using a four-item measure (Yeung & Fung, 2009), in
Measures. which participants rated their efficiency, quality, effectiveness, and
Daily work events. In each diary, the participants were asked overall performance on a 5-point scale (1 ⫽ unsatisfactory to 5 ⫽
to report the events happened at work. A list of nine positive and excellent). The measure had adequate reliability (␣ ⫽ .86).
nine negative work events was developed based on past studies on
work-related events (Mignonac & Herrbach, 2004) and interper-
Results
sonal tensions in the workplace (Yeung, Fung, & Chan, 2015).
Positive work events include “successfully completed a project or Preliminary analyses. Table 1 presents the descriptive and
a task” and “received praise from your supervisor.” Negative work correlation statistics of the person-level and day-level variables.
events include “job performance has been criticized,” and “had Age was positively correlated with organizational tenure (r ⫽ .57),
disagreement with other colleague about the task or organizational and gender was negatively associated with age (r ⫽ ⫺.28), orga-
policy.” The option “no event” was also given if none of the nizational tenure (r ⫽ ⫺.19), and rank of position (r ⫽ ⫺.25),
positive or negative events happened on the day. The reports with ps ⬍ .05. Positive and negative affect were moderately intercor-
no event (n ⫽ 954) were excluded from the following analyses as related; both at the person-level (r ⫽ ⫺.46) and day-level
affective responses were not assessed on these sampling days. (r ⫽ ⫺.66), ps ⬍ .05. Age was unrelated to job satisfaction and
After the participants had reported which event(s) happened performance.
during work, they were asked to give a rating from 1 ⫽ very bad Age differences in daily affect and work events. We tested
to 6 ⫽ very good to indicate the valence of each event. If the Hypotheses 1 (age differences in mean levels of affect) and 3 (age
participant had reported more than one event on a given day, only differences in event occurrence), as well as our research question
the event with the highest intensity was selected to provide further (age differences in affective reactivity) simultaneously in one
questions on emotional reactions to this event. For example, if two multilevel structural equation modeling (MSEM) using MPlus 7.
events were experienced by a participant during the day, one was Level 1 comprises day-level variables including positive and neg-
rated as 2 (bad) and the other as 1 (very bad), the latter event was ative affect and valence of the work event (0 ⫽ positive event, 1 ⫽
selected by the system and treated as the most salient event that negative event) experienced by each participant, and weekday and
happened on that day. This score was used to indicate the valence time of day as covariates. Level 2 includes the participants’ age. A
of the work event, with the score of 3 or below being categorized cross-level interaction was computed between age and valence of
as a negative event and the score of 4 or above being grouped as the work event to test age differences in affective reactivity. The
a positive event. The total numbers of positive and negative event number of diary entry, gender, organizational tenure, job position,
Table 1
Descriptive Statistics and Correlations Between Day-Level and Person-Level Variables, Study 1 (Chinese Managers)
Variable Mean (SD) 1 2 3 4 5 6 7 8 9 10 11 12
Variables assessed at person-level

1. Age 42.45 (8.97) —
2. Gender — ⴚ.28 —
3. Organizational tenure 11.31 (9.94) .57 ⴚ.19 —
4. Job position 0.61 (0.49) .02 ⴚ.25 ⴚ.17 —
5. Job satisfaction 4.91 (1.29) .03 .00 ⫺.07 ⫺.02 —
6. Task performance 3.87 (0.54) .14 ⫺.05 .03 ⫺.02 .41 —
Variables assessed at day-level
7. Positive affecta 2.49 (0.64) .32 ⫺.03 .13 ⫺.06 .20 .17 —
8. Negative affecta 1.66 (0.48) ⴚ.26 ⫺.01 ⴚ.16 .02 ⫺.13 ⴚ.17 ⴚ.46 —
9. iSD of positive affect 0.74 (0.29) ⫺.11 .05 ⴚ.25 .06 .06 .03 .07 .09 —
10. iSD of negative affect 0.58 (0.31) ⴚ.27 .14 ⴚ.22 ⫺.02 ⫺.03 ⫺.06 ⴚ.20 .73 .41 —
11. Positive reactivity slopeb ⫺1.31 (0.33) .17 ⫺.08 .19 .05 ⫺.01 ⫺.04 ⫺.04 ⫺.09 ⴚ.75 ⴚ.30 —
12. Negative reactivity slopeb 0.94 (0.39) ⴚ.35 .10 ⴚ.21 ⫺.11 ⫺.01 ⫺.05 ⫺.11 .76 .17 .83 ⴚ.19 —
13. Event valence 0.50 (0.23) ⫺.08 ⫺.06 ⫺.05 .17 ⴚ.22 ⴚ.19 ⴚ.65 .59 .05 .23 ⫺.02 .01
Note. N ⫽ 150. Gender was coded as 0 ⫽ men and 1 ⫽ women. iSD denotes intraindividual standard deviation. Job position was coded as 0 ⫽ officer
and 1 ⫽ manager. Event valence was coded as 0 ⫽ positive event and 1 ⫽ negative event occurred on the sampling day. Correlation coefficient displayed
in bold are significant at p ⬍ .05.
a
The scores of day-level variables were aggregated from the data of the daily diary study. b Positive and negative reactivity slopes refer to the difference
in affect when comparing days with a negative event and days with a positive event.
and total number of event days were included in the model as events relative to those in the lower position (B ⫽ .074, SE ⫽ .037,
covariates. In the multilevel analyses, the maximum likelihood p ⫽ .042).
method was used as the estimating technique. Below we report Results further show that compared with younger workers, older
unstandardized coefficients (labeled B) for all analyses. workers reported more positive affect (B ⫽ .020, SE ⫽ .006, p ⫽
As shown in Table 2, the multilevel analysis revealed that age .001) though no significant effect appeared for negative affect
was unrelated to the occurrence of daily negative events, failing to (B ⫽ ⫺.005, SE ⫽ .003, p ⫽ .075). This partially supports
support Hypothesis 3. Weekday was negatively related to the Hypothesis 1. Affect was also predicted by event occurrence. On
occurrence of daily negative events, B ⫽ ⫺.022, SE ⫽ .009, p ⫽ days with a negative work event (relative to days with a positive
.021, suggesting that the participants experienced fewer negative event), participants reported higher negative affect and lower pos-
events toward the end of the week. However, the participants in the itive affect (B ⫽ .931, SE ⫽ .044; and B ⫽ ⫺1.304, SE ⫽ .046,
managerial position were more likely to experience negative respectively, ps ⬍ .001). The relationship between valence of the
Table 2
Results from Multilevel Analyses Predicting Positive and Negative Affect and Occurrence of
Negative Events, Study 1 (Chinese Managers)
B (SE)
Measure Negative event occurrence Positive affect Negative affect
ⴱ ⴱⴱⴱ
Intercept 0.354 (0.141) 2.958 (0.215) 1.184 (0.149)ⴱⴱⴱ
Level 1 (Day-level)
Number of diary entry ⫺0.007 (0.004)ⴱ ⫺0.003 (0.004) ⫺0.001 (0.004)
Event valence — ⫺1.506 (0.648)ⴱ 0.617 (0.290)ⴱ
Weekday ⫺0.022 (0.009)ⴱ ⫺0.001 (0.009) ⫺0.002 (0.009)
Time of day 0.036 (0.057) ⫺0.023 (0.062) ⫺0.040 (0.056)
Level 2 (Person-level)
Age ⫺0.003 (0.002) 0.020 (0.006)ⴱⴱⴱ ⫺0.005 (0.003)
Gender ⫺0.009 (0.037) 0.001 (0.083) ⫺0.007 (0.045)
Organizational tenure 0.002 (0.002) ⫺0.005 (0.005) 0.001 (0.003)
Job position 0.074 (0.037)ⴱ 0.019 (0.083) 0.007 (0.044)
Number of event days 0.015 (0.006)ⴱⴱ 0.027 (0.012)ⴱ 0.011 (0.007)
Cross-level interaction
Age ⫻ Event valence — 0.006 (0.005) ⫺0.015 (0.005)ⴱⴱⴱ
Note. Valence of the event was coded as 0 ⫽ positive event and 1 ⫽ negative event. Weekday was coded 0 ⫽
Monday, 1 ⫽ Tuesday, 2 ⫽ Wednesday, etc. Time of day was coded as 1 ⫽ morning and 2 ⫽ evening. Gender
was coded as 0 ⫽ male and 1 ⫽ female. Job position was coded as 0 ⫽ officer and 1 ⫽ manager. The ⫺2log
likelihood value is ⫺2862.413 (df ⫽ 45).
ⴱ
p ⬍ .05. ⴱⴱ p ⬍ .01. ⴱⴱⴱ p ⬍ .001.
event and negative affect was moderated by age (B ⫽ ⫺.015, SE ⫽ operator in Germany by distributing flyers and through postings on
.005, p ⫽ .001), revealing that older participants experienced a the hospitals’ intranet and newsletter. One person was excluded
lower increase in negative affect than younger participants on days because of extensive missing data. The effective sample (N ⫽ 125)
with negative events. The relationship between valence of work ranged in age from 21 to 65 years (M ⫽ 46.45, SD ⫽ 9.95), and
events and positive affect was not moderated by age. Thus, in 58.7% participants were women. Average organizational tenure
response to our research question we found older adults to react was 13.06 (SD ⫽ 10.05) years. The sample comprised 28 physi-
less strongly to negative work events in terms of negative affect, cians, 53 nurses/health care professionals, 27 administrative work-
but equally strongly as younger adults in terms of positive affect. ers, and 18 participants with other professions (e.g., pedagogics).
Age differences in affect stability. To test Hypothesis 2 Upon recruitment, participants received a link to a baseline
regarding age differences in stability of affect, we calculated the online survey, which contained information about the study, an
intraindividual standard deviation (iSD) of positive and negative informed consent, demographic questions, as well as measures of
affect across days for each person. We regressed the iSD of job satisfaction and work engagement. Participants further indi-
positive and negative affect on age in separate analyses, account- cated their e-mail address, which was used to send them person-
ing for gender, tenure, and total number of diary days. Age alized links to the daily questionnaires. Over the next 10 workdays,
negatively predicted iSD of negative affect (B ⫽ ⫺.008, SE ⫽ participants received an individualized link at 11 a.m. (accessible
.003, p ⫽ .025), but not iSD of positive affect (B ⫽ .000, SE ⫽ for 24 hr) to a daily online survey assessing that days’ experienced
.003, p ⫽ .890). Thus, Hypothesis 2 was partially supported. When positive and negative affect and work events. Participants com-
additionally controlling for mean levels of negative affect, the age pleted the surveys at the end of the work shift or after work. In
effect was no longer significant, probably because of the high total, participants provided 1128 valid daily entries. A system error
correlation between mean and stability of negative affect (see caused 16 participants to provide data for more than 10 days, with
Table 1). a maximum of 13 entries. The number of valid diary entries per
Indirect associations of age with work-related outcomes. person ranged from 1 to 13 (M ⫽ 9.02, SD ⫽ 2.81). We recorded
To examine Hypothesis 4 about the indirect effect of age on the day of the week (0 ⫽ Monday, 1 ⫽ Tuesday, 3 ⫽ Wednesday,
work-related outcomes through affective experiences, mediation etc.) and the time that diary entries were completed (in hours since
analyses was conducted using PROCESS 2.12 (Hayes, 2012). As
midnight), which were used as covariates in the analyses. Upon
mediation requires a significant relationship between predictor
study completion, participants were provided personalized feed-
(age) and mediator (affect indicator), we only considered indirect
back on their job satisfaction and work engagement and how it
effects for those affect indicators that in the prior analyses yielded
differed on days with high/low positive and negative affect, which
a relation with age. In Study 1, this concerned the mean of positive
they could retrieve on website via a self-generated code. As further
affect and mean and iSD of negative affect, as well as the negative
incentive, the research team donated €5 for each completed base-
reactivity slope. The latter was obtained by saving the event ¡
line questionnaire and an additional €5 if the participant had
negative affect slope from a multilevel analysis, in which affect
completed at least seven daily surveys to a partner hospital in
was predicted by event valence and number of diary entry (see also
Israel. The research procedure was approved by the Ethical Com-
Charles, Piazza, Mogle, Sliwinski, & Almeida, 2013). For the
mittee Psychology at the University of Groningen.
mediation model, we thus specified age as predictor, the mean of
positive affect, the mean and iSD of negative affect, and the Measures.
affective reactivity slope as parallel mediators, and job satisfaction Daily work events. In contrast to Study 1, participants in this
and performance as outcomes. Organizational tenure and number study could list multiple events per day, which allows for an
of days were entered as covariates. Below, we report unstandard- independent assessment of positive and negative events. In each
ized coefficients (labeled B). daily survey, participants were asked to list up to six pleasant or
In this analysis, age was predictive of the mean of positive affect unpleasant work events they had experienced on that particular day
(B ⫽ .027, SE ⫽ .007, CI [.013, .040]), the mean and iSD of in an open-answer format (see Kuba & Scheibe, 2017). Examples
negative affect (B ⫽ ⫺.015, SE ⫽ .005, CI [⫺.025, ⫺.006]; and of reported events included interpersonal events (“conflict with a
B ⫽ ⫺.008, SE ⫽. 003, CI [⫺015, ⫺.002]), and the negative colleague,” “good conversation with colleagues during lunch
reactivity slope (B ⫽ ⫺.015, SE ⫽ .004, CI [⫺.024, ⫺.006]). break”), events related to the task or work environment (“many
The indirect effect of age on task performance via mean of positive emergency surgeries,” “good division of labor in close consulta-
affect (B ⫽ .006, SE ⫽ .002, CI [.002, .011]) and mean of negative tion with my colleagues”), and events related to personal issues
affect (B ⫽ .004, SE ⫽ .002, CI [.000, .009]) were significant. (“slept in and experienced time pressure”). On average, 1.66
However, the indirect effect of age on job satisfaction via affect events (SD ⫽ 1.60, range 0 – 6) were listed per day. Next, partic-
dynamics was not significant. These findings imply that older ipants rated each listed event as very negative, negative, neutral,
workers’ experience of more positive affect and less negative positive, or very positive, which we used to determine the valence
affect can contribute positively to their task performance. of events. To facilitate comparability with Study 1 (in which only
the most intense event on a given day was selected for further
Study 2 (German Hospital Employees) ratings), we only considered those events rated as either very
negative or very positive. We derived two event occurrence vari-
ables (positive vs. negative) where we compared days with at least
Method
one very negative or very positive event (coded as 1) from days
Participants and procedure. A sample of 126 health care without such events (coded 0). Across the 1128 daily reports, there
workers were recruited from four hospitals of one large hospital were 651 (57.7%) days with no events, 172 (15.2%) days with only
negative events, 242 (21.5%) days with only positive events, and Engagement Scale (UWES, Schaufeli, Bakker, & Salanova, 2006).
63 (5.6%) days with both negative and positive events. Items represented all three subscales (vigor, dedication, and ab-
Daily affect. Rather than measuring affective responses to sorption). A sample item is “I am proud of the work that I do.”
daily events as in Study 1, in this study affect was assessed Responses were given on a 7-point scale ranging from 1 (never) to
independent of events. At the beginning of each diary entry, 7 (always). Reliability was acceptable (␣ ⫽ .82).
participants indicated for each of eight affect terms on a 5-point
Likert scale (1 ⫽ never, 2 ⫽ rarely, 3 ⫽ sometimes, 4 ⫽ often, 5 ⫽
Results
very often) how often they had experienced these on that particular
day. Items included low-arousal positive states (LAP: calm, re- Preliminary analyses. Table 3 presents descriptive informa-
laxed), high-arousal positive states (HAP: joyous, enthusiastic), tion and intercorrelations for central study variables. The four
low-arousal negative states (LAN: down, sluggish), and high- types of affect were moderately intercorrelated both at the day
arousal negative states (HAN: upset, restless). Extending Study 1, level (absolute correlations ranging from .32 to .47; all ps ⬍ .001)
we were able to distinguish positive and negative affect in terms of and the person level (.30 to .63, all ps ⬍ .001). As expected,
arousal, a dimension that is receiving increased attention in re- positive affect (LAP and HAP) was negatively related to negative
search on emotional aging (e.g., Sands & Isaacowitz, 2017). A affect (LAN and HAN). Age was correlated with organizational
CFA performed with MPlus 7 showed that the fit of the four-factor tenure, r ⫽ .50, p ⬍ .001; both age and tenure were unrelated to
model distinguishing affect by valence and arousal (␹2 ⫽ 157; gender. The intraclass correlations indicate that 80% and 83%,
df ⫽ 32; CFI ⫽ .93; RMSEA ⫽ .06; ␻ ⫽ .67) outmatched the fit respectively, of the variance of negative and positive events, were
of a single-factor model (␹2 ⫽ 499; df ⫽ 40; CFI ⫽ .74; located at the day level. Moreover, between 47% and 63% of the
RMSEA ⫽ .59; ␻ ⫽ .04; Satorra-Bentler scaled ␹2 ⫽ 313; df ⫽ 8; variance in affect were located at the day level.
p ⬍ .001) and a two-factor model distinguishing positive and Age differences in daily affect and work events. We again
negative affect (␹2 ⫽ 354; df ⫽ 38; CFI ⫽ .82; RMSEA ⫽ .086; tested Hypotheses 1 to 3 and our research questions simultane-
␻ ⫽ .819; Satorra-Bentler scaled ␹2 ⫽ 161; df ⫽ 6; p ⬍ .001). The ously in one MSEM using MPlus 7. We predicted positive and
iSD was also calculated for each of the four types of affect. negative event occurrence by age. We predicted the four types of
Person-level variables. We assessed participants’ age (in affect (LAP, HAP, LAN, HAN) by positive and negative event
years), gender (0 ⫽ men, 1 ⫽ women), and organizational tenure occurrence, age, as well as the cross-level interaction between
(in years). We also assessed rank of position (0 ⫽ no leadership event occurrence and age. As covariates we included the number
function, 1 ⫽ leadership function), yet it was excluded in the of diary entry, weekday, and time of day at Level 1, as well as
following analyses because of too many missing cases. In the gender, tenure, and overall number of events at Level 2.
baseline survey, participants further rated their job satisfaction As shown in Table 4, age was unrelated to negative event
over the past four weeks on a 7-point scale ranging from 1 (very occurrence, yet it was positively related to positive event occur-
dissatisfied) to 7 (very satisfied). Previous validation work sup- rence (B ⫽ .006, SE ⫽ .003, p ⫽ .021). Thus, older participants
ports the use of single-item measures for job satisfaction (G. G. reported positive events more frequently than younger participants,
Fisher, Matthews, & Gibbons, 2016). Participants rated their work providing partial support for Hypothesis 3. Gender was positively
engagement using a shortened 3-item version of the Utrecht Work related to event occurrence, such that women reported both more
Table 3
Descriptive Statistics and Correlations Between Central Study Variables, Study 2 (German Hospital Employees)
Variable Mean (SD) 1 2 3 4 5 6 7 8 9 10 11 12 13 14

1. Age 46.45 (9.95) —
2. Gender — ⫺.11 —
3. Organizational tenure 13.06 (10.05) .50 ⫺.03 —
4. Job satisfaction 4.56 (1.57) .06 .03 ⫺.08 —
5. Work engagement 4.82 (1.09) .15 .04 .03 .45 —
Variables assessed at day-levela
6. Negative event(s) .21 (.24) .05 .22 .05 ⴚ.22 ⫺.04 —
7. Positive event(s) .27 (.28) .16 .18 .07 .13 .21 .20 —
8. LAP 3.35 (.63) .03 ⫺.08 ⫺.02 .12 .22 ⴚ.34 .11 —
9. HAP 2.83 (.68) ⫺.11 .07 ⫺.07 .41 .42 ⴚ.33 .32 .34 —
10. LAN 1.77 (.60) ⫺.08 ⫺.11 .08 ⴚ.28 ⴚ.30 .25 ⴚ.23 ⴚ.36 ⴚ.40 —
11. HAN 2.01 (.59) ⫺.03 .00 .12 ⴚ.27 ⫺.17 .44 .01 ⴚ.42 ⴚ.27 .56 —
12. iSD of LAP .60 (.23) ⫺.09 .21 .02 ⫺.09 ⫺.13 .36 .11 ⴚ.24 ⫺.16 .20 .28 —
13. iSD of HAP .55 (.24) .02 .14 ⫺.05 ⫺.09 ⫺.06 .01 .00 ⫺.03 .01 .15 .05 .39 —
14. iSD of LAN .51 (.29) ⫺.14 ⫺.03 .09 ⫺.17 ⴚ.21 .18 ⫺.10 ⫺.14 ⴚ.23 .45 .36 .37 .28 —
15. iSD of HAN .63 (.26) .11 .18 .11 ⫺.03 ⫺.06 .26 .07 ⫺.10 ⫺.13 .18 .28 .45 .37 .33
Note. N ⫽ 125. LAP ⫽ low-arousal positive affect; HAP ⫽ high-arousal positive affect; LAN ⫽ low-arousal negative affect; HAN ⫽ high-arousal
negative affect. iSD denotes the intraindividual standard deviation. Event occurrence was coded as 0 ⫽ no negative/positive event and 1 ⫽ one or more
negative/positive event(s). Gender was coded as 0 ⫽ men and 1 ⫽ women. Correlations displayed in bold are significant at p ⬍ .05.
a
Scores of day-level variables are aggregates of the daily ratings.
Table 4
Unstandardized Coefficients From Multilevel Structural Equation Modeling Analyses Predicting Affective Responses to Daily Work
Events, Study 2 (German Hospital Employees)
B (SE)
Measure Negative event(s) Positive event(s) LAP HAP LAN HAN
ⴱⴱⴱ ⴱⴱⴱ ⴱⴱⴱ ⴱⴱⴱ ⴱⴱⴱ
Intercept 0.178 (0.037) 0.212 (0.042) 3.442 (0.090) 2.838 (0.094) 1.907 (0.084) 2.083 (0.083)ⴱⴱⴱ
Level 1 (Day-level)
Negative event(s) — — ⫺0.496 (0.062)ⴱⴱⴱ ⫺0.353 (0.054)ⴱⴱⴱ 0.3842 (0.058)ⴱⴱⴱ 0.625 (0.078)ⴱⴱⴱ
Positive event(s) — — 0.235 (0.057)ⴱⴱⴱ 0.411 (0.059)ⴱⴱⴱ ⫺0.226 (0.050)ⴱⴱⴱ ⫺0.241 (0.058)ⴱⴱⴱ
Number of diary entry ⫺0.008 (0.004)ⴱ ⫺0.010 (0.004)ⴱ ⫺0.004 (0.006) ⫺0.018 (0.006)ⴱⴱ ⫺0.002 (0.006) ⫺0.020 (0.006)ⴱⴱ
Weekday 0.001 (0.007) 0.018 (0.008)ⴱ 0.006 (0.013) ⫺0.007 (0.012) ⫺0.013 (0.012) ⫺0.028 (0.013)ⴱ
Time of day 0.007 (0.003)ⴱ 0.007 (0.003)ⴱ ⫺0.002 (0.006) ⫺0.005 (0.005) 0.000 (0.005) ⫺0.002 (0.006)
Age 0.002 (0.002) 0.006 (0.003)ⴱ 0.003 (0.006) ⫺0.006 (0.007) ⫺0.005 (0.006) ⫺0.006 (0.006)
Gender 0.109 (0.040)ⴱⴱ 0.105 (0.046)ⴱ ⫺0.070 (0.107) 0.065 (0.114) ⫺0.213 (0.099)ⴱ ⫺0.006 (0.093)
Organizational tenure 0.001 (0.002) 0.000 (0.003) ⫺0.003 (0.006) ⫺0.003 (0.006) 0.008 (0.005) 0.009 (0.005)⫹
Number of negative event days — — ⫺0.637 (0.360)⫹ ⫺1.083 (0.362)ⴱⴱ 0.827 (0.337)ⴱ 0.782 (0.324)ⴱ
Number of positive event days — — 0.391 (0.306) 0.981 (0.315)ⴱⴱ ⫺0.484 (0.271)⫹ 0.031 (0.260)
Cross-level interaction — —
Age ⫻ Negative event(s) — — 0.011 (0.006)⫹ 0.004 (0.006) ⫺0.016 (0.006)ⴱ 0.006 (0.008)
Age ⫻ Positive event(s) — — ⫺0.003 (0.006) ⫺0.006 (0.006) 0.000 (0.005) 0.000 (0.006)
Note. Level 1 N ⫽ 1128; Level 2 N ⫽ 125. LAP ⫽ low-arousal positive affect; HAP ⫽ high-arousal positive affect; LAN ⫽ low-arousal negative affect;
HAN ⫽ high-arousal negative affect. Event occurrence was coded as 0 ⫽ no very negative/positive event and 1 ⫽ one or more very negative/positive
event(s). Weekday was coded 0 ⫽ Monday, 1 ⫽ Tuesday, 2 ⫽ Wednesday, etc. Gender was coded as 0 ⫽ male and 1 ⫽ female. All hypotheses were tested
simultaneously in one multilevel structural equation model; residual variances of the four affect variables were allowed to covary. The ⴚ2log likelihood
value is ⫺5243.15 (df ⫽ 107).
⫹
p ⬍ .10. ⴱ p ⬍ .05. ⴱⴱ p ⬍ .01. ⴱⴱⴱ p ⬍ .001.
positive events (B ⫽ .105, SE ⫽ .046, p ⫽ .022) but also more tively predicted iSD of LAN (B ⫽ ⫺.008, SE ⫽ .003, p ⫽ .012),
negative events (B ⫽ .109, SE ⫽ .040, p ⫽ .007) than men. but none of the other three iSD indicators (all ps ⬎ .10). Thus,
Positive events were more often reported toward the end of the there was only partial support for Hypothesis 2. When additionally
week (B ⫽ .018, SE ⫽ .008, p ⫽ .020), and both types of events accounting for mean affect levels, these results were robust. The
were more often reported when the survey was filled in at a later iSD of LAP and HAN were further predicted by gender (B ⫽ .096,
time of day (positive events: B ⫽ .007, SE ⫽ .003, p ⫽ .036; SE ⫽ .042; p ⫽ .024 and B ⫽ .112, SE ⫽ .049, p ⫽ .024),
negative events: B ⫽ .007, SE ⫽ .003, p ⫽ .022). The latter effect indicating stronger affect variability for women than men. The iSD
may result from the fact that people with longer workdays—which of LAN was further positively predicted by organizational tenure
could increase the chance that affective events occur— completed (B ⫽ .006, SE ⫽ .003, p ⫽ .035).
the diary entry later in the day. Indirect associations of age with work-related outcomes.
LAP and HAP were lower on days with negative events, relative The prior analyses showed that age was related to a higher fre-
to days without negative events, whereas LAN and HAN were quency of positive event days, lower LAN reactivity to negative
higher. At the same time, age moderated the relationship between events, and lower variability of LAN (iSD). To examine Hypoth-
negative event occurrence and LAN such that older participants esis 4 that age had indirect effects on work-related outcomes
experienced less of an increase in LAN than younger participants through these affective variables, we again conducted mediation
(B ⫽ ⫺.015, SE ⫽ .006, p ⫽ .012). No significant moderation analyses using PROCESS. We included all three potential medi-
effect appeared for the remaining three types of affect. Thus, there ators simultaneously, and included gender, organizational tenure,
was partial evidence that higher worker age is associated with and number of event days as covariates. As in Study 1, LAN
lower reactivity to negative events. The occurrence of positive reactivity to negative event scores were obtained by saving the
events was related to increased LAP and HAP and reduced LAN negative event ¡ LAN slope from a multilevel analysis performed
and HAN; there was no moderation effect of age on reactivity to with MPlus, in which we modeled the four types of affect as a
positive events. Hence, in response to our research question about function of negative and positive event occurrence and day-in-
affective reactivity, we found older workers to react less strongly survey.
to negative work events than younger workers on some dimensions The PROCESS analysis revealed positive effects of age on
of affect, but there were no age differences in reactivity to positive number of positive events (B ⫽ .008, SE ⫽ .003, CI [.002, .014]),
events. Weekday predicted levels of HAN negatively (B ⫽ ⫺.028, LAN reactivity to negative events (B ⫽ ⫺.005, SE ⫽ .002, CI
SE ⫽ .013, p ⫽ .028), suggesting improved well-being as the [⫺.008, ⫺.001]), and iSD of LAN (B ⫽ ⫺.008, SE ⫽ .003, CI
weekend nears. [⫺.015, ⫺.002]). Job satisfaction was not significantly predicted
Age differences in affect stability. Addressing Hypothesis 2 by either age, affective variables, or any covariates. Accordingly,
(age differences in affect stability), we regressed the iSD of each none of the indirect effects were significant. Work engagement
type of affect on age in separate analyses, accounting for gender, was negatively predicted by iSD of LAN (B ⫽ ⫺.739, SE ⫽ .370,
organizational tenure, and total number of diary days. Age nega- CI [⫺1.471, ⫺.007]), and at trend-level by number of positive
events (B ⫽ .669, SE ⫽ .365, CI [⫺.055, 1.392]), although the elicited. For positive events, we assessed four positive emotional
effects of LAN reactivity, age, and the covariates were all nonsig- responses, including serene, at ease (both representing LAP),
nificant. The indirect effects of age on work engagement via iSD elated, and delighted (both representing HAP). Emotion ratings
of LAN (B ⫽ .006, SE ⫽ .004, CI [.001, .017]) and number of were made on a scale from 1 (not at all) to 5 (very much). The
positive events (B ⫽ .005, SE ⫽ .003, CI [.001, .014]) were both average correlation between elated and delighted across events
significant. This suggests that older workers’ experience of more was .67, and the average correlation between serene and at ease
positive events and higher stability of LAN contributes positively was .52; therefore, we deemed it appropriate to combine these into
to their work engagement (though not to their job satisfaction). HAP and LAP subscales, respectively.
For negative events, we assessed four negative emotional re-
Study 3 (German Office Workers) sponses, including down, sluggish (both representing LAN), an-
noyed, and worried (both representing HAN). The average corre-
Method lation between down and sluggish across events was .43 and we
combined the two items into a LAN subscale. The correlation
Participants and procedure. The sample consisted of 121 between annoyed and worried was, however, low (mean r ⫽ .21
German office workers. The majority (n ⫽ 94) were recruited from across events, range from .02 to .50); therefore, we opted to
a municipal administrative center. The remaining 27 participants analyze the two items separately.
were recruited from the personal network of a research assistant. Person-level variables. As covariates, we assessed partici-
Seven participants did not provide data in the diary study (see pants’ age (in years), gender (0 ⫽ men, 1 ⫽ women), organiza-
below) and two participants reported no events across all diary tional tenure (in years), and supervisory function (0 ⫽ no super-
days; these were excluded from analysis, resulting in an effective visor, 1 ⫽ supervisor). As in this study, multiple events and
sample of 112 employees. Of the sample, 64% were female and associated emotional responses were rated in the same daily sur-
33% were in supervisor positions. Participants ranged in age from vey, these may be confounded by mood experienced in the morn-
20 to 64 years (M ⫽ 44.8, SD ⫽ 10.8) and worked for their ing of the working day (Rothbard & Wilk, 2011). To account for
company on average 18.7 years (SD ⫽ 11.8). this, we assessed morning level of affect each day (“How did you
After reaching agreement with the municipal administrative start into your workday this morning?” 1 ⫽ very exhausted to 5 ⫽
center on conducting the study, a link to a sign-up form was posted very rested) and mood (1 ⫽ very negative to 5 ⫽ very positive).
on the center’s intranet. Participants from outside the organization The two items were moderately to strongly positively correlated
were contacted directly via e-mail and were provided the link to across the three days (rs ⫽ .53, .45, and .59, all ps ⬍ .001) and
the sign-up form. After enlisting for the study and providing were therefore combined into a morning affect score. For reasons
informed consent, participants left their e-mail address and indi- of simplicity we further aggregated morning affect across days,
cated their work days. Participants received a personalized link via turning it into a person-level variable.
e-mail to a baseline survey (assessing demographic information,
In the baseline survey, participants also rated their job satisfac-
job satisfaction, work engagement, and further variables not in-
tion over the past month on a 7-point scale (1 ⫽ very dissatisfied,
cluded in this paper) and to daily surveys on the following work-
7 ⫽ very satisfied) and their work engagement using the nine-item
days (assessing that days’ experienced positive and negative work
UWES (Schaufeli et al., 2006) on a 7-point scale (1 ⫽ never, 7 ⫽
events and emotional responses to each event). Participants were
always).
instructed to complete at least three daily surveys at the end of
their workday or later at home. The daily links were sent at 11 a.m.
and were valid for 24 hr. In total, participants provided 358 valid Results
daily entries (M ⫽ 3.13, SD ⫽ 0.96, range 1 to 7) and reported a
total number of 966 events (M ⫽ 8.6, SD ⫽ 5.0, range 1 to 30). We Preliminary analyses. Table 5 presents descriptive informa-
recorded the day of the week (0 ⫽ Monday, 1 ⫽ Tuesday, 3 ⫽ tion and intercorrelations for central study variables. Note that
Wednesday, etc.) as a covariate for analyses. At the end of the positive affect was rated only for positive events, whereas negative
study, participants received personalized feedback on their emo- affect was rated only for negative events. Therefore, it is not
tion regulation style (this measure is not part of the current report) possible to correlate them at the event-level. Instead, we present
and took part in a raffle of two online vouchers each worth €100. intercorrelations for scores aggregated to the person-level. Inter-
The research procedure was approved by the Ethical Committee correlations between the five types of affect were all significant
Psychology at the University of Groningen. and positive (ranging from .20 to .63), except for the correlations
Measures. between LAP with Worried and LAN, which were nonsignificant.
Daily work events and affective responses. In each daily This suggests that workers who tended to react more strongly to
survey, participants were asked to list up to six pleasant and six positive events also tended to react more to negative events.
unpleasant work events they had experienced on that particular day The number of reported positive events was strongly correlated
(see Kuba & Scheibe, 2017). Pleasant and unpleasant work events with the number of reported negative events, r ⫽ .59, p ⬍ .05,
were assessed in separate blocks and event valence was counter- likely because event reports are attributable to the number of daily
balanced between daily entries. In total, participants reported 553 surveys completed. We therefore controlled for number of daily
positive and 413 negative daily events across all study days. surveys in the following analyses. Age was correlated with orga-
For each event, participants indicated the time period when the nizational tenure, r ⫽ .64, p ⬍ .05. Intraclass correlations indicate
event happened (choosing between 12 two-hour windows ranging that 80% and 83%, respectively, of the variance in affective
from 0:00 –2:00 to 22:00 –24:00) and which emotions these events responses were located at the event level.
116
Table 5
Descriptive Statistics and Correlations Between Central Study Variables, Study 3 (German Office Workers)
Variable Mean (SD) 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18

1. Age 44.76 (10.85) —
2. Gender — ⴚ.23 —
3. Org. tenure 18.66 (11.82) .64 ⫺.12 —
4. Supervisor — .17 ⴚ.23 .20 —
5. Job satisfaction 4.55 (1.69) .04 .01 ⫺.04 .02 —
6. Work engagement 4.19 (1.37) .08 .08 ⫺.12 .11 .54 —
7. Morning affect 3.32 (0.66) .15 .00 .03 ⫺.02 .31 .32 —
Variables assessed at event-level
8. No. positive events 4.94 (2.92) .14 .05 ⫺.02 ⫺.06 .18 .18 .05 —
9. No. negative events 3.69 (2.71) .07 ⫺.07 ⫺.01 .03 ⫺.12 ⫺.15 ⴚ.29 .59 —
10. LAP 3.31 (0.85) ⫺.02 ⫺.03 ⫺.03 ⫺.15 .22 .07 .20 .10 .15 —
11. HAP 3.15 (0.89) ⫺.03 .12 .03 ⫺.14 .23 .14 .07 .22 .19 .43 —
12. LAN 2.15 (0.92) .10 ⫺.11 .10 .02 ⴚ.27 ⫺.17 ⴚ.35 .08 .27 .11 .30 —
13. HAN: Annoyed 3.11 (0.91) .04 ⫺.09 ⫺.02 ⫺.14 .05 .08 .06 .14 .15 .43 .20 .24 —
14. HAN: Worried 2.15 (0.93) .14 .01 ⫺.01 ⫺.07 ⴚ.21 ⫺.09 ⴚ.42 .12 .29 .13 .22 .63 .30 —
15. iSD HAP .57 (0.33) .01 .10 .08 .00 ⫺.07 ⫺.11 ⴚ.24 ⫺.02 .02 ⴚ.30 ⫺.14 ⫺.01 ⴚ.26 .09 —
SCHEIBE, YEUNG, AND DOERWALD
16. iSD LAP 0.66 (0.38) ⫺.08 .17 ⫺.05 ⫺.05 ⫺.06 ⫺.02 ⴚ.23 .19 .31 .19 .02 .12 .22 .25 .31 —
17. iSD LAN 0.68 (0.041) ⫺.05 ⫺.08 .03 ⫺.14 ⫺.18 ⫺.21 ⫺.17 .04 .11 .02 .18 .43 .08 .13 ⫺.06 .04 —
18. iSD annoyed 1.09 (0.58) ⫺.17 .09 ⫺.15 ⫺.08 ⫺.13 ⫺.10 .00 ⫺.03 .09 .16 .16 .04 ⫺.10 .13 ⫺.11 .05 .13 —
19. iSD worried 0.90 (0.59) .00 .05 ⫺.04 .01 ⫺.17 ⫺.17 ⫺.15 .11 .18 .17 .14 .19 .01 .54 .13 .20 .32 .29
Note. N ⫽ 112. LAP ⫽ low-arousal positive affect; HAP ⫽ high-arousal positive affect; LAN ⫽ low-arousal negative affect; HAN ⫽ high-arousal negative affect. Event-level and day-level variables
were aggregated to the person-level. Gender was coded as 0 ⫽ male and 1 ⫽ female. Supervisor was coded 0 ⫽ no supervisory function and 1 ⫽ supervisory function. Correlations displayed in bold
are significant at p ⬍ .05.
Age differences in work events and affective responses. We (B ⫽ ⫺.440, SE ⫽ .133, p ⫽ .001), such that people who started
tested age differences in number of reported events with regression their workday in a more positive mood responded with less neg-
analysis. We accounted for gender, tenure, supervisory position, ative affect. Morning affect did not predict annoyed responding.
number of daily surveys, and (average) morning affect. In partial None of the other person-level covariates were related to the
support of Hypothesis 3 and consistent with Study 2, age was indicators of positive and negative affective responses.
positively related to the number of positive events (B ⫽ .076, SE ⫽ Age differences in affect stability. We regressed the iSD of
.034, p ⫽ .028) but unrelated to the number of negative events each type of affect on age in separate analyses, accounting for
(B ⫽ .049, SE ⫽ .031, p ⫽ .12). Of the covariates morning affect gender, organizational tenure, supervisory function, and total num-
was a significant negative predictor of reporting negative events ber of daily reports. Some participants only reported one negative
(B ⫽ ⫺1.365, SE ⫽ .386, p ⫽ .001); all other covariates yielded or positive event, thus their iSD could not be meaningfully com-
nonsignificant effects (all ps ⬎ .10). puted, reducing the effective sample to n ⫽ 80. Contrasting Hy-
We tested age differences in affective responses using multi- pothesis 2, age was unrelated to the iSD’s of all five types of affect
level modeling with MPlus, with events representing Level 1 and (all ps ⬎ .10). Findings were robust when additionally accounting
person-level differences modeled at Level 2. To account for our for mean affect levels.
data structure (positive responses were only assessed for positive Indirect associations of age with work-related outcomes.
events and negative responses were only assessed for negative The prior analyses showed that age was related to higher number
events), we modeled positive and negative affective responses in of positive events and higher worried responses to negative events.
two separate models. In both models, we accounted for time of day We therefore tested indirect effects of age on work-related out-
and weekday at level 1 and for gender, tenure, supervisory posi- comes via number of positive events and worried responding,
tion, number of daily entries, and average morning affect at Level using the PROCESS macro. We included both potential mediators
2. Results are shown in Table 6. in the same model, and accounted for gender, tenure, supervisory
As in the prior studies, age was unrelated to positive affective function, number of daily reports, and morning affect.
responses (both HAP and LAP). Among event-level covariates, The PROCESS analysis revealed positive predictive effects of
weekday predicted LAP positively (B ⫽ 0.055, SE ⫽ .027, p ⫽ age on number of positive events (B ⫽ .089, SE ⫽ .038, CI [.014,
.043), suggesting that LAP increased from Monday to Friday. .064]) and worried responses to negative events (B ⫽ ⫺.028, SE ⫽
Among the person-level covariates, morning affect predicted LAP .012, CI [.007, .049]). Job satisfaction was not significantly pre-
(B ⫽ .287, SE ⫽ .124, p ⫽ .021), indicating that people who dicted by number of positive events or worried responding, though
started their workday in a more positive mood responded with morning affect emerged as a significant covariate (B ⫽ .673, SE ⫽
more LAP to positive events. Morning affect did not predict HAP .284, CI [.108, 1.238]). None of the indirect effects on job satis-
responses. Regarding negative affective responses, age was unre- faction were significant. Work engagement was positively pre-
lated to being annoyed and to LAN. In contrast to the prior studies, dicted by number of positive events at trend level (B ⫽ 0.080,
age was positively related to the experience of being worried (B ⫽ SE ⫽ .045, CI [⫺0.010, .169]), but not by worried responding.
.029, SE ⫽ .010, p ⫽ .004). Time of day predicted LAN positively Organizational tenure emerged as a negative predictor (B ⫽
(B ⫽ .079, SE ⫽ .029, p ⫽ .006), suggesting that the LAN ⫺.041, SE ⫽ .015, CI [⫺.072, ⫺.011]), whereas morning affect
response increased across the workday. Morning affect predicted was a positive predictor (B ⫽ .564, SE ⫽ .223, CI [.109, 1.018]).
worried (B ⫽ ⫺.654, SE ⫽ .131, p ⫽ .001) and LAN responses None of the other covariates were significant. The indirect effect of
Table 6
Unstandardized Coefficients From Multilevel Structural Equation Modeling Analyses Predicting Affective Responses to Daily Work
Events, Study 3 (German Office Workers)
B (SE)
Measure LAP HAP LAN HAN: Annoyed HAN: Worried
ⴱⴱⴱ ⴱⴱⴱ ⴱⴱⴱ ⴱⴱⴱ
Intercept 3.379 (0.170) 3.121 (0.178) 2.185 (0.175) 3.112 (0.193) 2.039 (0.182)ⴱⴱⴱ
Level 1 (Event-level)
Time of day 0.003 (0.022) 0.032 (0.026) 0.079 (0.029)ⴱⴱ 0.016 (0.040) 0.046 (0.036)
Weekday 0.055 (0.027)ⴱ 0.021 (0.032) 0.046 (0.037) 0.079 (0.053) 0.010 (0.049)
Age 0.002 (0.010) 0.000 (0.010) 0.007 (0.010) 0.003 (0.010) 0.029 (0.010)ⴱⴱ
Gender ⫺0.129 (0.177) 0.145 (0.182) ⫺0.170 (0.177) ⫺0.114 (0.184) 0.238 (0.176)
Organizational tenure 0.001 (0.009) 0.007 (0.009) 0.006 (0.009) ⫺0.002 (0.010) ⫺0.008 (0.009)
Supervisory function ⫺0.321 (0.180)⫹ ⫺0.279 (0.185) ⫺0.098 (0.177) ⫺0.186 (0.183) ⫺0.123 (0.175)
Number of days 0.024 (0.087) 0.005 (0.092) ⫺0.117 (0.102) ⫺0.026 (0.113) ⫺0.051 (0.106)
Morning affect 0.287 (0.124)ⴱ 0.110 (0.130) ⫺0.440 (0.133)ⴱⴱⴱ 0.041 (0.134) ⫺0.654 (0.131)ⴱⴱⴱ
Note. Level 1 N ⫽ 557 for positive events and 419 for negative events; Level 2 N ⫽ 114; LAP ⫽ low-arousal positive affect; HAP ⫽ high-arousal positive
affect; LAN ⫽ low-arousal negative affect; HAN ⫽ high-arousal negative affect; Weekday was coded 0 ⫽ Monday, 1 ⫽ Tuesday, 2 ⫽ Wednesday, etc.
Gender was coded as 0 ⫽ male and 1 ⫽ female. Supervisory function was coded 0 ⫽ no supervisory function and 1 ⫽ supervisory function. Hypotheses
regarding positive affect (HAP, LAP) and negative affect (HAN, LAN) were tested in two separate multilevel models; residual variances of the affect
variables were allowed to covary. The ⫺2log likelihood value is ⫺1237.66 (df ⫽ 24) for positive affect and ⫺1628.89 (df ⫽ 39) for negative affect.
⫹
p ⬍ .10. ⴱ p ⬍ .05. ⴱⴱ p ⬍ .01. ⴱⴱⴱ p ⬍ .001.
age on work engagement via number of positive events was younger colleagues. However, whether or not older employees
significant (B ⫽ .007, SE ⫽ .006, CI [.001, .027]), but not via experience higher well-being than their younger colleagues
worried responding. This replicates Study 2 that older workers’ may be due to moderating factors. In our studies, it was the
experience of more positive events contributes positively to their Hong Kong Chinese sample of managers that showed more
work engagement (though not their job satisfaction). robust older-age advantages in affective well-being than the
German sample of hospital workers. Possible reasons lie either
in cultural or occupational characteristics. Chinese people em-
Discussion
phasize interpersonal harmony to a greater extent than their
In this paper, our goal was to comprehensively examine age- Western counterparts (Oyserman, Coon, & Kemmelmeier,
related differences in levels and dynamics of workplace affect by 2002), which motivates them to constantly control their nega-
investigating data from three diary studies with workers of differ- tive emotions once they occur. With the use of effective emo-
ent cultural and occupational backgrounds. We distinguished four tion regulation strategies, older Chinese workers could maintain
components of daily affective experience at work, mean levels, better affective well-being than their younger peers (Yeung et
variability of affect, occurrence of affective daily events, and the al., 2011). In addition, managerial jobs often come with high
reactivity to such events, and tested indirect effects of age—via job autonomy and status (Savery, 1988); both of these factors
mean levels and dynamics of affect— on individual differences in could trickle down to positively affect older employees’ daily
work outcomes. Overall, the patterns of age differences were not well-being. Larger samples representing a variety of cultural
entirely consistent across samples. Yet remarkably, when age and occupational backgrounds are needed to more fully test
effects were found, they nearly always pointed to an older-age moderating factors of age effects on workplace affect, to deter-
advantage in affective experience and the associated work out- mine when older employees tend to experience more positive
comes. and stable affect, and when they do not.
Notably, age differences in levels of affect found in the Hong
Kong Chinese sample appeared to account for higher self-rated
Findings on Levels and Variability of Affect
task performance with age. This is not surprising in light of
Based on theories and earlier findings from the life span earlier work showing that levels of affect predict job attitudes
development and organizational literatures, we had hypothe- (Spector, 1997) and performance (Kaplan et al., 2009). It sug-
sized higher age to be associated with higher levels of positive gests that better affective well-being may be a contributing
affect and lower levels of negative affect. We had further factor to more favorable job attitudes at higher worker age (Ng
predicted that affect is more stable (i.e., less variable) with age, & Feldman, 2010), highlighting a potential mechanism under-
which represents a correlate of mental well-being (Houben et lying age-related benefits in employee well-being and effec-
al., 2015). We tested these hypotheses in Studies 1 and 2 (the tiveness.
design of Study 3 precluded testing for age differences in levels
and variability of affect, as affect was assessed in relation to
Findings on Affective Events and Reactivity
multiple specific events per day). In Study 1 with Hong Kong
Chinese managers, hypotheses were largely confirmed using To derive a fuller picture of age differences in the affective
indicators of positive and negative affect. Older managers re- lives of workers, we further examined in all three samples the
ported higher affective well-being across sampling days, as occurrence and affective reactivity to daily events experienced
indicated by higher positive affect, lower negative affect, and at work. Whereas the bulk of research on aging and affective
higher stability of negative affect. In Study 2 with German reactivity so far has focused on negative events (aka daily
hospital employees, which allowed us to distinguished subtypes stressors), our studies are among the few that considered daily
of affect based on valence and arousal, we did not find age events of both negative and positive valence. Based on life span
differences in mean levels of affect. However, LAN as one theories of well-being and earlier research outside the work
subtype of negative affect was again more stable in older than setting, we had predicted that higher age is associated with
younger employees. In neither study did we find age differences reporting fewer negative and more positive daily work events.
in positive affect variability. We did not specify hypotheses on the relationship between
These findings are actually quite consistent with earlier re- worker age and affective reactivity to daily work events, as life
search. Regarding levels of affect, earlier studies in the work span theories make nuanced predictions based on distance to
setting also have found either higher affective well-being as a events (i.e., whether affect during or surrounding the negative
function of age (e.g., Dahling & Perez, 2010; Yeung et al., event is measured) and contextual factors (e.g., levels of cog-
2011) or no age differences in affect (Bindl et al., 2012; nitive functioning or chronic stress, Charles, 2010), and because
Tavares, 2016), whereas lower affective well-being with age is prior studies have yielded mixed results (Sliwinski & Scott,
rarely if ever found. Regarding variability of affect, prior stud- 2014). Among our studies, two assessed affect intensity during
ies in the work setting do not exist. Yet, our finding that age events, whereas the third one assessed affect frequency on the
predicts lower variability of (subtypes of) negative affect but day surrounding an event.
not positive affect dovetails with two earlier studies in the life In contrast to earlier studies (e.g., Nägel, Sonnentag, &
span development literature (Brose et al., 2013; Grühn et al., Kühnel, 2015) but consistently across our three samples, there
2013). were no age differences in the frequency of negative daily work
For the work setting, these findings imply that older employ- events. This is important as it indicates that any well-being
ees are unlikely to have lower affective well-being than their advantages we found cannot be attributed to the fact that older
participants have less stressful lives (see also Brose et al., events and how they are reflected in measures of affect intensity
2013). We did, however, find in two of the three studies that versus frequency.
older employees report more positive daily events than younger A finding that was robust across the three samples was the
adults. A similar age trend was reported in a longitudinal study lack of an association between age and affective reactivity to
with older men aged 50s and above (Aldwin et al., 2014); our positive events, no matter whether affect intensity during the
study thus extends these findings to samples of workers span- event or affect frequency on the day surrounding the event were
ning young and middle adulthood. Notably, daily events were measured. There is a dearth of research on aging and positive
assessed through self-report; therefore, it is unclear whether event reactivity (see Gunaydin et al., 2016 for an exception),
older employees actually experience more positive events. It is and our studies are the first test of this relationship in the work
equally plausible that they attend more than younger employees setting. The fact that the lack of age differences in positive
to positive events in their daily work or remember positive event reactivity was robust across the three samples is therefore
events more when filling in the daily survey. These predictions an important contribution. It fits well with the notion that older
are in fact consistent with SST (Carstensen, 2006) and research adults are motivated to maintain positive affect but are not
on the age-related positivity effect in attention and memory necessarily more motivated than younger adults to enhance it
(Reed et al., 2014). Whatever the cause of age differences, (Riediger et al., 2009). Thus, older employees’ “recipe” for
positive daily work events deserve more attention as they are higher work engagement may be the frequent experience of
experienced frequently during the workday and can potentially positive daily work events rather than drawing more affective
reduce resource drain resulting from daily negative events and benefits from positive events once they occur.
chronic stressors (Gross et al., 2011). Moreover, in both studies
that yielded age differences in positive event occurrence, these
Limitations and Future Directions
were in turn associated with person-level work engagement, and
there was evidence of indirect effects of age on work engage- The combination of three samples differing in cultural and
ment via positive events. This suggests that positive daily occupational background, and the variations in the measure-
events are a contributing factor to higher work engagement with ment of events and affect, allowed us to identify age effects that
age. were robust and generalizable across at least two studies, such
Regarding affective reactivity, there was no clear-cut pattern as the higher stability of negative affect with age, the higher
of age differences for negative daily events, which is in line frequency of positive events (which in turn predicted higher
with prior findings (Sliwinski & Scott, 2014). In Study 1, the work engagement), and lower reactivity to negative events.
older Chinese managers’ negative affect responses to negative Moreover, the lack of age differences in number of negative
work events appeared less pronounced than those of younger daily events and in affective reactivity to positive events was
managers (with no differences for positive affect). In other robust across all samples, which points at their generalizability.
words, they appeared to be more resilient and better able to However, our sample composition and design differences also
keep negative affect down when in the midst of negative work limit our ability to understand why some age effects differed
events. In Study 2, which measured affect frequency across the between studies, which could be attributable to cultural factors,
whole day, the older German hospital employees also reacted occupational characteristics, measurement differences, or a
less strongly to negative daily events than younger employees combination of the three. A fruitful direction therefore would be
in terms of a subtype of negative affect, LAN (with no differ- to conduct a large-scale study where cultural and occupational
ences in HAN, LAP, and HAP reactivity). In contrast, older background of participants are systematically varied, and where
German office workers in Study 3 showed higher worry inten- events and affect both during and surrounding daily events are
sity in response to negative daily events than younger workers systematically assessed. Such a study may also include more
(with no age differences in LAN and “annoyed” intensity). direct measures of cultural differences in emotion regulation
Again, these divergent findings may be attributable to occupa- and occupational demands, to test mechanisms underlying dif-
tional differences. It is possible that office workers in Study 3 ferent age-related patterns of affect. It is possible that cultures
have the lowest level of autonomy among the three samples; in which emotion regulation goals are more chronically acti-
they also may have the lowest levels of job demands. Challeng- vated, and occupations that stimulate workers to develop their
ing job demands are important as they may stimulate workers to emotional competencies over time, will facilitate age-related
improve their affect regulation skills and resilience as they get benefits in affective functioning. It is also possible that age-
older (Ilies, Aw, & Pluut, 2015). related advantages are more pronounced before and after affec-
Our findings further imply that the distinction between affect tive events than they are during affective events, as the SAVI
intensity during the event and affect frequency on the day model would suggest.
surrounding the event does not provide a clear-cut picture of Another limitation is that our studies assessed affect and
when age differences are most likely to emerge. A possible daily events once in the evening using self-report. End-of-day
reason could be that in all three studies—as in prior studies on reports of daily events may be biased by retrospective recall,
daily stress— both affect during the event and affect surround- either because people forget events or because their outcomes
ing daily events were measured once a day, at the end of or after or resolution make them no longer worth reporting (Sliwinski &
work time, which could lead to a blending of processes occur- Scott, 2014). We also have no information on the objective
ring during and surrounding events. More attention should be severity of experienced events and it is possible that the thresh-
devoted in future studies on daily affective events to disentan- old to categorize an event as severe enough to be worth report-
gling the different processes involved during and surrounding ing differs from one person to the other (although this is less
likely the case in Study 1 that used an event checklist). Atten- stress (Ong, Bergeman, Bisconti, & Wallace, 2006). These
tion to these issues is especially important when studying age psychological strengths enable older workers to continue to be
differences, as older adults were found to recall the past more successful at work, especially in service-oriented industries
positively than younger adults (Reed et al., 2014). Experience- which require emotional stability and resilience to interpersonal
sampling studies that measure event occurrence and affect stressors. Another valuable consequence of more positive work-
intensity and frequency multiple times per day, either at prede- place affect with age may be a stronger affective commitment to
termined intervals or event-based, offer a higher temporal res- the organization (Fisher, 2002). Workers with high affective
olution and precision in event reporting. Future studies would commitment have an emotional tie to the organization that
further benefit from more diverse measures, such as checklists motivates them to remain, not because it would be too costly to
of daily work events prerated for severity (for a parallel mea- leave, but because they have a sense of contribution and iden-
sure of life events, see Hobson & Delunas, 2001), objective tification with the organization (Meyer, Stanley, Herscovitch, &
indicators of task performance (see, e.g., Yeung & Fung, 2012), Topolnytsky, 2002). Our findings also provide insights for job
peer reports (Bruk-Lee & Spector, 2006), or physiological assignments in organizational contexts: higher work engage-
indicators of affect (Ilies, Dimotakis, & Watson, 2010). ment and performance may result from assigning older workers
Additionally, any cross-sectional study of aging at work is tasks that utilize their emotional competencies and knowledge
inherently limited in its possibility to disentangle age and (Kanfer & Ackerman, 2004).
generational effects. Age differences shown in the present stud-
ies may be due to aging, but also to generational effects—
Conclusion
historical trends toward lower well-being that would lead to
older workers appearing better off than younger workers; or Older workers comprise an increasing proportion of today’s
healthy worker effects—the tendency of low-functioning older workforce, and such trends often worry organizational practitio-
workers to retire (Fisher, Chaffee, & Sonnega, 2016). Impor- ners—many of whom associate aging with vulnerability and de-
tantly, cross-sectional and longitudinal findings on age and clining productivity. The current set of studies contributes to
affect may not converge. For example, one 10-year longitudinal increasing evidence that such negative views are unjustified. Older
study with a general-population sample yielded increases of workers are likely to function as well as or even better than
negative affect over time, while at any given point in time, older younger workers at the emotional level. Several of the indicators of
people reported lower negative affect than younger people affective experience emerged as mediators of positive age differ-
(Sliwinski et al., 2009). Measurement-burst designs that repeat ences in work engagement and self-rated performance. At the same
daily diary studies over time are needed to investigate aging time, it became clear that some of the positive age effects were
effects in affect dynamics while ruling out generational effects. present in only one or two of the three samples, and one finding in
The current set of studies tested indirect effects of age—via the sample with office workers pointed at stronger negative reac-
affect levels and dynamics— on self-reported work engage- tivity to daily events at higher age. Future research is needed that
ment, job satisfaction, and performance. Interestingly, indirect systematically investigates contextual moderators of age effects at
effects were found only for affect variables situated at the the individual, occupational, or cultural level.
person-level (e.g., mean levels of affect or number of events),
but less so for variables reflecting affect dynamics (e.g., event
reactivity). It is possible that affect dynamics would be more References
predictive if the work outcomes were assessed at the day level. Aldwin, C. M., Jeong, Y.-J., Igarashi, H., & Spiro, A. (2014). Do hassles
For example, a momentary increase in negative affect after a and uplifts change with age? Longitudinal findings from the VA nor-
negative event may predict subsequent momentary task perfor- mative aging study. Psychology and Aging, 29, 57–71. http://dx.doi.org/
mance but not necessarily average levels of performance (Beal 10.1037/a0035042
et al., 2005). Future research should measure both affect and Almeida, D. M., & Horn, M. C. (2004). Is daily life more stressful during
work outcomes at the day level. middle adulthood? In O. G. Brim, C. D. Ryff, & R. C. Kessler (Eds.),
How healthy are we? A national study of well-being at midlife (pp.
425– 451). Chicago, IL: University of Chicago Press.
Practical Implications Barsade, S. G., & Knight, A. P. (2015). Group affect. Annual Review of
Organizational Psychology and Organizational Behavior, 2, 21– 46.
Daily affective experiences play an important role in organi- http://dx.doi.org/10.1146/annurev-orgpsych-032414-111316
zational effectiveness, as they drive employees’ attitudes, de- Beal, D. J., & Ghandour, L. (2011). Stability, change, and the stability of
cisions, and behaviors (Brief & Weiss, 2002). The older-age change in daily workplace affect. Journal of Organizational Behavior,
advantages in affective well-being found in the present studies 32, 526 –546. http://dx.doi.org/10.1002/job.713
provide important implications to management regarding future Beal, D. J., Weiss, H. M., Barros, E., & MacDermid, S. M. (2005). An
recruitment, staff retention, and job assignment. Older employ- episodic process model of affective influences on performance. Journal
ees are often less prioritized in selection processes because of of Applied Psychology, 90, 1054 –1068. http://dx.doi.org/10.1037/0021-
9010.90.6.1054
their changes in physical and cognitive abilities. However, our
Bindl, U. K., Parker, S. K., Totterdell, P., & Hagger-Johnson, G. (2012).
studies reveal that relative to younger workers, older workers Fuel of the self-starter: How mood relates to proactive goal regulation.
are more emotionally stable and can quickly recover from Journal of Applied Psychology, 97, 134 –150. http://dx.doi.org/10.1037/
negative setbacks at work. They also show a greater tendency to a0024368
perceive their daily work experiences as positive, which will Birditt, K. S., Fingerman, K. L., & Almeida, D. M. (2005). Age differences
protect them from experiencing high levels of overall work in exposure and reactions to interpersonal tensions: A daily diary study.
Psychology and Aging, 20, 330 –340. http://dx.doi.org/10.1037/0882- Journal of Occupational Health Psychology, 21, 3–23. http://dx.doi.org/
7974.20.2.330 10.1037/a0039139
Bledow, R., Schmitt, A., Frese, M., & Kühnel, J. (2011). The affective shift Gross, S., Semmer, N. K., Meier, L. L., Kälin, W., Jacobshagen, N., &
model of work engagement. Journal of Applied Psychology, 96, 1246 – Tschan, F. (2011). The effect of positive events at work on after-work
1257. http://dx.doi.org/10.1037/a0024532 fatigue: They matter most in face of adversity. Journal of Applied
Bono, J. E., Glomb, T. M., Shen, W., Kim, E., & Koch, A. J. (2013). Psychology, 96, 654 – 664. http://dx.doi.org/10.1037/a0022992
Building positive resources: Effects of positive events and positive Grühn, D., Lumley, M. A., Diehl, M., & Labouvie-Vief, G. (2013).
reflection on work stress and health. Academy of Management Journal, Time-based indicators of emotional complexity: Interrelations and cor-
56, 1601–1627. http://dx.doi.org/10.5465/amj.2011.0272 relates. Emotion, 13, 226 –237. http://dx.doi.org/10.1037/a0030363
Brief, A. P., & Weiss, H. M. (2002). Organizational behavior: Affect in the Gunaydin, G., Selcuk, E., & Ong, A. D. (2016). Trait reappraisal predicts
workplace. Annual Review of Psychology, 53, 279 –307. http://dx.doi affective reactivity to daily positive and negative events. Frontiers in
.org/10.1146/annurev.psych.53.100901.135156 Psychology. Advance online publication. http://dx.doi.org/10.3389/
Brose, A., Scheibe, S., & Schmiedek, F. (2013). Life contexts make a fpsyg.2016.01000
difference: Emotional stability in younger and older adults. Psychology Hayes, A. F. (2012). PROCESS: A versatile computational tool for ob-
and Aging, 28, 148 –159. http://dx.doi.org/10.1037/a0030047 served variable mediation, moderation, and conditional process model-
Bruk-Lee, V., & Spector, P. E. (2006). The social stressors- ing [White paper]. Retrieved from http://www.afhayes.com/public/
counterproductive work behaviors link: Are conflicts with supervisors process2012.pdf
and coworkers the same? Journal of Occupational Health Psychology, Hobson, C. J., & Delunas, L. (2001). National norms and life-event
11, 145–156. http://dx.doi.org/10.1037/1076-8998.11.2.145 frequencies for the revised Social Readjustment Rating Scale. Interna-
Carstensen, L. L. (2006). The influence of a sense of time on human tional Journal of Stress Management, 8, 299 –314. http://dx.doi.org/10
development. Science, 312, 1913–1915. http://dx.doi.org/10.1126/ .1023/A:1017565632657
science.1127488 Houben, M., Van Den Noortgate, W., & Kuppens, P. (2015). The relation
Charles, S. T. (2010). Strength and vulnerability integration: A model of between short-term emotion dynamics and psychological well-being: A
emotional well-being across adulthood. Psychological Bulletin, 136, meta-analysis. Psychological Bulletin, 141, 901–930. http://dx.doi.org/
1068 –1091. http://dx.doi.org/10.1037/a0021232 10.1037/a0038822
Charles, S. T., & Almeida, D. M. (2007). Genetic and environmental Ilies, R., Aw, S. S. Y., & Pluut, H. (2015). Intraindividual models of
effects on daily life stressors: More evidence for greater variation in later employee well-being: What have we learned and where do we go from
life. Psychology and Aging, 22, 331–340. http://dx.doi.org/10.1037/ here? European Journal of Work and Organizational Psychology, 24,
0882-7974.22.2.331 827– 838. http://dx.doi.org/10.1080/1359432X.2015.1071422
Charles, S. T., Luong, G., Almeida, D. M., Ryff, C., Sturm, M., & Love, Ilies, R., Dimotakis, N., & Watson, D. (2010). Mood, blood pressure, and
G. (2010). Fewer ups and downs: Daily stressors mediate age differences heart rate at work: An experience-sampling study. Journal of Occupa-
in negative affect. The Journals of Gerontology Series B, Psychological tional Health Psychology, 15, 120 –130. http://dx.doi.org/10.1037/
Sciences and Social Sciences, 65, 279 –286. http://dx.doi.org/10.1093/ a0018350
geronb/gbq002 Kanfer, R., & Ackerman, P. L. (2004). Aging, adult development, and
Charles, S. T., Piazza, J. R., Mogle, J., Sliwinski, M. J., & Almeida, D. M. work motivation. The Academy of Management Review, 29, 440 – 458.
(2013). The wear and tear of daily stressors on mental health. Psycho- http://dx.doi.org/10.5465/amr.2004.13670969
logical Science, 24, 733–741. http://dx.doi.org/10.1177/0956797612 Kaplan, S., Bradley, J. C., Luchman, J. N., & Haynes, D. (2009). On the
462222 role of positive and negative affectivity in job performance: A meta-
Dahling, J. J., & Perez, L. A. (2010). Older worker, different actor? analytic investigation. Journal of Applied Psychology, 94, 162–176.
Linking age and emotional labor strategies. Personality and Individual http://dx.doi.org/10.1037/a0013115
Differences, 48, 574 –578. http://dx.doi.org/10.1016/j.paid.2009.12.009 Kooij, D. T. A. M., de Lange, A. H., Jansen, P. G. W., Kanfer, R., &
Dimotakis, N., Scott, B. A., & Koopman, J. (2011). An experience sam- Dikkers, J. S. E. (2011). Age and work-related motives: Results of a
pling investigation of workplace interactions, affective states, and em- meta-analysis. Journal of Organizational Behavior, 32, 197–225. http://
ployee well-being. Journal of Organizational Behavior, 32, 572–588. dx.doi.org/10.1002/job.665
http://dx.doi.org/10.1002/job.722 Kuba, K., & Scheibe, S. (2017). Let it be and keep on going! Acceptance
Doerwald, F., Scheibe, S., Zacher, H., & Van Yperen, N. W. (2016). and daily occupational well-being in relation to negative work events.
Emotional competencies across adulthood: State of knowledge and Journal of Occupational Health Psychology, 22, 59 –70. http://dx.doi
implications for the work context. Work, Aging and Retirement, 2, .org/10.1037/a0040149
159 –216. http://dx.doi.org/10.1093/workar/waw013 Kunin, T. (1955). The construction of a new type of attitude measure.
Eid, M., & Diener, E. (1999). Intraindividual variability in affect: Reli- Personnel Psychology, 8, 65–77. http://dx.doi.org/10.1111/j.1744-6570
ability, validity, and personality correlates. Journal of Personality and .1955.tb01189.x
Social Psychology, 76, 662– 676. http://dx.doi.org/10.1037/0022-3514 Kuppens, P., Oravecz, Z., & Tuerlinckx, F. (2010). Feelings change:
.76.4.662 Accounting for individual differences in the temporal dynamics of
Elizur, D., Borg, I., Hunt, R., & Beck, I. M. (1991). The structure of work affect. Journal of Personality and Social Psychology, 99, 1042–1060.
values: A cross cultural comparison. Journal of Organizational Behav- http://dx.doi.org/10.1037/a0020962
ior, 12, 21–38. http://dx.doi.org/10.1002/job.4030120103 Lazarus, R. S., & Folkman, S. (1984). Stress, appraisal and coping. New
Fisher, C. D. (2002). Antecedents and consequences of real-time affective York, NY: Springer.
reactions at work. Motivation and Emotion, 26, 3–30. http://dx.doi.org/ Lee, K., & Allen, N. J. (2002). Organizational citizenship behavior and
10.1023/A:1015190007468 workplace deviance: The role of affect and cognitions. Journal of
Fisher, G. G., Chaffee, D. S., & Sonnega, A. (2016). Retirement timing: A Applied Psychology, 87, 131–142. http://dx.doi.org/10.1037/0021-9010
review and recommendations for future research. Work, Aging and .87.1.131
Retirement, 2, 230 –261. http://dx.doi.org/10.1093/workar/waw001 Lin, S.-H., Wu, C.-H., Chen, M.-Y., & Chen, L. H. (2014). Why employees
Fisher, G. G., Matthews, R. A., & Gibbons, A. M. (2016). Developing and with higher challenging appraisals style are more affectively engaged at
investigating the use of single-item measures in organizational research. work? The role of challenging stressors: A moderated mediation model.
International Journal of Psychology, 49, 390 –396. http://dx.doi.org/10 Rothbard, N. P., & Wilk, S. L. (2011). Waking up on the right or wrong
.1002/ijop.12064 side of the bed: Start-of-workday mood, work events, employee affect,
Meyer, J. P., Stanley, D. J., Herscovitch, L., & Topolnytsky, L. (2002). and performance. Academy of Management Journal, 54, 959 –980.
Affective, continuance, and normative commitment to the organization: http://dx.doi.org/10.5465/amj.2007.0056
A meta-analysis of antecedents, correlates, and consequences. Journal of Sands, M., & Isaacowitz, D. M. (2017). Situation selection across adult-
Vocational Behavior, 61, 20 –52. http://dx.doi.org/10.1006/jvbe.2001 hood: The role of arousal. Cognition and Emotion, 31, 791–798.
.1842 Savery, L. K. (1988). Comparison of managerial and non-managerial
Mignonac, K., & Herrbach, O. (2004). Linking work events, affective employees’ desired and perceived motivators and job satisfaction levels.
states, and attitudes: An empirical study of managers’ emotions. Journal Leadership and Organization Development Journal, 9, 17–22. http://dx
of Business and Psychology, 19, 221–240. http://dx.doi.org/10.1007/ .doi.org/10.1108/eb053629
s10869-004-0549-3 Schaufeli, W. B., Bakker, A. B., & Salanova, M. (2006). The measurement
Morgan, E. S., & Scheibe, S. (2014). Reconciling cognitive decline and of work engagement with a short questionnaire: A cross-national study.
increased well-being with age: The role of increased emotion regulation Educational and Psychological Measurement, 66, 701–716. http://dx.doi
efficiency. In P. Verhaeghen & C. Hertzog (Eds.), Emotion, social .org/10.1177/0013164405282471
cognition, and everyday problem solving during adulthood (pp. 155– Scheibe, S., Wisse, B., & Schulz, A. (2017). Affect and emotion regulation
171). Oxford, UK: Oxford University Press. in aging workers. In N. A. Pachana (Ed.), Encyclopedia of geropsychol-
Muthén, L. K., & Muthén, B. O. (1998 –2015). Mplus user’s guide (7th ogy (pp. 33– 42). Singapore: Springer.
ed.). Los Angeles, CA: Author. Scheibe, S., & Zacher, H. (2013). A lifespan perspective on emotion
Nägel, I. J., Sonnentag, S., & Kühnel, J. (2015). Motives matter: A diary regulation, stress, and well-being in the workplace. In P. L. Perrewé, J.
study on the relationship between job stressors and exercise after work. Halbesleben, & C. C. Rosen (Eds.), Research in occupational stress and
International Journal of Stress Management, 22, 346 –371. http://dx.doi well being (Vol. 11, pp. 163–197). Bingley, UK: Emerald. http://dx.doi
.org/10.1037/a0039115 .org/10.1108/S1479-3555(2013)0000011010
Neupert, S. D., Almeida, D. M., & Charles, S. T. (2007). Age differences Sliwinski, M. J., Almeida, D. M., Smyth, J., & Stawski, R. S. (2009).
in reactivity to daily stressors: The role of personal control. The Journals Intraindividual change and variability in daily stress processes: Findings
from two measurement-burst diary studies. Psychology and Aging, 24,
of Gerontology Series B, Psychological Sciences and Social Sciences,
828 – 840. http://dx.doi.org/10.1037/a0017925
62, 216 –225. http://dx.doi.org/10.1093/geronb/62.4.P216
Sliwinski, M. J., & Scott, S. B. (2014). Boundary conditions for emotional
Ng, T. W. H., & Feldman, D. C. (2010). The relationships of age with job
well-being in aging: The importance of daily stress. In P. Verhaeghen &
attitudes: A meta-analysis. Personnel Psychology, 63, 677–718. http://
C. Hertzog (Eds.), Emotion, social cognition, and everyday problem
dx.doi.org/10.1111/j.1744-6570.2010.01184.x
solving during adulthood (pp. 128 –141). Oxford, UK: Oxford Univer-
Ong, A. D., Bergeman, C. S., Bisconti, T. L., & Wallace, K. A. (2006).
sity Press.
Psychological resilience, positive emotions, and successful adaptation to
Sliwinski, M. J., Smyth, J. M., Hofer, S. M., & Stawski, R. S. (2006).
stress in later life. Journal of Personality and Social Psychology, 91,
Intraindividual coupling of daily stress and cognition. Psychology and
730 –749. http://dx.doi.org/10.1037/0022-3514.91.4.730
Aging, 21, 545–557. http://dx.doi.org/10.1037/0882-7974.21.3.545
Oyserman, D., Coon, H. M., & Kemmelmeier, M. (2002). Rethinking
Spector, P. E. (1997). Job satisfaction: Application, assessment, causes,
individualism and collectivism: Evaluation of theoretical assumptions
and consequences (Vol. 3). Thousand Oaks, CA: Sage Publications.
and meta-analyses. Psychological Bulletin, 128, 3–72. http://dx.doi.org/
Spector, P. E., & Fox, S. (2002). An emotion-centered model of voluntary
10.1037//0033-2909.128.1.3
work behavior: Some parallels between counterproductive work behav-
Pelled, L. H., & Xin, K. R. (1999). Down and out: An investigation of the
ior and organizational citizenship behavior. Human Resource Manage-
relationship between mood and employee withdrawal behavior. Journal ment Review, 12, 269 –292. http://dx.doi.org/10.1016/S1053-4822(02)
of Management, 25, 875– 895. http://dx.doi.org/10.1177/014920639902 00049-9
500605 Stawski, R. S., Sliwinski, M. J., Almeida, D. M., & Smyth, J. M. (2008).
Piazza, J. R., Charles, S. T., & Almeida, D. M. (2007). Living with chronic Reported exposure and emotional reactivity to daily stressors: The roles
health conditions: Age differences in affective well-being. The Journals of adult age and global perceived stress. Psychology and Aging, 23,
of Gerontology Series B, Psychological Sciences and Social Sciences, 52– 61. http://dx.doi.org/10.1037/0882-7974.23.1.52
62, 313–321. http://dx.doi.org/10.1093/geronb/62.6.P313 Tavares, S. M. (2016). How does creativity at work influence employee’s
Reed, A. E., Chan, L., & Mikels, J. A. (2014). Meta-analysis of the positive affect at work? European Journal of Work and Organizational
age-related positivity effect: Age differences in preferences for positive Psychology, 25, 525–539. http://dx.doi.org/10.1080/1359432X.2016
over negative information. Psychology and Aging, 29, 1–15. http://dx .1186012
.doi.org/10.1037/a0035194 United States Department of Labor/Employment and Training Adminis-
Riediger, M., & Rauers, A. (2014). Do everyday affective experiences tration. (2018). OⴱNET 22.3 [Database]. Retrieved June 24, 2018 from
differ throughout adulthood? A review of ambulatory-assessment evi- http://www.onetcenter.org/database.html
dence. In P. Verhaeghen & C. Hertzog (Eds.), Emotion, social cognition, Wegge, J., van Dick, R., Fisher, G. K., West, M. A., & Dawson, J. F.
and everyday problem solving during adulthood (pp. 61–79). Oxford, (2006). A test of basic assumptions of affective events theory (AET) in
UK: Oxford University Press. call centre work. British Journal of Management, 17, 237–254. http://
Riediger, M., Schmiedek, F., Wagner, G. G., & Lindenberger, U. (2009). dx.doi.org/10.1111/j.1467-8551.2006.00489.x
Seeking pleasure and seeking pain: Differences in prohedonic and Weiss, H. M., & Cropanzano, R. (1996). Affective events theory: A
contra-hedonic motivation from adolescence to old age. Psychological theoretical discussion of the structure, causes and consequences of
Science, 20, 1529 –1535. http://dx.doi.org/10.1111/j.1467-9280.2009 affective experiences at work. In B. M. Staw & L. L. Cummings (Eds.),
.02473.x Research in organizational behavior (Vol. 18, pp. 1–74). Philadelphia,
Röcke, C., Li, S. C., & Smith, J. (2009). Intraindividual variability in PA: Elsevier Science/JAI Press.
positive and negative affect over 45 days: Do older adults fluctuate less World Values Survey Association. (2010 –2014). World Value Survey
than young adults? Psychology and Aging, 24, 863– 878. http://dx.doi Wave 6. Retrieved June 24, 2018, from http://www.worldvaluessurvey
.org/10.1037/a0016276 .org
Wrzus, C., Müller, V., Wagner, G. G., Lindenberger, U., & Riediger, M. Yeung, D. Y., Fung, H. H., & Chan, D. (2015). Managing conflict at work:
(2013). Affective and cardiovascular responding to unpleasant events Comparison between younger and older managerial employees. Inter-
from adolescence to old age: Complexity of events matters. Develop- national Journal of Conflict Management, 26, 342–364. http://dx.doi
mental Psychology, 49, 384 –397. http://dx.doi.org/10.1037/a0028325 .org/10.1108/IJCMA-06-2014-0044
Yeung, D. Y., & Fung, H. H. (2009). Aging and work: How do SOC Yeung, D. Y., Wong, C. K. M., & Lok, D. P. P. (2011). Emotion regulation
strategies contribute to job performance across adulthood? Psychology mediates age differences in emotions. Aging & Mental Health, 15,
and Aging, 24, 927–940. http://dx.doi.org/10.1037/a0017531 414 – 418. http://dx.doi.org/10.1080/13607863.2010.536136
Yeung, D. Y., & Fung, H. H. (2012). Impacts of suppression on emotional
responses and performance outcomes: An experience-sampling study in
younger and older workers. The Journals of Gerontology Series B, Received March 12, 2018
Psychological Sciences and Social Sciences, 67, 666 – 676. http://dx.doi Revision received July 20, 2018
.org/10.1093/geronb/gbr159 Accepted August 29, 2018 䡲

Retrieve 6

Uploaded by

Copyright:

Available Formats

Retrieve 6

Uploaded by

Document Information

Original Title

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

Retrieve 6

Uploaded by

Copyright:

Available Formats

Psychology and Aging

Age-Related Differences in Levels and Dynamics of Workplace Affect

Susanne Scheibe Dannii Y. Yeung

Levels and dynamics

Affective event reactivity

a time interval of 10 years (Sliwinski, Almeida, Smyth, & Stawski,

2009). The Present Studies

Variable Mean (SD) 1 2 3 4 5 6 7 8 9 10 11 12

Variables assessed at person-level

Variable Mean (SD) 1 2 3 4 5 6 7 8 9 10 11 12 13 14

Variables assessed at person-level

Variable Mean (SD) 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18

Variables assessed at person-level

You might also like