THE RELATIONSHIP BETWEEN EMOTIONAL LABOR, DIET, AND ALCOHOL
BEHAVIORS
By
Taylor K. Lauricella

A THESIS
Submitted to
Michigan State University
in partial fulfillment of the requirements
for the degree of
Psychology – Master of Arts
2017

ABSTRACT
THE RELATIONSHIP BETWEEN EMOTIONAL LABOR, DIET, AND ALCOHOL
BEHAVIORS
By
Taylor K. Lauricella
The present study examines the relationship between emotional labor strategies at work
and physical health behaviors, namely food and alcohol consumption, using a mood repairment
framework. An experience sampling methodology was utilized, and 38 servers/bartenders
completed three surveys a day for ten work days. Findings indicate that surface acting and
negative affect are positively related to snacking, high fat/high calorie food consumption, and
alcohol consumption after work. Furthermore, prevention focus moderated the indirect
relationships between surface acting, snacking, and alcohol consumption via negative affect.
Surface acting and negative affect, however, were not related to meal portion size or overall food
consumed. Results also suggest that deep acting is negatively related to high fat/high calorie food
consumption and snacking, but unrelated to meal size, total food consumption, and alcohol
consumption. Theoretical and practical implications, as well as future research directions are
discussed.
Keywords: emotional labor, alcohol, snacking, diet,

Copyright by
TAYLOR K. LAURICELLA
2017

ACKNOWLEDGMENTS

I would like to thank my advisor, Dr. Daisy Chang, for her developmental feedback and
support on this thesis, as well as her mentorship throughout my graduate school career. I am so
grateful for the time Daisy has invested in me over the past few years and am looking forward to
continuing working with her as I tackle the next phase of graduate school. Additionally, I would
like to thank my committee members, Dr. Richard DeShon and Dr. Brent Scott, for their helpful
comments and insightful ideas. I have no doubt that their input and unique perspectives made
this thesis a better study and made me a more thoughtful researcher.
In addition to my thesis committee, there were several faculty that helped me build
competencies along the way. I would like to thank Dr. Paul Bliese and Dr. James LeBreton for
teaching me multilevel analyses. Their guidance was fundamental in my understanding of
multilevel theory and analysis. I also would like to thank Dr. Betty Zhou for helping me navigate
the steep learning curve of MPlus.
Finally, I am grateful for my family for their constant love and encouragement
throughout my graduate school endeavors. I would not be where I am today without their
unwavering support.

iv

TABLE OF CONTENTS

LIST OF TABLES ............................................................................................................ vii
LIST OF FIGURES ......................................................................................................... viii
CHAPTER 1 ........................................................................................................................1
INTRODUCTION ...............................................................................................................1
CHAPTER 2 ........................................................................................................................4
THEORETICAL BACKGROUND .....................................................................................4
Emotional Labor ......................................................................................................4
Emotional Labor, Negative Affect, and Positive Affect ........................................10
Regulatory Focus ...................................................................................................12
Mood Repairment ..................................................................................................15
Reward-Based Stress Eating ..................................................................................16
Tension Reduction Theory .....................................................................................19
CHAPTER 3 ......................................................................................................................23
METHOD ..........................................................................................................................23
Participants.............................................................................................................23
Procedure ...............................................................................................................23
Measures ................................................................................................................25
Demographics (Baseline) ...........................................................................25
Regulatory focus (Baseline) .......................................................................25
Sleep quality and quantity (Time 1)...........................................................25
State positive and negative affect (Time 1 and Time 2) ............................26
Emotional labor (Time 2)...........................................................................26
Food quantity (Time 2 and Time 3) ...........................................................26
Food quality (Time 2 and Time 3) .............................................................27
Alcohol consumption (Time 3) ..................................................................28
Data Analysis .........................................................................................................28
Data structure .............................................................................................28
Analytic approach ......................................................................................29
CHAPTER 4 ......................................................................................................................30
RESULTS ..........................................................................................................................30
Supplementary analyses .........................................................................................34
CHAPTER 5 ......................................................................................................................35
DISCUSSION ....................................................................................................................35
Emotional Labor and State Affect .........................................................................35

!

v

Emotional Labor, Regulatory Focus, and State Affect ..........................................37
State Affect and Physical Health Behaviors ..........................................................38
Emotional Labor and Physical Health Behaviors ..................................................40
Practical Implications.............................................................................................42
Limitations .............................................................................................................42
Conclusion .............................................................................................................44
APPENDICES ...................................................................................................................45
APPENDIX A: Tables and Figures .......................................................................46
APPENDIX B: Consent Form ...............................................................................60
APPENDIX C: Baseline Survey ............................................................................61
APPENDIX D: Before Work Survey (Daily) ........................................................63
APPENDIX E: After Work Survey (Daily) ...........................................................64
APPENDIX F: Before Bed Survey (Daily) ...........................................................67
REFERENCES ..................................................................................................................69

!

vi

LIST OF TABLES

Table 1 Descriptive statistics, reliabilities, and intra-class correlation coefficients ..........46
Table 2 Intercorrelations between variables ......................................................................47
Table 3 Surface acting predicting negative affect with prevention focus moderation ......49
Table 4 Deep acting predicting positive affect with promotion focus moderation............51
Table 5 State affect predicting high fat/calorie food consumption ....................................52
Table 6 Moderated mediation for emotional labor, regulatory focus, state affect and
high fat/calorie consumption..............................................................................................53
Table 7 State affect predicting overeating .........................................................................54
Table 8 Moderated mediation for emotional labor, regulatory focus, state affect and
snacking .............................................................................................................................55
Table 9 Moderated mediation for emotional labor, regulatory focus, state affect and
meal size.............................................................................................................................56
Table 10 Moderated mediation for emotional labor, regulatory focus, state affect and
total food consumption ......................................................................................................57
Table 11 State affect predicting alcohol consumption .......................................................58
Table 12 Moderated mediation for emotional labor, regulatory focus, state affect and
alcohol consumption ..........................................................................................................59

!

vii

LIST OF FIGURES

Figure 1. Theoretical model ...............................................................................................22
Figure 2. Interaction of surface acting and prevention focus on negative affect ...............50

!

viii

CHAPTER 1
INTRODUCTION
In industrialized countries around the world, there has been a growth in the number of
jobs and the number of individuals employed in the service-providing sector, which include jobs
in healthcare, hospitality, and education, among others. In the United States, for example, as of
2011 there were over 112.8 million individuals employed in the service sector, and the US
Bureau of Labor Statistics predicts that this sector will continue to be the main source of
employment throughout 2018 (Bureau of Labor Statistics, 2013). This trend is also occurring
worldwide. As of 2012, over 40% of employed individuals work in the service sector in a
number of countries, including Australia, Canada, Japan, Korea, New Zealand, Netherlands, and
England (Bureau of Labor Statistics, 2013).
Meanwhile, there is also a growing emphasis on physical health and well-being,
particularly within the United States. With the implementation of new healthcare reforms, more
individuals are gaining access to health care services. In addition, government programs are
emphasizing the importance of preventive health. The Center for Disease Control and Prevention
introduced a national prevention strategy to transform America into a healthier nation. The
strategy identified a number of evidence-based practices that reduce the likelihood of major
illness and preventable death, which includes the consumption of a healthy diet (CDC, 2014).
Therefore, this study aims to address both of these trends by investigating whether the stress
associated with service jobs is associated with physical health behaviors, namely diet and alcohol
use.
A large body of research has been dedicated to the study of emotional labor over the past
three decades. Emotional labor refers to the process of regulating expressions, feelings, and

!

1

affect in order to meet organizational goals (Grandey, 2000; Hochschild, 1983) and is a
characteristic of jobs in the service-providing sector. Prior research has outlined two mechanisms
by which emotional labor occurs: surface acting and deep acting. When engaging in surface
acting, employees fake an emotional display without changing their underlying feelings, whereas
while deep acting, individuals try to alter their internal emotions to align with the expressions
they show (Grandey, 2000; Hochschild, 1983). Emotional labor has largely been investigated in
the context of organizational outcomes, such as turnover (Chau, Dahling, Levy, & Diefendorff,
2009; Goodwin, Groth, & Frenkel, 2011), performance (Grandey, 2003; Hulsheger, Lang, &
Maier, 2010), and employee psychological well-being (Goodwin et al., 2011; Holman, Chissick,
& Totterdell, 2002, Hulsheger et al., 2010; Morris & Feldman, 1997). However, far less research
has investigated the relationship between emotional labor and physical health behaviors.
Therefore in this study, I will examine whether there is an association between practicing
emotional labor and engaging in unhealthy behaviors, namely alcohol use and unhealthy food
consumption.
This study aims to contribute to the literature in a number of ways. First, this study uses
new outcome variables (i.e. food behaviors and alcohol consumption) that have yet to be studied
within the emotional labor framework. These behaviors are important to investigate as they can
directly and negatively impact employees’ physical health. Furthermore, this aligns with
Grandey and Gabriel’s (2015) call for more behavioral indicators of well-being to be
investigated as outcomes of emotional labor, other than burnout and job satisfaction. Doing so
expands the nomological network of emotional labor, and identifies important pathways through
which emotional labor may relate to workers’ health. Second, this study contributes to the worklife literature by investigating the spillover effects of work on health behaviors that occur in the

!

2

non-work domain. Lastly, this study adds to the growing emotional labor literature that utilizes
experience-sampling methodology (ESM) in order to capture, both, within- and between-person
variability. The use of ESM is particularly important within this study because in addition to the
intra-individual differences in emotional labor, diet and alcohol behaviors are also known to
fluctuate within-person (Armeli, Carney, Tennen, Affleck, & O’Neil, 2000; Mohr et al., 2005;
Steptoe, Lipsey, & Wardle, 1998). Therefore, using an ESM design will allow me to examine
how these phenomena fluctuate using a corresponding measurement frequency.

!

3

CHAPTER 2
THEORETICAL BACKGROUND
Emotional Labor
In order for service-providing organizations to both survive and thrive in the competitive
market that exists today, it is imperative that they deliver quality service, in addition to quality
products (Parasuraman, Zeithaml, & Berry, 1985). For this reason, among others, many
organizations choose to implement affective display rules which, either explicitly or implicitly,
guide and direct employees to display a particular emotion at any given time (Grandey, 2000;
Hochschild, 1983; Rafaeli & Sutton, 1987). The implementation of affective display rules has
many benefits for the organization. Prior research shows that display rules impact an
organization’s immediate gains (i.e. higher sales), encore gains (i.e. customer loyalty and repeat
visits; Rafaeli & Sutton, 1987), and influences overall service quality (Pugh, 2001).
There are several types of affective display rules, which vary greatly based on the nature
of the service provided and the job at hand. One type of display rule directs employees to solely
present positive emotions and to suppress all negative emotions (Grandey, 2000).This is known
as positive affective display rules and are often seen within the hospitality and retail industries.
For example, a waitress is told to always approach her table with a smile and a friendly greeting.
Meanwhile, another type of display rule guides employees to suppress positive emotions and
only display negative emotions (Grandey, 2000). For example, a bouncer is expected to act stern
while checking IDs at a bar in order to act as a rule enforcer. Lastly, display rules may encourage
employees to suppress all emotions altogether (Grandey, 2000). Examples of these jobs include
judges and therapists, whose position is to remain neutral.

!

4

Despite the implementation of display rules, employees naturally experience a widerange of emotions throughout any given day. Furthermore, these emotions do not always
coincide with the affective display rules. Therefore, when employees are at work, they have to
manage their own feelings and expressions in order to conform to the display rules. The process
by which this occurs is known as emotional labor.
Over three decades ago, Hochschild (1983) first introduced the concept of emotional
labor. She defined it as, “…require[ing] one to induce or suppress feeling in order to sustain the
outward countenance that produces the proper state of mind in others…This kind of labor calls
for coordination of mind and feeling” (p. 7). In her book, The Managed Heart, Hochschild
differentiates between managing emotions in an individual’s personal life and doing so within
the context of the workplace, where emotions become a “marketplace commodity” (Morris &
Feldman, 1997, p. 988). To be considered emotional labor, there are three requirements. First,
there must be either face-to-face or voice-to-voice interactions with customers. Second, the
employee must be trying to produce an emotional state in the customers as part of the interaction.
Examples of emotional states include happiness and gratitude. Third, the employer must have
control over the emotions and emotional activities of the employees (Hochschild, 1983).
There are two strategies by which employees engage in emotional labor: surface acting
and deep acting. Surface acting refers to a mechanism by which employees manage their
expressions (Grandey, 2000). When employees engage in surface acting, they avoid displaying
their true feelings by either suppressing, amplifying, or faking an emotion (e.g. “putting on a
happy face”; Grandey, 2000; Hochschild, 1983; Hulscheger & Schewe, 2011). While engaging
in surface acting, employees tend to experience emotional dissonance, due to the misalignment
between the emotions they are expressing and those they are feeling (Grandey, 2000;

!

5

Hochschild, 1983). Another consequence of surface acting is emotional suppression. The
combination of both emotional suppression and emotional dissonance leads to the experience of
a “discordant emotional state” (Mesmer-Magnus, DeChurch, & Wax, 2012).
Deep acting, on the other hand, is a mechanism through which employees manage their
feelings (Grandey, 2000). When employees engage in deep acting, they strive to experience the
desired emotions outlined by the affective display rules (e.g. “acting in good faith;” Grandey,
2000; Hochschild, 1983). This alignment of feelings and expressions leads to congruent
emotional states (Mesmer-Magnus et al., 2012). Despite the alignment between feelings and
expressions, deep acting is still a form of labor. Employees still have to exert a degree of effort
in order to genuinely change their feelings and act in an organizationally appropriate manner
(Morris & Feldman, 1997).
Emotional labor has been widely studied over the past thirty years in relation to a number
of organizational outcomes. Prior evidence demonstrates that the strategy chosen to carry out
emotional labor differentially relates to a number of job attitudes. For example, Chau et al.
(2009) found a positive relationship between surface acting and turnover intentions.
Furthermore, turnover intentions were predictive of actual turnover behavior. Deep acting, on the
other hand, was negatively related to turnover intentions, and thus, negatively related to actual
turnover behavior (Chau et al., 2009; Goodwin et al., 2011). Prior research also shows that
surface acting is negatively related to job satisfaction (Grandey, 2003; Judge, Wolf, & Hurst,
2009, Hulsheger & Schewe, 2011) and organizational commitment (Hulsheger & Schewe, 2011),
whereas deep acting is unrelated to both job satisfaction (Judge et al., 2009; Hulsheger &
Schewe, 2011) and organizational commitment (Hulsheger & Schewe, 2011).

!

6

Emotional labor also plays a role in job performance, although prior literature has found
mixed support. Some studies show that deep acting is positively related to performance
(Grandey, 2003; Hulsheger et al., 2010). However, in a recent meta-analysis by Hulsheger and
Schewe (2011), results indicated that there was no positive relationship between deep acting and
task performance. Deep acting was, however, positively related to emotional performance and
customer satisfaction. Furthermore, Hulsheger, Lang, Schewe, and Zilstra (2015) found that deep
acting was positively related to customer tips. Surface acting, on the other hand, was found to be
both negatively related (e.g., Grandey, 2003) and unrelated (e.g., Totterdell & Holman, 2003) to
job performance. Meta-analytic results show a small negative relationship between surface acting
and task performance (ρ = -.114), emotional performance (ρ = -.140), and customer satisfaction
(ρ = -.048). However, credibility intervals for task performance and emotional performance
include zero, indicating that between-study moderators exist to alter the magnitude of these
relationships (Hulsheger & Schewe, 2011).
Emotional labor has also been investigated within the context of employee psychological
well-being. Initially, Hochschild (1983) suggested that all emotional labor, regardless of strategy,
is harmful to well-being because “…the worker can become estranged or alienated from an
aspect of self – either the body or the margins of the soul” (p. 7). Since Hochshchild’s initial
work, there has been mixed support for her theory that all emotional labor is bad. Morris and
Feldman (1997) found a significant positive relationship between emotional dissonance, which is
experienced during surface acting, and emotional exhaustion. Furthermore, in their metaanalysis, Hulsheger and Schewe (2011) found that surface acting was significantly, positively
related to emotional exhaustion, depersonalization, and psychological strain. Therefore, the
evidence suggests that surface acting, overall, is harmful for employees’ well-being. Meanwhile,

!

7

the evidence for the relationship between deep acting and well-being is inconsistent. Prior
research has found a positive relationship between deep acting and stress-related outcomes (i.e.
burnout, emotional exhaustion; Goodwin et al., 2011; Holman et al., 2002), as well as no
relationship between deep acting and these well-being indicators (Hulsheger et al., 2010;
Wagner, Barnes, & Scott, 2014). Meta-analytics results showed that the relationship between
deep acting, emotional exhaustion, and psychological strain were approximately zero and the
credibility intervals included zero, suggesting that between-study moderators exist to explain the
relationship between deep acting and strains (Hulsheger & Schewe, 2011). Therefore, the
relationship between deep acting and employee psychological well-being may be more complex
than a bivariate relationship, and it is important to investigate boundary conditions that may alter
the nature of the relationship.
Far less research, however, has investigated the relationship between emotional labor and
employee physical health and/or behaviors that impact physical health. In their meta-analysis of
95 articles, Hulscheger and Schewe (2011) only found six studies that investigated emotional
labor and psychosomatic complaints. The meta-analytic estimates showed a significant, positive
relationship between surface acting and psychosomatic complaints (ρ =.435), and a small, but
positive relationship between deep acting and psychosomatic complaints (ρ = .175).
Additionally, two studies investigated emotional labor and physical health, although not
specifically surface and deep acting. Schaubroeck and Jones (2010) found that the demand to
express positive emotions at work, in general, was positively related to physical symptoms of
illness. This relationship was stronger among those who had low identification with their
organization. In their experimental study Hopp, Rorhmann, Zapf, and Hodapp (2010) found that
in conditions where emotional dissonance was high (i.e. positive display rules and rude

!

8

customers), there was an increase in blood pressure and heart rate, suggesting an aversive
physiological state. Lastly, in their experience-sampling study, Wagner et al. (2014) investigated
the spillover effect of emotional labor on a number of outcomes, including insomnia. Their
results indicated a significant, positive relationship between work-day surface acting and
insomnia at night. The relationship between deep acting and insomnia was not significant.
The lack of research investigating physical health and behaviors that influence health is a
major limitation of the emotional labor literature. In their recent review, Grandey and Gabriel
(2015) call for the use of more well-being outcomes. They argue that the majority of current
well-being outcomes are “work-centric” and “…ignores how emotional labor may affect the
whole person” (p. 337). Therefore, this study aims to start filling in the gaps, by assessing
behaviors that have an impact on physical health as an indicator of employee well-being. In
addition, this study aims to investigate spillover effects of emotional labor onto non-work
behaviors. To date, few studies have examined possible spillover effects of emotional labor. For
example, Krannitz, Grandey, Liu, and Almeida (2015) found that hotel managers’ surface acting
was directly related to their marital partner wanting them to quit and indirectly related to marital
partners’ perceptions of work-family conflict. In addition, Wagner et al. (2014) found bus
drivers’ surface acting was positively related to feelings of work-life conflict, emotional
exhaustion experienced at home, and insomnia. Thus, this study will provide novel insight
regarding the possible spillover effects of emotional labor, while simultaneously contributing to
the work-life literature.
Furthermore, this study adds to the growing literature that investigates emotional labor
over time. Traditionally, emotional labor research was conducted with cross-sectional studies.
However, Judge et al. (2009) found that there was significant within-person variance as well as

!

9

between-person variance in emotional labor strategies utilized. More recently, Gabriel, Daniels,
Diefendorff, and Greguras (2015) identified five latent profiles for emotional labor actors. Their
analyses revealed that individuals vary in both the amount of acting they engage in, as well as the
type of acting they engage in (i.e. surface acting, deep acting, combination). For example,
regulators are individuals who practice a high amount of both surface and deep acting, nonactors practice a low amount of both surface and deep acting, and deep actors practice a high
amount of deep acting and a low amount of surface acting. These findings show that individuals
can vary with the amount of emotional labor they engage in and can switch between strategies
during and across shifts. In addition, Gabriel and Diefendorff (2015) found that felt emotions and
emotional regulation varied within a single interaction with a customer. These findings suggest
that employees switch between emotional labor strategies throughout the course of a single
interaction. Thus, it is only appropriate to use a within-person design to study emotional labor in
order to capture this variation.
In light of these findings, more emotional labor research has been utilizing the ESM
design to account for intra-individual differences (e.g., Judge et al., 2009; Scott & Barnes, 2011;
Wagner et al., 2013). In this study, the use of ESM is particularly important. In addition to the
predictor variables (i.e. emotional labor), the outcome variables being investigated (i.e. food and
alcohol consumption) also tend to vary on a daily basis within and across individuals (Armeli et
al., 2000; Mohr et al., 2005; Steptoe et al., 1998). Therefore, using an ESM design will allow me
to capture both the within- and between-person variance in the predictor and outcome variables.
Emotional Labor, Negative Affect, and Positive Affect
As aforementioned, employees experience a wide-variety of emotions throughout any
given day, which may not align with the affective display rules provided at work. As a result,

!

10

employees may choose to fake the appropriate affective display via surface acting or to modify
their underlying affective state via deep acting (Hochschild, 1983). As such, affect is both a
precursor to and consequence of the process of emotional labor and regulation (Judge et al.,
2009; Scott & Barnes, 2011; Totterdell & Holman, 2003).
Drawing on Gross’s (1998) emotional regulation taxonomy, surface acting is a form of
response-focused emotional regulation meaning individuals respond to a particular stimulus after
an emotion has already been developed. As a result, they have to change their facial, bodily, or
affective expressions after the fact. Research on response-focused emotional regulation has
found that changing an affective expression does not change the underlying affective experience
(John & Gross, 2004) and rather, attempting to suppress negative emotions can amplify negative
affect (Wegner, 1994). When individuals engage in surface acting, they avoid changing their
underlying affective state that is mismatched from the affective display rules. Rather, they
continue experiencing the affective state while demonstrating a different affective expression.
Therefore, when positive display rules are in place, surface acting is likely to be positively
related to state negative affect.
Hypothesis 1a: Daily surface acting will be positively related to state negative affect.
With respect to Gross’s (1998) emotional regulation taxonomy, deep acting is an example
of antecedent-focused emotional regulation. That is, emotional regulation occurs before an
emotion develops and is experienced. The aim of antecedent-focused regulation is to change the
perception of a situation in order to change the subsequent emotional response (Gross, 1998).
When individuals engage in deep acting, they aim to align their affective display with their
underlying affective state. When affective display rules are positive, engaging in deep acting
should be positively related to state positive affect.

!

11

Hypothesis 1b: Daily deep acting will be positively related to state positive affect.
Regulatory Focus
Regulatory focus theory is a self-regulation theory that describes the nature by which
individuals reach desired end-states, or goals. The underlying assumption of regulatory focus
theory is that individuals will self-regulate differently when pursuing fundamentally different
needs, such as those for nourishment and those for security. Regulatory focus theory suggests
that nourishment-related and security-related self-regulation differ in their regulatory focus.
Nourishment-related regulation activates a promotion focus, whereas security-related regulation
activates a prevention focus (Higgins, 1997).
Promotion focus is an approach orientation, where individuals, “…attain a desired endstate by approaching matches…to that end-state” (Higgins, 1997, p.1284). Individuals high in
promotion focus are primarily concerned with advancement, growth, and accomplishment. They
constantly strive to improve themselves and gain growth experiences, which ultimately allow
them to approach and attain their ideal goals. High promotion-focused individuals are most
sensitive to the presence or absence of positive outcomes, or gains. They typically experience
emotions ranging from cheerfulness to dejection, based on whether they are able to attain their
ideal, desired end-state (Higgins, 1997).
Prevention focus is an avoidant orientation, where individuals, “…attain a desired endstate by…avoiding mismatches to that end-state” (Higgins, 1997, p. 1284). Individuals high in
prevention focus are primarily interested in fulfilling their responsibilities and are concerned
with safety and security. High prevention-focused individuals are sensitive to the presence or
absence of negative outcomes, or losses. They tend to experience agitation-like emotions when

!

12

something impedes their ability to fulfill their duties. In general, these individuals experience
emotions ranging from quiescence to agitation (Higgins, 1997).
Within an organization, the affective display rules are the desired end-states for
employees to meet. Therefore, an employee’s regulatory focus may influence the emotional labor
strategy he/she uses in order to meet the display rules provided. Dahling and Johnson (2010)
conducted research investigating the relationship between regulatory focus and emotional labor
strategies. They found that trait promotion focus was positively related to deep acting, whereas
trait prevention focus was positively related to surface acting, after controlling for display rule
perceptions and trait affect. Dahling and Johnson also found that state regulatory focus, activated
by a writing prime, affected emotional labor strategy selection such that state promotion focus
led to the selection of deep acting and state prevention focus led to the selection of surface
acting, during a hypothetical service encounter.
In this study, I explore trait regulatory focus as the individual difference that moderates
whether employees experience positive or negative state affect after engaging in emotional labor.
First, prior research shows that there is a positive relationship between surface acting and
prevention focus, and between deep acting and promotion focus (Dahling & Johnson, 2010).
Although Dahling and Johnson’s study found positive associations between emotional regulation
strategies and regulatory focus, this does not preclude other types of relationships between the
two to exist. Second, the emotional spectrum associated with promotion and prevention focus
map onto affective states. That is, individuals with high vs. low levels of a particular regulatory
foci will experience different intensity of positive versus negative emotions in response to their
environment. Finally, prior research has found that regulatory focus can moderate the effects of
individuals’ perceptions of the environment. For example, regulatory focus can impact how

!

13

individuals’ cognitively and affectively appraise others (Pham & Avnet, 2009). Therefore,
regulatory focus is likely to relate to how individuals perceive the experience of emotional labor.
More specifically, I expect that prevention focus will moderate the relationship between
surface acting and state negative affect. When employees engage in surface acting, they do not
express the emotions they internally feel. Rather they fake a particular expression, leading to
feelings of inauthenticity and emotional dissonance (Grandey, 2000). Employees who surface act
are likely to experience negative affect because they suppress their true feelings, which
subsequently leads to the experience of emotional dissonance. This experience may be
heightened for high prevention-focused individuals. High prevention-focused individuals are
fixated on meeting their duties. Within the context of emotional labor, high prevention-focused
individuals likely see affective display rules as obligations they need to meet, rather than
something that they should internally align to. This emphasis on duties and obligations may
cause high prevention-focused individuals to feel even more detached and experience more
dissonance when engaging in surface acting because there is less internalization of the process.
Rather, they are only acting to meet an external requirement. In addition, surface acting occurs
when employees avoid displaying their true feelings, and high prevention-focused individuals are
more sensitive to avoidance-related strategies. Furthermore, the emotional spectrum associated
with prevention focus may be another factor. High prevention-focused individuals are already
more susceptible to experiencing agitation-related emotions (Higgins, 1997), which overlap with
negative affect. Therefore, these individuals may be more sensitive to negative affect while
engaging in surface acting.
Hypothesis 2: The positive relationship between daily surface acting and state negative
affect will be stronger for those with high versus low prevention focus.

!

14

Individuals high in promotion focus, however, aim to approach and achieve their desired
end states. These individuals are excited to gain growth opportunities, which enhance their wellbeing (Higgins, 1997). When individuals engage in deep acting, they strive to genuinely
experience the emotion that they are expressing. Employees who engage in deep acting are likely
to experience state positive affect because they internalize the positive feelings they are
expressing and aim to develop meaningful, authentic experiences with their customers. I expect
this experience will be heightened for high promotion-focused individuals. High promotionfocused individuals will view deep acting as more of an opportunity to gain personal growth and
mastery in their jobs. These individuals will also seek more meaningful connections with
customers than those low in promotion focus. In addition, deep acting occurs when employees
approach experiencing the emotion they want to display, and high promotion-focused individuals
are more sensitive to approach-oriented strategies. Furthermore, high promotion-focused
individuals tend to experience cheerful-related emotions, which overlaps with positive affect.
Therefore, I believe that individuals high in promotion focus will experience more positive affect
than individuals low in promotion focus when engaging in deep acting.
Hypothesis 3: The positive relationship between daily deep acting and state positive
affect will be stronger for those with high versus low promotion focus.
Mood Repairment
When individuals experience negative affect or other aversive states, they want to
alleviate the negative feelings. In order to do so, individuals often choose to engage in behaviors
that bring them pleasure. These behaviors help to restore hedonic balance and act as mood
repairment (Zillman & Bryant, 1985). However, the pleasurable behaviors individuals choose to
engage in are not always healthy behaviors. These behaviors may act as a quick fix, but then

!

15

have negative long-term implications (Zillman & Bryant, 1985). In this study, I examine two
mood repairment behaviors - food and alcohol consumption – which have often been used to
mitigate negative feelings. I integrate the reward-based stress eating framework (Adam & Epel,
2007) to explain the mechanism by which food acts as mood repairment and the tension
reduction theory (Conger, 1951; 1956) to explain the mechanism by which alcohol acts as mood
repairment. Below I discuss the proposed relationship between state affect with food and alcohol
consumption.
Reward-Based Stress Eating
The impact of stress and negative affect on food consumption is widely known, both in
popular press and among researchers. In daily life, the phrases “comfort food” and “stress
eating” are frequently used to describe the use of food to cope with negative emotions. In
research, the literature is consistent regarding the relationship between the two variables; that is
that when individuals experience negative affect, they tend to eat more food and unhealthy food.
Allen and Armstrong (2006), for example, found that work-family conflict was related to
unhealthy food consumption. Ng and Jeffrey (2003) found that working adults who perceived
high stress consumed more high fat foods than working adults who perceived less stress. These
results were consistent across both genders. Other research shows a positive relationship between
negative emotions and the consumption of fast food (Hoffman, Lee, & Mendez-Luck, 2012;
Mouchacca, Abbott, & Ball, 2013; Pak, Olsen, & Mahoney, 2000; Spillman, 1990), and soda
intake (Hoffman et al., 2012; Spillman, 1990). These results suggest that consuming food high in
fat or sugar seems to be a mechanism for mood repair for individuals who experience negative
affect.

!

16

According to the reward-based stress eating framework (Adam & Epel, 2007), the
relationship between emotions and food is largely due to how individuals appraise stressors and
the subsequent biological response. If a stressor is perceived as a threat, the individual
experiences a negative emotional response. The hypothalamic-pituitary-adrenal (HPA) axis is
activated, which is how the body communicates that it is undergoing a harmful situation. The
HPA system is activated by messages from the central nervous system to the hypothalamus,
which begins the secretion of various hormones by the hypothalamus and the pituitary gland
(Straub, 2012). One of the main hormones secreted during the activation of the HPA axis is
cortisol. Cortisol is released to signal to the body to stop releasing other stress hormones.
However, increased cortisol in the body is associated with higher levels of glucose in the blood
and an increase in caloric intake, particularly with highly palatable food (Adam & Epel, 2007).
Cortisol also plays a role in drug addiction. Prior research shows that cortisol mediates
the relationship of stress on drug acquisition (Goeders, 2002). More recently, research has found
that highly caloric food has properties that promote addiction. Highly palatable food can activate
the brain reward system, which leads to the release of a number of chemicals in the brain,
including opioids. The release of opioids decreases activity of the HPA axis, which then
attenuates the negative emotional response. Therefore, over time, it seems that the overeating of
highly palatable food leads to decreased negative affect, which reinforces this eating behavior
(Adam & Epel, 2007).
Therefore, based on this theoretical framework, I expect that negative affect will be
related to both an increase in caloric intake, as well as the intake of highly palatable foods, due to
the release of cortisol. As such, I expect that surface acting will be related to both the

!

17

consumption of high fat/sugar food as well as overeating. This relationship will be mediated by
negative affect.
Hypothesis 4a: State negative affect will be positively related to the consumption of high
fat/sugar food.
Hypothesis 4b: State negative affect will mediate the positive relationship between daily
surface acting and high fat/sugar food consumption.
Hypothesis 5a: State negative affect will be positively related to overeating.
Hypothesis 5b: State negative affect will mediate the positive relationship between daily
surface acting and overeating.
On the other hand, if a stressor is perceived as a challenge, the individual experiences a
positive emotional response, and the sympathetic-adrenomedullary (SAM) system is activated.
The SAM system is the body’s initial rapid response to stress (i.e. “fight or flight” response).
Because the SAM system is designed for immediate and short-term stressors, it does not produce
cortisol. Rather, it produces catecholamines, such as adrenaline and nonadrenaline, which allows
the body to sustain a short-term stress response. While the SAM system is activated, other parts
of the body undergo various physical responses. Physical responses include the acceleration of
the heart and lungs, dilation of blood vessels, digestion slows, and hunger decreases. These
physical responses are designed to give the body increased strength, speed, and focus in order to
address the stressor at hand (Straub, 2012).
Given that the activation of the SAM system is associated with a decrease in hunger, the
slowing of the digestive system, and occurs when individuals perceive a stressor as positive and
challenging, I hypothesize that positive affect will be negatively related to overeating.

!

18

Furthermore, I expect that deep acting will be negatively related to overeating, and this
relationship will be mediated by positive affect.
Hypothesis 6a: State positive affect will be negatively related to overeating.
Hypothesis 6b: State positive affect will mediate the negative relationship between daily
deep acting and overeating.
Although there is evidence demonstrating the negative relationship between
catecholamines and quantity of food consumption, there is very little research investigating the
relationship between the release of catecholamine and the quality of food consumed. On the one
hand, being in a state of arousal with adrenaline running through the body may leave individuals
feeling lethargic and exhausted afterwards. In order to cope with this, individuals may choose to
consume high sugar/fat foods in order to re-energize. However, it is also possible that individuals
may want to avoid highly palatable because their hunger subsides during this state of arousal. It
may take longer for individuals to begin experiencing feelings of hunger again. Given that there
is no theoretical explanation for this potential relationship, I explore whether positive affect is
related to the quality of food consumed in a research question.
Research Question 1: Is there a relationship between state positive affect and the
consumption of high fat/sugar foods?
Tension Reduction Theory
Tension reduction theory points to another possible mood repair behavior—consumption
of alcoholic beverages. Conger (1951, 1956) was the first to introduce the tension reduction
theory, which states that alcohol serves to reduce tension and arousal in the body. Tension
reduction theory has two propositions. First, consumption of alcohol will reduce tension after
being exposed to any type of stressor. This proposition is known as “stress-response dampening”

!

19

(Conger, 1951, 1956; Greeley & Oei, 1999; Kushner, Sher, Wood, & Wood, 1994). Second,
exposure to stressors will promote the use of alcohol as a way to mitigate feelings of tension.
This proposition is known as “stress-induced substance abuse” (Conger, 1951; 1956; Greeley &
Oei, 1999; Kushner et al., 1994). Therefore, this theory suggests that the consumption of alcohol
relieves feelings of tensions, which thereby acts as negative reinforcement, and the cycle of
alcohol use continues and potentially escalates. The underlying mechanism by which alcohol use
reduces tension is through the reduction of self-awareness. When individuals consume alcohol,
their inhibitions are lowered, and their attention is often redirected away from the stressful
situation at hand (Conger, 1951, 1956; Greeley & Oei, 1999).
The tension reduction framework has been applied to a number of studies within the
context of work. For example, in their daily diary study, Liu, Wang, Zhan, and Shi (2009) found
that daily work stress was positively related to daily alcohol consumption, as well as the desire to
consume alcohol. Wang, Liu, Zhan, and Shi (2010) found that work-family conflict had a
significant within-person effect on daily alcohol consumption. Ballanger et al. (2010) found that
work stress and post-traumatic stress disorder predicted alcohol consumption in police officers,
regardless of gender.
Tension reduction theory states that individuals choose to consume alcohol to alleviate
tension in the body. Given that negative affect is an effect that occurs after being exposed to a
stressor and causes tension in the body, individuals are likely to consume alcohol for mood repair
after experiencing negative affect. As such, I expect that negative affect will be positively related
to alcohol consumption. Furthermore, I hypothesize that surface acting will be positively related
to alcohol consumption, and that this relationship will be mediated by negative affect.
Hypothesis 7a: State negative affect will be positively related to alcohol consumption.

!

20

Hypothesis 7b: State negative affect will mediate the positive relationship between daily
surface acting and alcohol consumption.
However, the tension reduction theory does not operationally define which types of
tensions or arousals lead to the consumption of alcohol, although the majority of research on this
topic has investigated negative arousal states (i.e. anxiety, distress; see Greeley & Oei, 1999 for a
review). Based on the ambiguity of this theory, it is unclear whether state positive affect will be
associated with alcohol use. From a biological perspective, positive affect can lead to
physiological arousal in the body. The SAM system is activated, and the body releases
adrenaline (Adam & Epel, 2007). On the other hand, because positive affect is a positive arousal
state, individuals may not perceive this as tension. Thereby in this study, I explore the
relationship between positive affect and alcohol use as exploratory research questions.
Research Question 2: Is state positive affect related to alcohol consumption?
Research Question 3: Is daily deep acting related to alcohol consumption via state
positive affect?

!

21

Figure 1. Theoretical model

!

22

CHAPTER 3
METHOD
Participants
The sample consisted of servers and bartenders working in restaurants in the Midwestern
and northeastern United States. To participate in the study, participants had to be 21+ years of
age and work either part-time or full-time as a server or bartender. Over 100 employees were
invited to participate and 44 signed up for the study by completing the baseline survey. Of the 44
participants that completed surveys, 6 individuals were eliminated for failing to complete at least
60% of the daily surveys. In both the Midwest and northeastern sample, the participants were
mostly female (71% and 83%, respectively), Caucasian (85% and 83%), and the average ages
were comparable (27.52 and 24.22). Additionally, the average tenure (1.75 years and 3.31 years)
and the number of shifts worked each week (3.62 and 5.04) were similar. Because of these
similarities, the samples were combined. The final sample consisted of 38 servers and bartenders
that were predominantly female (76%), 81.4% Caucasian, 12% Hispanic/Latino, 5.6% Asian,
and 1% other. Participants’ mean age was 26 (M = 25.70, SD = 6.61). On average, participants
worked 5 shifts a week (M = 4.80, SD = 1.93) and had been at their current job for 2.61 years (M
= 2.61, SD = 1.10).
Procedure
Participants were recruited in-person at the restaurants where they worked. The author
hung up flyers in restaurants in the Midwestern and northeastern United States and spoke with
employees at the restaurants. Flyers contained information about the study, as well as the contact
information of the author. Interested participants emailed the author and received the consent

!

23

form (see Appendix A) and baseline survey shortly after. Once the baseline survey was
completed, the daily portion of the surveys began. All surveys were completed online.
The study used an experience sampling methodology where three surveys were
completed each workday for 10 days following the baseline. The baseline survey captured
demographic information including job tenure, age, and gender, as well as regulatory focus,
alcohol expectancy, and food choice expectancy. On each workday, participants filled out the
morning survey (Time 1) which assessed sleep quality, sleep quantity, negative affect, and
positive affect This survey was completed before the start of their shift. Participants completed
the second daily survey immediately after their shift ended (Time 2), where they filled out the
extent to which they engaged in emotional labor during the shift, state affect felt during the shift,
and the quantity and quality of food consumed during the shift. Finally, participants completed
the third survey before going to bed (Time 3), which captured alcohol consumption after work
and the quantity and quality of food consumed after work.
All survey emails sent were tailored to meet participants’ individual work schedules. For
example, if a participant worked from 11-4pm, they received the Time 1 survey and 1-2
reminders before 11am, they received the Time 2 at 4pm with reminders shortly after, and they
received the Time 3 survey at 8:45pm with reminders shortly after. Participants were allowed to
complete the nighttime survey the following morning, but Times 1 and 2 had to be completed on
their respective day. In addition, scales were randomized in each survey to prevent ordering
effects.
Lastly, participants were compensated $2 per survey completed. If all surveys were
completed, participants received $60. As an incentive to complete all surveys, for each day

!

24

participants complete all three surveys, they were entered into a lottery to win an additional
$100. One person was selected as the lottery winner.
Measures
All measures can be found in Appendix B-D. Additionally, the variance decomposition
for all level-one variables are listed below and can be found in Table 1. These values were
calculated by examining the intra-class correlation (ICC-1) based on an unconditional random
coefficient model (Bliese, 2006; Bliese & Ployhart, 2002).
Demographics (Baseline). Participants reported their gender, age, ethnicity, job tenure,
job position, job location, and whether they work full-time or part-time.
Regulatory focus (Baseline). Regulatory focus was measured using the Wallace,
Johnson, & Frazier (2009) work regulatory focus scale. The scale started with “Rate how often
you focus on these thoughts and activities when you are working.” Six items measured
promotion focus (α = .78), and six items measured prevention focus (α = .84). An example
promotion focus item include “Accomplishing a lot of work” and an example prevention focus
include “Following rules and regulations at work.” Items were scored on a 5-point scale, where
1= “Never” and 5 = “Constantly.”
Sleep quality and quantity (Time 1). Sleep quality and quantity were measured before
work each day. Sleep quality was measured with one item, “How would you rate the quality of
your sleep last night?” on a three-point scale where 1 = “Poor” and 3 = “Excellent.” Sleep
quantity was measured with one-item, “How many hours did you sleep last night?” The ICC(1)
for sleep quality was .16 and .27 for sleep quantity. Therefore 16% of the variation in sleep
quality and 27% in sleep quantity can be attributed to between-person variation.

!

25

State positive and negative affect (Time 1 and Time 2). State positive and negative
affect were measured twice each day: once before work and again immediately after work using
the Positive and Negative Affect Schedule (PANAS) scale (Watson et al., 1988). Ten items
measured positive affect (Time 1: α = .96, Time 2: α = .94), and ten items measured negative
affect (Time 1: α = .74, Time 2: α = .85). Participants were asked to indicate the extent they feel
the following emotions. Sample items include, “Interested” and “Inspired” (positive affect), and
“Hostile” and “Irritable” (negative affect). Items were scored on a scale from 1-5, where 1 =
“Very slightly/Not at all” and 5 = “Extremely.”. Items were summed, so scores ranged from 1050 for each scale. At Time 1, the ICC(1) for positive and negative affect were .65 and .29,
respectively. At Time 2, the ICC(1) was .58 and .50.
Emotional labor (Time 2). Emotional labor was measured immediately after an
employee’s shift using the six-item Brotheridge and Lee (2003) and Grandey (2003) measure.
Three items assessed surface acting (α = .88), and three items measured deep acting (α = .88).
Items included, “During my shift, I resisted expressing my true feelings” to measure surface
acting, and “During my shift, I made an effort to actually feel the emotions that I needed to
display to others” to measure deep acting. Items were scored on a five-point scale, where 1 =
“Not at all” and 5 = “All the time.” The ICC(1) was .51 for surface acting and .58 for deep
acting.
Food quantity (Time 2 and Time 3). Food quantity was measured twice each day:
immediately after an employee’s shift and before the employee went to bed in the evening. Food
quantity was assessed with three items, adapted from Oliver and Wardle (1999), which ask about
overall food quantity, meal size, and snacking. Items are measured on a 5-point scale, where 1 =
“None” and 5 = “Very much.” Each item was treated as a separate variable, as prior research has

!

26

found both within- and between-person differences in the quantity of snacks and meals
consumed throughout the day (O’Connor et al., 2008). Additionally, O’Connor et al. (2008) and
Oliver and Wardle (1999) found that stress had different relationships with meal consumption
and snacking. The ICC(1) was .35 for snacking, .07 for meal size, and .09 for total food
consumption.
Food quality (Time 2 and Time 3). Food quality was assessed with a modified checklist
twice each day: immediately after an employee’s shift and before the employee went to bed in
the evening. Participants were asked to select each type of food that they have consumed during
their shift and after leaving work. The list of food and beverages were created based on the
United States Department of Agriculture’s (USDA) MyPlate guideline descriptions with 30 food
and beverage options (15 healthy, 15 unhealthy). Examples of foods high in fat or sugar include
“cheese, butter, or cream-based sauces/dips” and “cakes, cookies, pastries, and donuts.”
Examples of healthy foods/foods low in fat and sugar include, “nuts and seeds” and “raw fruit.”
Items were summed to get an overall score of healthy foods and high fat/high calorie foods. The
ICC(1) for high fat/high calorie food was .46.
Additionally, the Time 2 and Time 3 food quality and quantity variables were
respectively aggregated (e.g. Time 2 and Time 3 snacking were aggregated) to capture total food
consumption during and after the shift. Unlike traditional 9-5 hours, servers and bartenders’
shifts often occur during normal dining hours, so it is likely that people might eat toward the end
of their shift rather than waiting until the shift is over. Combining the two time points provides a
clearer picture of participants’ total food consumption. Additionally, the correlations between
each of the variables at Time 2 and Time 3 were positive and significant (p < .01), suggesting

!

27

they are strongly related. Therefore, the Time 2 and Time values for each variable (i.e. high
fa/high calorie, snacking, meal size, and total food consumption), were summed.
Alcohol consumption (Time 3). Alcohol consumption was measured immediately after
an employee’s shift. Alcohol consumption was assessed through three questions: (1) Did you
consume an alcoholic beverage? (2) What type(s) of alcoholic beverage(s) did you consume? –
check all that apply and (3) Given serving size definitions provided, how many alcoholic
beverages did you consume? Number of alcoholic beverages was used to capture alcohol
consumption. If participants reported that they did not consume an alcoholic drink, they were
assigned a 0. This variable was treated as a count variable. The ICC(1) for alcohol consumption
was .31.
Data Analysis
Data structure. Data was structured such that daily variables were nested within
participants, creating a 2-level dataset. The dataset contained 20 level-one measures and 38
individuals. Per Scherbaum and Ferreter’s (2009) recommendations, this should be sufficient to
find achieve a power of .80 and find a medium effect size. From the 38 participants retained for
analyses, a total of 370 days of data were collected. Participants provided approximately 9.68
days of data. Of the 370 days of data, several were removed for compliance reasons. Some
participants completed the Time 1 survey in the morning and did not go into work that afternoon.
For this reason, 8 days were excluded from analyses. In addition, compliance with the required
timing of the surveys also resulted in 4 days being removed. For two of the days, participants
completed all three surveys within 15 minutes of each other at the end of the day. For the other
two days, Time 1 and Time 2 surveys were completed within 10 minutes of each other after the

!

28

shift. After removing these 12 days, 358 days of level-1 data were retained for analyses, or 96%
of collected surveys.
Analytic approach. Bivariate analyses were conducted using the multilevel package in R
(Bliese 2006; Bliese, 2016), given the nested nature of the data where daily measurements (level1 variables) were nested within participants (level-2 variables). In line with recommendations by
Nezlek (2012), coefficients were allowed to vary randomly, error terms were entered for each
level-1 coefficient in level-2 equations, and random intercepts and slopes were used. Level-1
predictors (i.e. emotional labor) variables were person-mean centered, whereas the level-2
variables (i.e. regulatory focus) were grand mean centered. Moderated mediation and mediation
analyses were tested with Mplus 7.0 (Muthén & Muthén, 2012) using the Bauer, Preacher, and
Gil (2006) method, where the covariances of the Level 1 random effects were estimated in order
to estimate random indirect effects and standard errors. Finally, the Monte Carlo method was
used to estimate confidence intervals for the indirect effects in the mediation relationships
(Preacher, Zyphur, & Zhang, 2010).

!

29

CHAPTER 4
RESULTS
Descriptive statistics, including means, standard deviations, and intra-class correlations
(ICCs) are provided in Table 1. Additionally, intercorrelations between variables can be seen in
Table 2. There are several bivariate relationships worth noting. Negative affect at time 1 is
positively related to surface acting (r = .14, p < .05), and positive affect at time 1 is positively
related to deep acting (r = .16, p < .01). Surface acting is positively related to negative affect at
time 2 (r = .46, p < .01), high fat/calorie consumption (r = .27, p < .01), and snacking (r = .42, p
< .01). Surface acting is also positively related to alcohol consumption after work (r = .27, p <
.01). Surface acting is not significantly related to meal size or total food consumption during and
after work. Deep acting is positively related to positive affect at time 2 (r = .37, p < .01) and
negatively related to high fat/calorie consumption (r = -.17, p < .01) and snacking during work (r
= -.13, p < .05). Deep acting is also negatively related to alcohol consumption after work (r = .18, p < .01). Deep acting is not significantly related to meal size or total food consumption
during and after work.
Hypothesis 1a proposed that within individuals, daily surface acting will be positively
related to state negative affect. Results indicate that surface acting during the workday is
positively related to state negative affect (ɣ = 2.17, p < .001) after controlling for sleep quantity,
sleep quality, and psychological negative affect before work (see Table 3). Hypothesis 1b
proposed that daily deep acting will be positively related to state positive affect. Results indicate
that deep acting is positively related to state positive affect (ɣ = 4.07, p < .001) after controlling
for sleep quantity, sleep quality, and positive affect before work (see Table 4). Therefore,
Hypotheses 1a and 1b were supported. Although not hypothesized, there was also a significant,

!

30

negative effect between daily surface acting and state positive affect (ɣ = -1.53, p < .01) and a
significant, negative effect between daily deep acting and state negative affect (ɣ =-0.69, p < .05)
after controlling for sleep quantity, sleep quality, positive affect before work, and negative affect
before work, respectively.1
Hypothesis 2 proposed that the positive relationship between surface acting and state
negative affect will be stronger for those with high versus low prevention focus. The results
indicate that the main effect of prevention focus on state negative affect was not significant (ɣ = 1.35, ns), but the interaction term was positively and significantly related to state negative affect
(ɣ =1.85, p = < .01). Simple slopes for the association between surface acting and state negative
affect were tested for low (-1 SD below the mean) and high (+1 SD above the mean) levels of
prevention focus (Preacher, Curran, & Bauer, 2006). The simple slopes test revealed a significant
positive association between surface acting and state negative affect for high levels of prevention
focus (simple slope = 7.87, p < .05), but was not significant for low levels of prevention focus
(simple slope = 0.22, ns). Therefore, Hypothesis 2 was supported.
Hypothesis 3 stated that the positive relationship between daily deep acting and state
positive affect will be stronger for those with high versus low promotion focus. Results indicate
that both the main effect of promotion focus on state positive affect (ɣ = 3.74, ns) and the
interaction term (ɣ = 0.19, ns) are in the hypothesized direction, but are not significant (see Table
4). Therefore, Hypothesis 3 was not supported.
Hypothesis 4a proposed that state negative affect will be positively related to the
consumption of high fat/sugar food. This hypothesis was supported (ɣ = .03, p < .01; see Table

!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!
1
!I also controlled for demographic variables (i.e. age, gender, location, position) across
hypotheses. This did not change the significant findings.
!

31

5). Hypothesis 4b proposed that state negative affect will mediate the relationship between daily
surface acting and high fat/sugar food consumption. The indirect effect of surface acting on high
fat/sugar food consumption was not significant, thereby providing no support for Hypothesis 4b.
A summary of the indirect and total effects can be found in Table 6.
Hypothesis 5a stated that state negative affect will be positively related to overeating.
This hypothesis was partially supported. State negative affect was positively and significantly
related to snacking (ɣ = .10, p < .01), but was not significantly related to meal size or overall
food consumption. A summary of these results can be found in Table 7. Hypothesis 5b proposed
that state negative affect will mediate the positive relationship between daily surface acting and
overeating. This hypothesis was partially supported. State negative affect mediated the
relationship between surface acting and snacking, and the indirect effect of surface acting on
snacking via state negative affect was significant (ɣ = .11, p < .05). Moreover, the indirect effect
was moderated by prevention focus. For individuals with a high prevention focus, the indirect
effect of surface acting on snacking via state negative affect was significant (ɣ = .17, p < .05),
but the indirect effect was not significant for those with a low prevention focus (ɣ = .06, ns). A
summary of results can be seen in Table 8. However, as can be seen in Tables 9 and 10, the
indirect effect of surface acting on meal size and total food consumption were not significant.
Therefore, Hypothesis 5b was only partially supported.
Hypothesis 6a proposed that state positive affect will be negatively related to overeating.
Results indicate that the relationship between state positive affect and snacking (ɣ = .00, ns),
meal portion size (ɣ = .001, ns), and overall food consumption (ɣ = .003, ns) were not significant
(see Table 7). Therefore, Hypothesis 6a was not supported. Hypothesis 6b stated that state
positive affect will mediate the negative relationship between daily deep acting and overeating.

!

32

Hypothesis 6b was not supported for any of the overeating variables. A summary of the indirect
effects can be found in Tables 8-10. Research Question #1 explored whether there was a
relationship between state positive affect and the consumption of high fat/sugar foods. Results
indicate a significant, negative relationship between state positive affect and high fat/sugar
consumption, after controlling for sleep quantity and quality (ɣ = -.01, p < .05). See Table 5 for
results.
Hypothesis 7a stated that state negative affect will be positively related to alcohol
consumption. Results demonstrated a positive and significant effect of negative affect on alcohol
consumption after controlling for sleep quantity and quality (ɣ = 0.06, p < .01; see Table 11).
Hypothesis 7b proposed that state negative affect will mediate the positive relationship between
daily surface acting and alcohol consumption. The indirect effect was moderated by prevention
focus. For individuals with a high prevention focus, the indirect effect of surface acting on
alcohol consumption via state negative affect was significant (ɣ = 0.10, p < .05), but the indirect
effect was not significant for those with a low prevention focus (ɣ = .03, ns). Therefore,
Hypothesis 7b was partially supported. See Table 12 for a summary of the indirect and total
effects.
Research Question #2 explored whether state positive affect was related to alcohol
consumption. Results indicate that the relationship between state positive affect and alcohol
consumption was not significant (ɣ = 0.01, ns). Research Question #3 explored whether daily
deep acting was related to alcohol consumption via state positive affect. Results indicate that the
indirect effect of deep acting on alcohol consumption was not significant (ɣ = -0.01, ns). A
summary of results can be found in Table 11 and Table 12, respectively.

!

33

Supplementary analyses. In addition to the hypothesized relationships, additional
mediation analyses were explored. Because surface acting was significantly negatively related to
state positive affect, I investigated whether state positive affect mediated the relationship
between surface acting and high fat/calorie consumption, snacking, and alcohol consumption.
Results indicated that the indirect effects were not significant for any of these relationships (see
Tables 6, 8, and 12). Similarly, I tested whether state negative affect mediated the relationship
between deep acting and high fat/high calorie consumption. Again, the indirect effect was not
significant.

!

34

CHAPTER 5
DISCUSSION
As the number of service jobs continues to increase across the US and worldwide
(Bureau of Labor Statistics, 2013), the prevalence of emotional labor is likely to increase as well.
As such, it is important for scholars to expand their knowledge of emotional labor to understand
how it relates to the whole actor, beyond work-related outcomes (Grandey & Gabriel, 2015). The
current study sought to do so by exploring the spillover effect of emotional labor at work on
behaviors relevant to physical health, specifically food and alcohol consumption. Results from
the study indicate that there is a significant relationship between emotional labor, state negative
affect, and state positive affect. Additionally, state negative affect is significantly related to
snacking, high fat/high calorie consumption, and alcohol intake, whereas state positive affect is
only significantly related to high fat/high calorie intake. Finally, prevention focus moderated the
indirect relationships between surface acting, snacking, and alcohol consumption via state
negative affect. The following sections provide a more comprehensive review of the study’s
finding and their theoretical implications. Finally, practical implications are suggested,
limitations are addressed and future research directions are discussed.
Emotional Labor and State Affect
The results from this study provide support that engaging in emotional labor daily is
associated with state affective response. More specifically, engaging in surface acting daily is
positively related to state negative affect. These findings are consistent with prior emotional
labor research which found surface acting to have a detrimental effect on psychological wellbeing via emotional exhaustion (Bono & Vey, 2005; Chau et al., 2009; Goodwin et al., 2011;
Hulsheger & Schewe, 2011; Judge et al., 2009; Totterdell & Holman, 2003; Wagner et al., 2014),

!

35

burnout (Brotheridge & Grandey, 2002), strain (Hulsheger et al., 2010; Hulsheger & Schewe,
2011), negative affect (Judge et al., 2009; Scott & Barnes, 2011), and stress (Grandey, 2003;
Pugliesi, 1999). Therefore, this study provides further evidence that engaging in surface acting at
work is associated with consequential affective responses, even after controlling for affective
states prior to work.
In addition to the positive relationship between surface acting and state negative affect,
results also indicated a significant negative relationship with surface acting and state positive
affect. This provides further support for the negative relationship of surface acting. Actors
engaging in surface acting are more likely to report negative affective responses and unlikely to
report positive affective responses. Given the accumulating evidence showing the harmful
relationships of surface acting, more research should investigate effective interventions
organizations can use to train employees to utilize other emotional labor techniques that are less
detrimental to the actors themselves, such as deep acting and expression of naturally felt
emotion.
Similarly, this study found that engaging in deep acting was positively associated with
experiencing state positive affect and negatively related to experiencing state negative affect at
work. These findings are interesting, given the inconsistencies in the literature regarding the
relationship between deep acting and state affect. For example, Judge et al. (2009) found that
deep acting was significantly negatively related to both state positive and negative affect,
whereas the results of Scott and Barnes (2011) were in line with the findings from the current
study. It is possible that the inconsistencies in the literature might be due to the type of
organization employees worked for. Judge et al. (2009)’s sample included individuals across
different types of jobs, whereas Scott and Barnes (2011) used a sample of bus drivers and the

!

36

current study used a sample of servers and bartenders. It is possible that the variation in job type
may explain the differences in findings. Future research may want to investigate the boundary
conditions of job type in the relationship between deep acting and positive affect.
Emotional Labor, Regulatory Focus, and State Affect
Results from this study found mixed support for the relationship between emotional
labor, regulatory focus, and state affect. Findings indicate that the relationship between surface
acting and state negative affect is stronger when individuals are high in prevention focus. When
individuals are low in prevention focus, the relationship between surface acting and experienced
negative affect is weaker. These results suggest that an individual’s goal orientation – especially
an avoidant orientation – relates to affective responses and surface acting. In the context of the
service industry, people who tend to be preoccupied with meeting obligations may be more
sensitive to having to comply with affective display rules. Further research could explore which
organizational resources may act as a buffer for high prevention-focused individuals so they
experience less negative affect when engaging in surface acting. Additionally, future research
could investigate methods to motivate high prevention-focused individuals to engage in deep
acting, rather than surface acting. One method might include reframing affective display rules.
Despite the significant the moderation effect of prevention focus on the relationship
between daily surface acting and state negative affect, the within-person relationship between
deep acting and state positive affect was similar regardless of individuals’ level of promotion
focus. This may be due to the nature of affective display rules. Emotional labor may be more
inherently about compliance to display rules, rather than opportunities to create positive
experiences for customers, thereby making it more related to prevention focus than promotion
focus. This is unfortunate given the positive emotions that are associated with a promotion focus

!

37

(Higgins,1997). Future research could investigate ways to make promotion focus more relevant
for emotional labor. This may include reframing the affective display rules or altering the
training for new servers and bartenders to activate a state promotion focus.
State Affect and Physical Health Behaviors
The study’s findings indicate a mixed relationship between state affect and food
consumption. First, results indicate that state negative affect at work was significantly and
positively related to snacking and consumption of high fat/calorie food at Time 2 and Time 3.
This aligns with the reward-based stress eating framework, which suggests that negative
emotional and stress responses lead to increased cortisol in the body, and higher levels of cortisol
are associated with an increase in caloric intake and highly palatable foods (Adam & Epel,
2007). These findings are important for several reasons. First, these results show the harmful
effects of negative affect at work. In addition to its negative effect on psychological well-being,
state negative affect is related to employees making poor diet choices during and after work. If
sustained, these behaviors are likely to create health problems in the long-run and negatively
impact employees’ performance and absenteeism at work (Collins et al., 2005). Second, these
findings contribute to the work-nonwork literature by exploring new outcomes that relate to
negative affect at work. Future research could examine other work context factors that spillover
and influence food choices.
However, state negative affect was not significantly related to meal size or overall food
consumption at either time point. These findings may be the result of a variety of factors. First,
wanting to consume highly palatable foods as a coping response to negative emotions may lend
itself more to snacking, rather than consuming larger meal portions. Similarly, Oliver and
Wardle (1999) found that during periods of high stress, there was an increase in consumption of

!

38

“snack-type” foods and a decrease in intake for “meal type” foods. Second, if people are
snacking more, they may not eat a large meal portion. It is possible that meal portions may be
smaller depending on the quantity of snacks consumed. Subsequently, if meal portions are
smaller than usual then perceptions of total food consumption may be affected. Finally, it is
easier to snack during work than to have a full meal. This is especially true for servers and
bartenders who need to attend to their customers regularly. Therefore, further research is needed
to better understand the nature of the relationship between negative affect, overall food
consumption, and meal sizes.
Results also indicate a significant positive relationship between experiencing state
negative affect and consuming alcohol after work. This is consistent with tension reduction
theory (Conger, 1951, 1956) and builds upon prior research that has found work-related stress to
influence alcohol intake (Liu et al., 2009; Wang et al., 2010). These findings provide support for
the mood repairment framework, suggesting that individuals ingest alcohol as a coping
mechanism to enhance their mood and affect. Future research could explore whether the
relationship between negative affect and alcohol consumption is cyclical over time and whether
consuming alcohol relates to negative affect on the following workday.
State positive affect, on the other hand, was not significantly related to any of the
overeating variables. These findings are contrary to the rewards based stress eating framework
which suggested that the SAM system activated by positive arousal is associated with a decrease
in hunger (Adam & Epel, 2007). The results from this study indicate that the relationships
between state positive affect and overeating were close to zero, suggesting that positive affect
was unrelated to quantity of food intake during and after work. This may have been the result of
our measure of positive affect and whether it was capturing the corresponding positive arousal.

!

39

Further research is needed to explore whether there is a relationship between state positive affect
and quantity of food consumed.
Additionally, the results from the study indicated that state positive affect and alcohol
consumption were not significantly related. This finding provides initial evidence for boundary
conditions of tension reduction theory. Tension reduction theory suggests that individuals
consume alcohol to reduce arousal and tension in the body (Conger, 1951, 1956). However, the
theory does not specify the types of tension, and research in this area has typically assessed
negative arousal states (see Greeley & Oei, 1999). This study provides initial evidence that the
type of tension experienced might be related to whether or not alcohol is consumed. Additional
physiological research is needed to provide support for this boundary condition.
Although explored as a research question, results indicated that state positive affect was
significantly and negatively related to high fat/high calorie consumption. This finding, along
with the aforementioned finding, suggest that when individuals report experiencing positive
affective responses, they also report engaging in fewer mood repairing behaviors. These results
suggest that state positive affect is negatively or unrelated to the maladaptive behavior that
typically are associated with negative affected responses. Further research is needed to examine
these relationships over time to discover whether the negative relationship between positive
affect and some maladaptive behaviors is sustainable.
Emotional Labor and Physical Health Behaviors
The findings of this study suggest that there is a relationship between emotional labor and
physical health behaviors, albeit a complex one. At the bivariate level, surface acting was
significantly and positively related to alcohol consumption after work, high fat/high calorie
consumption during and after work, and snacking during and after work. Surface acting was not

!

40

related to meal size or overall food consumption. Additionally, for individuals with a high
prevention focus, engaging in surface acting was associated with increased snacking and alcohol
behaviors after work, and these relationships were mediated by state negative affect. However,
state negative affect did not mediate the relationship between high fat/high calorie consumption
and surface acting.
At the bivariate level, deep acting was negatively related to snacking and high fat/high
calorie food consumption during work and alcohol consumption after work. However, deep
acting was not related to snacking and high fat/calorie food consumption after work, nor were
any of the mediation relationships via positive affect significant. This may have been due to
power issues given the smaller sample size. This issue will be addressed further in the limitations
section.
It is worthwhile to note the differences in spillover effects at the bivariate level. The
negative relationship between deep acting, snacking, and high fat/high calorie consumption was
only significant for the Time 2 variables (during work). Whereas for surface acting, the positive
relationship between surface acting, snacking, and high fat consumption was significant at both
Time 2 and Time 3 (during and after work). It is possible that the negative effects of surface
acting lasts longer than the positive effects of deep acting. Future research could explore under
which conditions does emotional labor have a lasting impact and how these relationships unfold
over time with autoregressive effects.
Overall, more research is needed to understand the relationship between emotional labor
and physical health behaviors. The significant bivariate relationships suggest that there are
associations between food and alcohol consumption and emotional labor. However, the lack of

!

41

support for many of the mediation analyses suggest issues with power or that other variables may
play a role in these relationships.
Practical Implications
The findings from this study suggest that experiencing negative affect at work is
associated with a number of negative health behaviors, including snacking, increased high
fat/high calorie food consumption, and alcohol consumption both during and after work.
Engaging in these behaviors regularly can lead to the deterioration of employees’ health over
time, as well as impacting their job performance and absenteeism rates (Collins et al., 2005). As
such, the findings from this study have important practical implications.
First, organizations in the service industry could encourage their employees to engage in
deep acting rather than surface acting. This can be done by adapting training and newcomer
onboarding, reframing affective display rules to activate approach orientations, and utilizing a
different hiring process to select candidates that are more likely to engage in deep acting.
Additionally, organizations should try to promote a healthy climate at work to encourage
employees to engage in healthy behaviors. One way to do so is to provide incentives for healthy
activities. For example, the “5 a Day Challenge” encourages employees to eat five different
colored fruits and vegetables every day for a week, and the winner gets a small prize (Center for
Disease Control and Prevention, 2015).
Limitations
There were several limitations to this study, with respect to sample and measurement.
First, although in line with the guidelines offered by Scherbaum and Ferretter (2009), the sample
size for this study was fairly small. Given the complexity of the model and the small sample size,
I was unable to obtain full model estimates and it is possible that additional effects were not

!

42

detected. Future research could utilize a larger sample for a similar design. In addition, the
sample consisted of both part-time and full-time employees. For that reason, the days were not
always consecutive, and I was unable to check for autoregressive effects. However, despite the
“recovery” days in between shifts, significant results were still found between emotional labor
and health behaviors. In addition, it is very common for restaurant employees to be a
combination of part-time and full-time workers. According to the Bureau of Labor Statistics
(2010), 40% of employees in bars and restaurants work part-time. Therefore, the sample was a
realistic representation for this work context. Finally, because of the nature of their work
participants had greater access to food and alcohol, so the generalizability of these findings is
questionable. However, many other service industry jobs are located in close proximity to bars
and restaurants (i.e. retail stores are typically located in malls and strip malls) or have access to
unhealthy food options (i.e. cafeterias in hospitals). Additionally, I controlled for food
consumption during work and examined whether surface acting and state negative affect still
predicted snacking and high fat/high calorie food consumption after work. The relationships
remained significant.
There were also limitations to the methodology and measurement utilized in the study.
First, only self-report surveys were used to gather data. However, no objectives measures exist to
assess emotional labor, so self-report is the most appropriate measurement. In addition, the
overeating variables were each assessed with one item. Findings might be more accurate if these
constructs were assessed with scales, rather than single items.
Finally, although it is not a limitation, it is important to recognize the boundary
conditions of this study. This study focused on jobs that require positive affective displays. It is
unclear whether the same findings would be found for jobs that require negative affective

!

43

displays, such as tax collectors and police officers. Further research is needed to examine these
relationships.
Conclusion
The present study aimed to explore the spillover effect of emotional labor on behaviors
that influence physical health, namely food and alcohol consumption. Although the path model
did not receive support, findings suggest that there is a relationship between emotional labor and
state affect, affective responses and physical health behaviors during and after work, and a
relationship between emotional labor and health behaviors – although the relationship may be
more complicated than hypothesized. As one of the first studies to examine emotional labor and
health behaviors, this study shows the importance of expanding the nomological network of
emotional labor in order to gain a holistic understanding of the impact of emotional regulation on
the individual actors.

!

44

APPENDICES

!

45

APPENDIX A
Tables and Figures
Table 1
Descriptive statistics, reliabilities, and intra-class correlation coefficients
Min
Max
Time 1
Sleep Quality
1.00
3.00
Sleep Quantity
3.00
12.00
State PA
10.00
45.00
State NA
10.00
28.00
Time 2
State PA
10.00
44.00
State NA
10.00
42.00
Surface Acting
1.00
5.00
Deep Acting
1.00
5.00

!

Reliability

ICC(1)

ICC(2)

0.96
0.74

0.16
0.27
0.65
0.29

0.63
0.77
0.94
0.79

0.94
0.85
0.88
0.88

0.58
0.50
0.51
0.58

0.92
0.90
0.91
0.93

Time 2 & 3
High Fat/High Calorie Foods
Amount Snacking
Meal Size
Overall Food Consumption

0.00
2.00
2.00
2.00

12.00
8.00
8.00
9.00

-

0.46
0.35
0.07
0.09

0.89
0.83
0.39
0.47

Time 3
Alcohol Consumption

0.00

12.00

-

0.31

0.80

Baseline
Prevention Focus
Promotion Focus

3.17
3.33

5.00
5.00

0.84
0.78

-

-

46

Table 2
Intercorrelations between variables
M(SD)

1

2

3

4

5

6

7

8

9

10

11

1. Sleep Quantity
2. Sleep Quality
3. State Negative Affect (T1)

7.20(1.42) -1.97(0.65) .47
12.78(3.40) -.09

.30
--.15

-.27
.08
--

-.15
.17
.16

-.08
.-15
.30

.06
.11
.15

.10
-.42
.58

-.15
.20
.30

.10
.17
.35

.11
.15
.20

-.12
.14
-.29

4. State Positive Affect (T1)
5. Surface Acting

22.00(9.67) .05
2.67(0.98) -.08

.25
-.03

-.07
.14

--.20

-.31
--

.15
-.06

-.02
.56

.68
-.36

-.18
.45

-.24
.58

.02
-.13

6. Deep Acting
7. State Negative Affect (T2)
8. State Positive Affect (T2)
9. High fat/ calorie (T2/T3)

3.14(0.89)
14.58(5.19)
26.52(9.59)
2.35(1.99)

.05
-.03
-.03
.06

.04
.08
.10
.02

.01
.33
.07
.09

.16
-.06
.59
-.18

-.03
.46
-.30
.27

--.19
.37
-.17

-.27
--.16
.35

.28
-.12
--.15

-.18
.58
-.07
--

-.43
.51
-.23
.60

.02
-.20
.03
-.27

10. Snacking (T2/T3)

3.94(1.45)

.03

.06

.11

-.08

.42

-.13

.37

-.11

.57

--

-.30

11. Meal Size (T2/T3)
12. Overall Food (T2/T3)
13. Alcohol Quantity (T3)
14. Prevention Focus
15. Promotion Focus

4.21(1.38)
4.70(1.30)
1.39(2.02)
4.47(0.47)
4.21(0.51)

.04
.04
.02
-.10
-.02

-.05
.01
.15
-.06
.00

-.13
.03
.03
-.03
.03

-.03
-.01
-.02
.20
.36

-.08
.03
.27
-.06
-.20

.04
-.05
-.18
.10
.10

-.08
.14
.27
-.18
-.16

.03
.01
-.03
.03
.23

.28
.55
.25
-.15
-.14

.24
.58
.21
.01
-.05

-.77
.02
-.04
-.07

16. Age
17. Gender
18. Tenure

25.70(6.72) -.20
0.76(0.43) .17
2.61(1.11) -.01

.10
-.05
.03

-.17
.01
-.09

.30
-.27
.12

-.05
.01
.27

-.22
.24
-.06

-.09
.03
.13

.21
-.09
.01

-.06
.13
.07

-.01
.14
.09

-.02
-.03
-.08

19. Location
20. Position

3.87(3.78)
1.14(0.38)

.04
.04

-.06
-.18

.06
.01

-.14
-.13

-.08
-.04

-.15
-.25

-.07
-.01

-.09
-.12

-.02
.00

-.13
.01

.03
.02

21. Shift

1.91(0.63)

-.04

.01

-.14

.01

.10

-.05

.07

.05

.20

.03

.07

! Note: N = 358 days from N = 38 !participants, p < .01, .05; Values below the diagonal represent correlations at the between-person level, while variables
above the diagonal represent correlations at the within-person level

!
!

47

Table 2 (cont’d).

1. Sleep Quantity
2. Sleep Quality
3. State Negative Affect
(T1)
4. State Positive Affect
(T1)
5. Surface Acting
6. Deep Acting
7. State Negative Affect
(T2)
8. State Positive Affect
(T2)
9. High fat/ calorie
(T2/T3)
10. Snacking (T2/T3)
11. Meal Size (T2/T3)
12. Overall Food (T2/T3)
13. Alcohol Quantity
(T3)
14. Prevention Focus
15. Promotion Focus
16. Age
17. Gender
18. Tenure
19. Location
20. Position

12

13

14

15

16

17

18

19

20

21

.09
.20
.13

.09
-.46
.27

-.18
-.27
.06

.05
-.08
.14

-.41
.27
-.24

.43
-.04
-.04

-.07
.18
-.16

.19
-.18
.17

.08
-.26
-.16

-.16
.18
-.17

.10

-.08

.30

.42

.40

-.28

.20

-.21

-.18

-.07

-.08
-.19
.15

.44
-.38
.55

-.03
.04
-.27

-.33
.03
-.26

-.08
-.36
-.11

.06
.45
.06

.36
-.24
.21

-.08
.06
-.14

-.03
-.38
.03

.28
-.40
.12

.14

-.09

.02

.27

.27

-.10

-.01

-.02

-.21

-.14

.53

.58

-.28

-.21

-.11

.28

.18

-.06

-.10

.19

.38
.47
-.10

.40
-.27
.34
--

.08
.07
.07
-.17

-.06
.03
.02
-.19

-.01
-.09
-.17
.16

.29
-.14
.25
-.19

.33
-.43
-.22
.42

-.17
.41
.19
-.42

-.02
.07
-.13
.04

.35
-.31
-.01
.15

-.05
-.07
-.06
.10
-.05
.02
.01

-.11
-.14
.11
-.13
.21
-.20
.05

-.33
.20
-.08
-.06
-.01
-.04

.33
-.07
.01
-.06
-.04
-.06

.23
.07
--.53
.10
.19
.06

-.10
.02
-.40
--.14
.05
-.23

-.06
-.09
.38
-.13
--.68
.19

-.10
.02
-.29
.04
-.39
--.02

-.09
-.08
.05
-.27
.32
-.09
--

-.23
.01
.27
.09
.32
-.30
-.20

Note: N = 358 days from N = 38 participants, p < .01, .05; Values below the diagonal represent correlations at the between-person level, while variables
above the diagonal represent correlations at the within-person level

!
!

48

Table 3.
Surface acting predicting negative affect with prevention focus moderation
State NA (Time 2)
Predictor
Intercept
Negative Affect (Time 1)
Sleep Quantity
Sleep Quality
Surface Acting
Deep Acting

Coefficient
13.40
0.16
-0.38
-1.00
2.17
-0.69

Pseudo R2

0.16

Intercept
Negative Affect (Time 1)
Sleep Quantity
Sleep Quality
Deep Acting
Surface Acting
Prevention Focus
Promotion Focus
Surface Acting*Prevention Focus
Deep Acting*Promotion Focus
Surface Acting*Promotion Focus
Deep Acting*Prevention Focus

13.50
0.17
-.039
-0.96
-0.73
2.15
-1.35
-1.17
1.85
0.21
0.22
-0.25

Pseudo R2

0.22

SE
1.56
0.07
0.19
0.41
0.32
0.35

t
8.55
2.36
-1.94
-2.42
6.82
-1.96

p-value
< 0.001**
0.02**
.05*
0.02*
< 0.001**
.05*

1.56
0.07
0.19
0.41
0.35
0.31
1.26
1.18
0.71
0.73
0.67
0.84

8.65
2.47
-2.05
2.35
-2.08
6.94
-1.07
-0.99
2.60
0.28
0.32
-0.30

< 0.001**
0.01*
0.04*
0.04*
0.04*
< 0.001**
0.29
0.32
0.01*
0.78
0.75
0.76

Note: N = 358 days from N = 38 participants, deep acting and surface acting were centered at individuals’ means, promotion focus and prevention focus were
grand mean centered, p < .05*, .01*, coefficients are unstandardized from the R multilevel package, pseudo-R2 calculated using Snijder and Bosker’s (1999)
formula

!

49

Negative Affect

50

Low Prevention Focus
High Prevention Focus

30

10
Low

High

Surface Acting

Figure 2. Interaction of surface acting and prevention focus on negative affect
!

!

50

Table 4
Deep acting predicting positive affect with promotion focus moderation
Predictor
Intercept
Positive Affect (Time 1)
Sleep Quantity
Sleep Quality
Surface Acting
Deep Acting

Coefficient
21.18
0.28
0.14
0.96
-1.53
4.07

Pseudo R2

0.24

Intercept
Positive Affect (Time 1)
Sleep Quantity
Sleep Quality
Deep Acting
Promotion Focus
Surface Acting
Prevention Focus
Deep Acting*Promotion Focus
Surface Acting*Prevention Focus
Deep Acting*Prevention Focus
Surface Acting*Promotion Focus

21.51
0.25
0.16
0.94
4.07
3.74
-1.58
-1.46
0.19
0.87
-1.94
-0.14

Pseudo R2

0.25

State Positive Affect (Time 2)
SE
t
2.09
10.15
0.05
5.15
0.29
0.47
0.29
1.52
0.46
-3.32
0.79
5.17

p-value
<0.001**
<0.001**
0.64
0.13
0.001**
< 0.001**

2.11
0.06
0.29
0.63
0.80
2.08
0.47
2.17
1.64
1.29
1.80
1.05

< 0.001**
< 0.001**
0.59
0.14
< 0.001**
0.07
< 0.001**
0.50
0.91
0.50
0.28
0.90

10.22
4.57
0.55
1.49
5.07
1.79
-3.39
-0.67
0.11
0.68
-1.08
-0.13

Note: N = 358 days from N = 38 participants, deep acting and surface acting were centered at individuals’ means, promotion focus and prevention focus were
grand mean centered, p < .05*, .01*, coefficients are unstandardized from the R multilevel package, pseudo-R2 calculated using Snijder and Bosker’s (1999)
formula

!

51

Table 5
State affect predicting high fat/calorie food consumption
Predictor
Intercept
Sleep Quantity
Sleep Quality
Positive Affect (Time 2)
Negative Affect (Time 2)
Pseudo R2

Coefficient
0.76
0.01
-0.10
-0.01
0.03

High Fat/Calorie Food Consumption (Time 2 & 3)
SE
z
Pr(>|z|)
0.31
2.49
0.01*
0.03
0.34
0.73
0.07
-1.47
0.14
0.01
-2.04
0.04*
0.01
3.24
0.001**

0.22

Note: N = 358 days from N = 38 participants, p < .05*, .01*, coefficients are unstandardized from the R multilevel package

!

52

Table 6
Moderated mediation for emotional labor, regulatory focus, state affect, and high fat/calorie consumption
Indirect Effect
Coefficient
SE
Surface Acting ! NA ! High Fat/Calorie Consumption
0.05
0.03
Surface Acting*High Prevention ! NA ! High Fat/Calorie Consumption

0.07

0.05

[-0.07, 0.11]

Surface Acting*Low Prevention ! NA ! High Fat/Calorie Consumption
Surface Acting ! PA ! High Fat/Calorie Consumption

0.02
0.02

0.02
0.011

[-0.09, 0.13]
[-0.03, 0.16]

Deep Acting !PA ! High Fat/Calorie Consumption
Deep Acting*High Promotion ! PA ! High Fat/Calorie Consumption
Deep Acting *Low Promotion ! PA ! High Fat/Calorie Consumption

-0.04
-0.04
-0.04

0.03
0.03
0.03

[-0.12, 0.03]
[-0.11, 0.03]
[-0.13, 0.04]

Deep Acting ! NA ! High Fat/Calorie Consumption

-0.02

0.02

[-0.07, 0.02]

Total Effect
Surface Acting ! NA ! High Fat/Calorie Consumption
Surface Acting*High Prevention ! NA ! High Fat/Calorie Consumption
Surface Acting*Low Prevention ! NA ! High Fat/Calorie Consumption
Surface Acting ! PA ! High Fat/Calorie Consumption
Deep Acting ! PA ! High Fat/Calorie Consumption
Deep Acting*High Promotion ! PA ! High Fat/Calorie Consumption
Deep Acting *Low Promotion ! PA ! High Fat/Calorie Consumption
Deep Acting ! NA ! High Fat/Calorie Consumption

0.18*
0.20*
0.16*
0.15*
-0.06
-0.06
-0.06
-0.05

0.05
0.05
0.06
0.06
0.07
0.07
0.07
0.08

[0.05, 0.31]
[0.07, 0.34]
[0.02, 0.29]
[0.01, 0.30]
[-0.24, 0.12]
[-0.24, 0.12]
[-0.24, 0.13]
[-0.25, 0.16]

Note: N = 358 days from N = 38 participants, p < .05*, .01*, coefficients are unstandardized from MPlus

!

95% CI
[-0.07, 0.10]

53

Table 7
State affect predicting overeating
Predictor
Intercept
Sleep
Quantity
Sleep Quality
NA (T2)
PA (T2)
Pseudo R2

Amount of Snacking
(Time 2 & 3)
Coefficient
SE
t
p-value
2.77
0.61 4.57
0.001**
-0.02
0.06 -0.36
0.72
-0.02
0.10
0.00
0.41

0.14
0.02
0.01

-0.20
3.84
0.03

0.84
0.001**
0.97

Portion Size
(Time 2 & 3)
Coefficient SE
t
4.44
0.53 7.75
0.05
0.07 0.74
-0.17
-0.02
-0.001

0.14
0.02
0.01

0.07

-1.20
-0.94
-0.04

p-value
0.001**
0.46
0.23
0.34
0.97

Overall Food Consumption
(Time 2 & 3)
Coefficient
SE
t
p-value
4.07
0.61
6.73
0.001**
0.03
0.06
0.44
0.66
-0.07
0.03
0.003

0.13
0.02
0.01

-0.54
1.93
0.30

0.59
0.06
0.76

0.17

Note: N = 358 days from N = 38 participants, p < .05*, .01*, coefficients are unstandardized from the R multilevel package, pseudo-R2 calculated using Snijder
and Bosker’s (1999) formula

!

54

Table 8
Moderated mediation for emotional labor, regulatory focus, state affect, and snacking
Indirect Effect
Coefficient
Surface Acting ! NA ! Snacking
0.11*

95% CI
[0.02, 0.23]

Surface Acting*High Prevention ! NA ! Snacking

0.17*

0.08

[0.02, 0.22]

Surface Acting*Low Prevention ! NA ! Snacking
Surface Acting ! PA ! Snacking

0.06
-0.01

0.04
0.02

[-0.03, 0.16]
[-0.06, 0.04]

Deep Acting ! PA ! Snacking
Deep Acting*High Promotion ! PA ! Snacking
Deep Acting *Low Promotion ! PA ! Snacking
Deep Acting ! NA ! Snacking

0.02
0.02
0.02
-0.04

0.04
0.04
0.05
0.03

[-0.09, 0.14]
[-0.09, 0.14]
[-0.09, 0.14]
[-0.10, 0.04]

Total Effect
Surface Acting ! NA ! Snacking
Surface Acting*High Prevention ! NA ! Snacking

0.46*
0.52*

0.14
0.14

[0.11, 0.81]
[0.16, 0.87]

Surface Acting*Low Prevention ! NA ! Snacking
Surface Acting ! PA ! Snacking
Deep Acting ! PA ! Snacking
Deep Acting*High Promotion ! PA ! Snacking
Deep Acting *Low Promotion ! PA ! Snacking
Deep Acting ! NA ! Snacking

0.41*
0.34*
0.03
0.03
0.03
-0.03

0.14
0.13
0.12
0.12
0.13
0.12

[0.06, 0.77]
[0.01, 0.67]
[-0.29, 0.35]
[-0.09, 0.14]
[-0.09, 0.13]
[-0.34, 0.28]

Note: N = 358 days from N = 38 participants, p < .05*, .01*, coefficients are unstandardized from MPlus

!

SE
0.05

55

Table 9
Moderated mediation for emotional labor, regulatory focus, state affect, and meal size
Indirect Effect
Coefficient
Surface Acting ! NA ! Meal Size
0.05

SE
0.18

95% CI
[-0.08, 0.17]

Surface Acting*High Prevention ! NA ! Meal Size

0.07

0.07

[-0.12, 0.26]

Surface Acting*Low Prevention ! NA ! Meal Size
Deep Acting ! PA ! Meal Size
Deep Acting*High Promotion ! PA! Meal Size
Deep Acting *Low Promotion ! PA ! Meal Size

0.03
-0.08
-0.08
-0.08

0.03
0.07
0.06
0.07

[-0.04, 0.09]
[-0.26, 0.10]
[-0.25, 0.09]
[-0.27, 0.11]

Total Effect
Surface Acting ! NA ! Meal Size
Surface Acting*High Prevention ! NA ! Meal Size
Surface Acting*Low Prevention ! NA ! Meal Size
Deep Acting ! PA ! Meal Size
Deep Acting*High Promotion ! PA ! Meal Size
Deep Acting *Low Promotion ! PA! Meal Size

-0.06
-0.03
-0.08
0.05
0.05
0.05

0.15
0.15
0.16
0.17
0.18
0.18

[-0.44, 0.33]
[-0.41, 0.35]
[-0.48, 0.33]
[-0.63, 0.27]
[-0.41, 0.52]
[-0.40, 0.51]

Note: N = 358 days from N = 38 participants, p < .05*, .01*, coefficients are unstandardized from MPlus

!

56

Table 10
Moderated mediation for emotional labor, regulatory focus, state affect, and total food consumption
Indirect Effect
Coefficient
SE
Surface Acting ! NA ! Total Food
0.08
0.05

95% CI
[-0.03, 0.11]

Surface Acting*High Prevention ! NA ! Total Food

0.12

0.07

[-0.04, 0.19]

Surface Acting*Low Prevention ! NA ! Total Food
Deep Acting ! PA ! Total Food
Deep Acting*High Promotion ! PA ! Total Food
Deep Acting *Low Promotion ! PA ! Total Food

0.04
-0.05
-0.05
-0.05

0.03
0.08
0.08
0.09

[-0.03, 0.30]
[-0.20, 0.10]
[-0.22, 0.08]
[-0.21, 0.11]

Total Effect
Surface Acting ! NA ! Total Food
Surface Acting*High Prevention ! NA ! Total Food
Surface Acting*Low Prevention ! NA ! Total Food
Deep Acting ! PA ! Total Food
Deep Acting*High Promotion ! PA ! Total Food
Deep Acting *Low Promotion ! PA ! Total Food

0.18
0.22
0.14
0.05
0.06
0.05

0.14
0.14
0.14
0.19
0.18
0.19

[-0.18, 0.53]
[-0.14, 0.57]
[-0.23, 0.50]
[-0.40, 0.55]
[-0.42, 0.49]
[-0.41, 0.51]

Note: N = 358 days from N = 38 participants, p < .05*, .01*, coefficients are unstandardized from MPlus

!

57

Table 11
State affect predicting alcohol consumption
Alcohol (Time 3)
Predictor
Intercept
Sleep Quantity
Sleep Quality
PA (Time 2)
NA (Time 2)
Pseudo R2

Coefficient
0.42
-0.07
0.23
0.01
0.06

SE
0.80
0.09
0.20
0.02
0.02

z
0.51
-0.76
1.16
0.35
2.60

Pr(>|z|)
0.61
0.45
0.25
0.72
0.01**

0.10

Note: N = 358 days from N = 38 participants, p < .05*, .01*, coefficients are unstandardized from the R multilevel package, pseudo-R2 calculated using Snijder
and Bosker’s (1999) formula

!

58

Table 12
Moderated mediation for emotional labor, regulatory focus, state affect, and alcohol consumption
Indirect Effect
Coefficient
SE
Surface Acting ! NA ! Alcohol
0.06
0.03
Surface Acting*High Prevention ! NA ! Alcohol

0.10*

0.05

[0.001, 0.15]

Surface Acting*Low Prevention ! NA ! Alcohol
Surface Acting ! PA ! Alcohol
Deep Acting ! PA ! Alcohol
Deep Acting*High Promotion ! PA ! Alcohol
Deep Acting *Low Promotion ! PA ! Alcohol

0.03
-0.02
-0.01
-0.01
-0.02

0.06
0.02
0.06
0.06
0.06

[-0.03, 0.09]
[-0.07, 0.04]
[-0.17, 0.14]
[-0.17, 0.13]
[-0.17, 0.14]

Total Effect
Surface Acting ! NA ! Alcohol
Surface Acting*High Prevention ! NA ! Alcohol
Surface Acting*Low Prevention ! NA ! Alcohol
Surface Acting ! PA ! Alcohol
Deep Acting ! PA ! Alcohol
Deep Acting*High Promotion ! PA ! Alcohol
Deep Acting *Low Promotion ! PA ! Alcohol

0.12
0.15*
0.09
0.04
-0.10
-0.10
-0.11

0.06
0.06
0.06
0.06
0.15
0.16
0.15

[-0.03, 0.26]
[0.02, 0.29]
[-0.07, 0.24]
[-0.12, 0.19]
[-0.74, 0.06]
[-0.74, 0.06]
[-0.74, 0.06]

Note: N = 358 days from N = 38 participants, p < .05*, .01*, coefficients are unstandardized from MPlus

!

95% CI
[-0.03, 0.14]

59

APPENDIX B
Consent Form
Consent to Participate in a Research Study: Working in the service industry and health behaviors
Please read the information below completely and carefully:
This research study investigates the affective experiences and health behaviors of people who work in the
service industry, namely servers and bartenders. The research is designed to help us better understand
how work experiences can impact employees’ health behaviors during and outside of work. The research
contains two components:
1. A baseline survey (about 15 minutes) of demographic and background information
2. For 10 work days, you will fill out 3 short surveys: (~2-3 minutes) before work, immediately after
work, and before bed.
You will be compensated $2 for each short survey. Therefore, if all surveys are completed, you will be
compensated $60. However, if all surveys are not completed, you will not receive this amount (i.e. if
you fill out 20 surveys, you will receive 20 x $2 = $40). In addition, for each day you complete all three
surveys, you will be entered into a “lottery” to win $100. You will receive your initial compensation after
you have completed your 10 days. The “lottery” drawing will occur after all participants have completed
their 10 days.
You will use your birthday and phone number as an identifier, in order to be properly compensated you
and link your survey responses. All information you provide will be confidential. Only the primary
investigator, the co-investigator, and the Institutional Review Board (should the project be selected for
audit) will have access to the data and/or the identification numbers. Data will be stored for five years on
the investigator’s password protected computer in an encrypted folder. Please be aware that your
confidentiality will be protected to the maximum extent allowable by law.
Please note that your participation is voluntary. This means that refusal to participate will
involve NO penalty. We will not be keeping record of who does not participate. You are free to withdraw
your consent and discontinue participation at any point during the survey. You may choose not to answer
specific questions or to stop participating at any time.
We appreciate your time and participation! If you have concerns or questions about this study, such
as scientific issues, how to do any part of it, or to report an injury due to your participation in the research
(i.e., physical, psychological, social, financial, or otherwise), please contact Taylor Lauricella,
Psychology Building, 316 Physics, Room 302, East Lansing, MI 48824. E-mail: laurice5@msu.edu,
Phone: (201) 264-1256. If you have questions or concerns about your role and rights as a research
participant, would like to obtain information or offer input, or would like to register a complaint about
this study, you may contact, anonymously if you wish, the Michigan State University's Human Research
Protection Program at 517-355-2180, Fax 517-432-4503, or e-mail irb@msu.edu or regular mail at 207
Olds Hall, MSU, East Lansing, MI 48824.
Please write your name below to indicate that you voluntarily agree to participate in this research study.

!

60

APPENDIX C
Baseline Survey
Demographic Information
•! How old are you?
•! What is your sex?
o! Male
o! Female
•! What is your ethnicity?
o! Hispanic or Latino
o! Not Hispanic of Latino
•! What is your race?
o! White
o! Black or African American
o! American Indian or Alaska Native
o! Asian
o! Native Hawaiian or Pacific Islander
o! Other
•! What is the name of the restaurant you currently work at?
•! How long have you been working at this restaurant?
o! Drop down menu:
o! < 1 year
o! 1-2 years
o! 3-4 years
o! 5+ years
•! How many shifts do you work each week?
•! What is your work schedule?
•! Do you work full-time or part-time?
o! Full-time
o! part-time
•! Job Position (check all that apply)
o! Server
o! Bartender
Regulatory Focus
Rate how often you focus on these thoughts and activities when you are working.
1.! Following rules and regulations at work
2.! Completing my work tasks correctly
3.! Doing my duty at work
4.! My work responsibilities
5.! Fulfilling my work obligations
6.! On the details of my work
7.! Accomplishing a lot of work
8.! Getting my work done no matter what
9.! Getting a lot of work finished in a short amount of time
!

61

10.!Work activities that allow me to get ahead at work
11.!My work accomplishments
12.!How many job tasks I can complete

!

62

APPENDIX D
Before Work Survey (Daily)
Right now, I feel:
!! Interested
!! Distressed
!! Excited
!! Upset
!! Strong
!! Guilty
!! Scared
!! Hostile
!! Enthusiastic
!! Proud
!! Irritable
!! Alert
!! Ashamed
!! Inspired
!! Nervous
!! Determined
!! Attentive
!! Jittery
!! Active
!! Afraid
1 = Very slightly or not at all
2 = A little
3 = Moderately
4 = Quite a bit
5 = Extremely
How many hours did you sleep LAST NIGHT?_____
How would you rate the quality of your sleep LAST NIGHT? When responding, please consider
the quality of your sleep last night relative to what’s typical for you.
- Poor
- Average
- Excellent

!

63

APPENDIX E
After Work Survey (Daily)
Which shift did you work? (Drop down)
!! Lunch
!! Dinner
!! Double
What position were you? (Drop down)
!! Server
!! Bartender
** logic, if server: Approximately how many tables did you have?
If bartender, Approximately, how many customers did you serve at the bar?
During my shift, I… (Brotheridge & Lee, 2003; Grandey, 2003)
!! Resisted expressing my true feelings
!! Pretended to have emotions that I didn’t really have
!! Hid my true feelings about a situation
!! Made an effort to actually feel the emotions that I needed to display to others
!! Tried to actually experience the emotions that I must show
!! Really tried to feel the emotions I have to show as part of my job
1 = Never
2 = Rarely
3 = Sometimes
4 = Often
5 = Always
During my shift, I felt…
!! Interested
!! Distressed
!! Excited
!! Upset
!! Strong
!! Guilty
!! Scared
!! Hostile
!! Enthusiastic
!! Proud
!! Irritable
!! Alert
!! Ashamed
!! Inspired
!! Nervous
!! Determined
!! Attentive
!

64

!! Jittery
!! Active
!! Afraid
1 = Very slightly or not at all
2 = A little
3 = Moderately
4 = Quite a bit
5! = Extremely
What foods and beverages did you consume DURING your shift today? Please check all that
apply. (*add in servings)
Beverages
Diet and Non-diet soda (Coca-cola, Pepsi, Gingerale, Diet Coke, Sweet tea, Sprite,)
Energy drinks (RedBull, Monster)
Sports drinks (Gatorade, PowerAid)
Sugar-sweetened fruit juices/cocktails/mixers (e.g. cranberry juice cocktail, lemonade, fruit
punch, margarita mix)
Sweetened tea
Full fat (whole) milk, cream, half-and-half
Water and carbonated water (unsweetened)
Low-fat or fat-free milk, soy milk, almond milk
100% Juice
Unsweetened tea (green tea, black tea, mint tea)
Foods
Cakes, cookies, pastries, and donuts (e.g. cupcakes)
Candies and candy bars (e.g. Sour Patch Kids, M&Ms)
Fried foods (e.g., french fries, fried chicken, mozzarella sticks, fried vegetables, fried seafood)
Chips and related bagged snack foods (e.g., Doritos, Cheetos, Tortilla Chips)
Cheese, butter, or cream-based sauces/dips/spreads (e.g., nacho cheese, alfredo sauce, ranch
dressing, sour cream)
Pizza (with cheese and/or meat)
Sausages, bacon, regular hot dogs, ribs, regular (less than 85% lean) ground beef, deli meat
Ice cream, frozen full-fat yogurt, or frozen dairy desserts/novelties (e.g., Klondike bar, popsicles)
Refined grains (e.g. white bread, white rice, pasta, flour tortillas, pretzels, crackers, bagel, pita,
couscous)
Whole grains (whole wheat bread, brown rice, popcorn, oatmeal, corn tortilla, whole-wheat
tortilla, whole-wheat pasta, whole-wheat crackers, quinoa)
Raw fruit (e.g. apple, banana, strawberries, mango, avocado)
Canned fruit in light syrup (e.g. canned peaches, canned pineapples)
Raw vegetables (e.g. peppers, carrots, lettuce)
Cooked vegetables (e.g. sautéed vegetables, baked vegetables) and canned vegetables (e.g.
canned green beans, canned corn)
Chicken, >85% lean ground beef/turkey

!

65

Dried fruit (e.g. dried cranberries, dried apricots)
Nuts and seeds (e.g. cashews, almonds, peanuts)
Beans (e.g. black beans, refried beans)
Eggs
Seafood (e.g. shrimp, salmon)
The amount of snacking I did during my shift was:
a)! None
b)! Very little
c)! Moderate
d)! Quite a bit
e)! A lot
During my shift, the portion size of my meal was:
a)! I did not eat a meal during my shift
b)! Very small
c)! Moderate
d)! Large
e)! Very large
Overall, the amount of food I consumed during my shift was:
a)! None
b)! Very little
c)! Moderate
d)! Quite a bit
e)! A lot
Overall, how would you rate your customer interactions today?
a.! Very poor
b.! Poor
c.! Satisfactory
d.! Good
e.! Excellent

!

66

APPENDIX F
Before Bed Survey (Daily)
1.! Since leaving work today, did you consume an alcoholic beverage?
a.! Yes
b.! No
c.! ** add logic into qualtrics, If yes…**
2.! What type of alcoholic beverages did you consume (check all that apply)?
a.! Wine
b.! Beer
c.! Hard Liquor
3.! One drink serving is defined as 12 oz of beer (one can or bottle), 5 oz of wine (one glass),
or 1.5 oz of liquor (one shot). Given these definitions, how many alcoholic beverages did
you consume? (fill in) __________
4.! Since leaving work today, what foods and beverages did you consume? Please check all
that apply.
Beverages
Diet and Non-diet soda (Coca-cola, Pepsi, Gingerale, Diet Coke, Sweet tea, Sprite,)
Energy drinks (RedBull, Monster)
Sports drinks (Gatorade, PowerAid)
Sugar-sweetened fruit juices/cocktails/mixers (e.g. cranberry juice cocktail, lemonade, fruit
punch, margarita mix)
Sweetened tea
Full fat (whole) milk, cream, half-and-half
Water and carbonated water (unsweetened)
Low-fat or fat-free milk, soy milk, almond milk
100% Juice
Unsweetened tea (green tea, black tea, mint tea)
Foods
Cakes, cookies, pastries, and donuts (e.g. cupcakes)
Candies and candy bars (e.g. Sour Patch Kids, M&Ms)
Fried foods (e.g., french fries, fried chicken, mozzarella sticks, fried vegetables, fried seafood)
Chips and related bagged snack foods (e.g., Doritos, Cheetos, Tortilla Chips)
Cheese, butter, or cream-based sauces/dips/spreads (e.g., nacho cheese, alfredo sauce, ranch
dressing, sour cream)
Pizza (with cheese and/or meat)
Sausages, bacon, regular hot dogs, ribs, regular (less than 85% lean) ground beef, deli meat
Ice cream, frozen full-fat yogurt, or frozen dairy desserts/novelties (e.g., Klondike bar, popsicles)
Refined grains (e.g. white bread, white rice, pasta, flour tortillas, pretzels, crackers, bagel, pita,
couscous)
Whole grains (whole wheat bread, brown rice, popcorn, oatmeal, corn tortilla, whole-wheat
tortilla, whole-wheat pasta, whole-wheat crackers, quinoa)

!

67

Raw fruit (e.g. apple, banana, strawberries, mango, avocado)
Canned fruit in light syrup (e.g. canned peaches, canned pineapples)
Raw vegetables (e.g. peppers, carrots, lettuce)
Cooked vegetables (e.g. sautéed vegetables, baked vegetables) and canned vegetables (e.g.
canned green beans, canned corn)
Chicken, >85% lean ground beef/turkey
Dried fruit (e.g. dried cranberries, dried apricots)
Nuts and seeds (e.g. cashews, almonds, peanuts)
Beans (e.g. black beans, refried beans)
Eggs
Seafood (e.g. shrimp, salmon)
The amount of snacking I did after leaving work was:
f)! None
g)! Very little
h)! Moderate
i)! Quite a bit
j)! A lot
After leaving work, the portion size of my meal was:
f)! I did not eat a meal during my shift
g)! Very small
h)! Moderate
i)! Large
j)! Very large
Overall, the amount of food I consumed after leaving work was:
f)! None
g)! Very little
h)! Moderate
i)! Quite a bit
j)! A lot

!

68

REFERENCES

!

69

REFERENCES

Adam, T.C., & Epel, E.S. (2007). Stress, eating and the reward system. Physiology & Behavior,
91, 339-458. doi:10.1016/j.physbeh.2007.04.011
Allen, T.D., & Armstrong, J. (2006). Further examination of the link between work=family
conflict and physical health: The role of health-related behaviors. American Behavioral
Scientist, 49, 1204-1221. doi:10.1177/0002764206286386
Armeli, S., Carney, M.A., Tennen, H., Affleck, G., & O’Neil, T.P. (2000). Stress and alcohol
use: A daily process examination of the stressor-vulnerability model. Journal of
Personality and Social Psychology, 78, 979-994. doi:10.1037//0022-3514.78.5.979
Ballanger, J.F., Best, S.R., Metzler, T.J., Wasserman, D.A., Mohr, D.C., Liberman, A., Delucchi,
K., Weiss, D.S., Fagan, J.A., Waldrop, A.E., & Marmar, C.R. (2010). Patterns and
predictors of alcohol use in male and female urban police officers. The American Journal
on Addiction, 20, 21-29. doi:10.1111/j.1521-0391.2010.00092.x
Bauer, D.J., Preacher, K.J., & Gil, K.M. (2006). Conceptualizing and testing random indirect
effects and moderated mediation in multilevel models: New procedures and
recommendations. Psychological Methods, 11, 142-162. doi:10.1037/1082989X.11.2.142
Bliese, P.D. (2006). Multilevel modeling in R (2.2): A brief introduction to R, the multilevel
package and the nlme package. Washington, DC: Walter Reed Army Institute of
Research.
Bliese, P.D., & Ployhart, R.E. (2002). Growth modeling using random coefficient models: Model
building, testing, and illustrations. Organizational Research Methods, 5, 362-387.
doi:10.1177/109442802237116
Bono, J. E., & Vey, M. A. (2005). Toward understanding emotional management at work: A
quantitative review of emotional labor research. In C.E. Hartel, W.J. Zerbe, & N.M.
Ashkanasy (Eds.) Emotions in organizational behavior (pp. 213-233). Mahwah, NJ:
Lawrence Erlbaum Associated Publishers.
Brotheridge, C.M., & Grandey, A.A. (2002). Emotional labor and burnout: Comparing two
perspectives of “people work.” Journal of Vocational Behavior, 60, 17-39.
doi:10.1006/jvbe.2001.1815

!

70

Brotheridge, C.M., & Lee, R.T. (2003). Development and validation of the emotional labour
scale. Journal of Occupational and Organizational Psychology, 76, 365-379.
doi:10.1348/096317903769647229
Bureau of Labor Statistics (2013). International comparisons of annual labor force statistics,
1970-2012. Retrieved from http://www.bls.gov/fls/flscomparelf/lfcompendium.pdf.
Chau, S.L., Dahling, J.J., Levy, P.E., & Diefendorff, J.M. (2009). A predictive study of
emotional labor and turnover. Journal of Organizational Behavior, 30, 1151-1163.
doi:10.1002/job.617
Center for Disease Control and Prevention (2014). National prevention strategy: America’s plan
for better health and wellness. Retrieved from
http://www.cdc.gov/features/preventionstrategy/
Centers for Disease Control and Prevention (2015). Workplace health strategies. Retrieved from
https://www.cdc.gov/workplacehealthpromotion/health-strategies/index.html
Collins, J.J. Baase, C.M., Sharda, C.E., Ozminkowski, R.J., Nicholson, S., Billotti, G.M., &
Berger, M.L. (2005). The assessment of chronic health conditions on work performance,
absence, and total economic impact for employers. Journal of Occupational and
Environmental Medicine, 47, 547-557.
Conger, J.J. (1951). The effect of alcohol on conflict behavior in the albino rat. Quarterly
Journal of Studies on Alcohol, 12, 1-49.
Conger, J.J. (1956). Reinforcement theory and the dynamics of alcoholism. Quarterly Journal of
Studies on Alcohol, 17, 296-305.
Dahling, J.J. & Johnson, H.A. (2010). Promoting good service: Effects of trait and state
regulatory focus on emotional labor. In 25th Annual Conference of the Society for
Industrial and Organizational Psychology, Atlanta, GA.
Dahling, J.J., & Perez, L.A. (2010). Older worker, different actor? Linking age and emotional
labor strategies. Personality and Individual Differences, 48, 574-578.
doi:10.1016/j.paid.2009.12.009
Diestel, S., Rivkin, W., & Schmidt, K-H. (2015). Sleep quality and self-control capacity as
protective resources in the daily emotional labor process: Results from two diary studies.
Journal of Applied Psychology, 100, 809-827. doi:10.1037/a0038373
Esquirol, Y., Bongard, V., Mabile, L., Jonnier, B., Soulat, J-M., & Perret, B. (2009). Shift work
and metabolic syndrome: Respective impacts of job strain, physical activity, and dietary
rhythms. The Journal of Biological and Medical Rhythm Research, 26, 544-559.

!

71

Gabriel, A.S., Daniels, M.A., Diefendorff, J.M., & Greguras, G.J. (2015). Emotional labor
actors: A latent profile analysis of emotional labor strategies. Journal of Applied
Psychology, 100, 863-882. doi:10.1037/a0037408
Gabriel, A.S., & Diefendorff, J.M. (2015). Emotional labor dynamics: A momentary approach.
Academy of Management Journal, 58, 1804-1825. doi:10.5465/amj.2013.1135
Goeders, N.E. (2002). Stress and cocaine addiction. Journal of Pharmacology and Experimental
Therapeutics, 301, 785-789. doi:10.1124/jpet.301.3.785
Goodwin, R.E., Groth, M., & Frenkel, S.J. (2011). Relationships between emotional labor, job
performance, and turnover. Journal of Vocational Behavior, 79, 538-548.
doi:10.1016/j.jvb.2011.03.001
Grandey, A.A. (2000). Emotion regulation in the workplace: A new way to conceptualize
emotional labor. Journal of Occupational Health Psychology, 5, 95-110.
doi:10.1037/1076-8998.5.1.95
Grandey, A.A. (2003). When “the show must go on”: Surface acting and deep acting as
determinants of emotional exhaustion and peer-rated service delivery. Academy of
Management Journal, 46, 86-96. doi:10.2307/30040678
Grandey, A.A., & Gabriel, A.S. (2015). Emotional labor at a crossroads: Where do we go from
here? Annual Review of Organizational Psychology and Organizational Behavior, 2,
323-349. doi:10.1146/annurev-orgpsych-032414-111400
Greeley, J., & Oei, T. (1999). Alcohol and tension reduction. Psychological Theories of Drinking
and Alcoholism, 2, 14-53.
Gross, J.J. (1998). The emerging field of emotional regulation: An integrative review. Review of
General Psychology, 2, 271-299. doi:10.1037/1089-2680.2.3.271
Higgins, E.T. (1997). Beyond pleasure and pain. American psychologist, 52, 1280-1300.
doi:10.1037/0003-066X.52.12.1280
Hochschild, A. R. (1983). The managed heart: Commercialization of human feeling. Berkeley:
University of California Press.
Hoffman, G.J., Lee, J., & Mendez-Luck, C.A. (2012). Health behaviors among baby boomer
informal caregivers. The Gerontologist, 52, 219-230. doi:10.1093/geront/gns003
Holman, D., Chissick, C., & Totterdell, P. (2002). The effects of performance monitoring on
emotional labor and well-being in call centers. Motivation and Emotion, 26, 57-81.
doi:10.1023/A:1015194108376

!

72

Hopp, H., Rohrmann, S., Zapf, D., & Hodapp, V. (2010). Psychophysiological effects of
emotional dissonance in a face-to-face service interaction. Anxiety, Stress, & Coping, 23,
399-414. doi:10.1080/10615800903254091
Hulsheger, U.R., Lang, J.W.B., & Maier, G.W. (2010). Emotional labor, strain, and
performance: Testing reciprocal relationships in a longitudinal panel study. Journal of
Occupational Health Psychology, 15, 505-521. doi:10.1037/a0021003
Hulsheger, U.R., Lang, J.W.B., Schewe, A.F., & Zijlstra, F.R.H. (2015). When regulating
emotions at work pays off: A diary and an intervention study on emotional regulation and
customer tips in service jobs. Journal of Applied Psychology, 100, 263-277.
doi:10.1037/a0038229
Hulsheger, U.R., & Schewe, A.F. (2011). On the costs and benefits of emotional labor: A metaanalysis of three decades of research. Journal of Occupational Health Psychology, 16,
361-389. doi:10.1037/a0022876
Judge, T.A., Woolf, E.F., & Hurst, C. (2009). Is emotional labor more difficult for some than for
others? A multilevel, experience-sampling study. Personnel Psychology, 62, 57-88.
doi:10.1111/j.1744-6570.2008.01129.x
Krannitz, M.A., Grandey, A.A., Liu, S., & Almeida, D.A. (2015). Workplace surface acting and
marital partner discontent: anxiety and emotional exhaustion spillover mechanisms.
Journal of Occupational Health Psychology, 20, 314-325. doi:10.1037/a0038763
Kushner, M.G., Sher, K.J., Wood, M.D., & Wood, P.K. (1994). Anxiety and drinking behavior:
Moderating effects of tension-reduction alcohol outcomes expectancies. Alcoholism:
Clinical and Experimental Research, 18, 852-860. doi:10.1111/j.15300277.1994.tb00050.x
Liu, S., Wang, M., Zhan, Y., & Shi, J. (2009). Daily work stress and alcohol use: Testing the
cross-level moderation effects of neuroticism and job involvement. Personnel
Psychology, 62, 575-597. doi:10.1111/j.1744-6570.2009.01149.x
Lowden, A., Moreno, C., Holmback, U., Lennernas, M., & Tucker, P. (2010). Eating and shift
work – effects on habits, metabolism, and performance. Scandinavian Journal of Work,
Environment, & Health, 36, 150-162.
Mesmer-Magnus, J.R., DeChurch, L.A., & Wax, A. (2012). Moving emotional labor beyond
surface and deep acting: A discordance-congruence perspective. Organizational
Psychology Review, 2, 6-53. doi:10.1177/2041386611417746
Mohr, C.D., Armeli, S., Tennen, H., Temple, M., Todd, M., Clark, J., & Carney, M.A. (2005).
Moving beyond the keg party: a daily process study of college student drinking
motivations. Psychology of Addictive Behaviors, 19, 392-403. doi:10.1037/0893164X.19.4.392

!

73

Morris, J.A., & Feldman, D.C. (1997). The dimensions, antecedents, and consequences of
emotional labor. Academy of Management Review, 21, 986-1010.
Mouchacca, J., Abbott, G.R., & Ball, K. (2013). Associations between psychological stress,
eating, physical activity, sedentary behaviours and body weight among women: A
longitudinal study. BMC Public Health, 13, 828-839. doi:10.1186/1471-2458-13-828
Muthén, L. K., & Muthén, B. O. (2012). Mplus user’s guide (7th ed.). Los Angeles, CA: Author.
Nezlek, J.B. (2012). Multilevel modeling analyses of diary-style data. Handbook of research
methods for studying daily life, 357-383.
Ng, D.M., & Jeffery, R.W. (2003). Relationships between perceived stress and health behaviors
in a sample of working adults. Health Psychology, 22, 638-642. doi:10.1037/02786133.22.6.638
O’Connor, D.B., Jones, F., Conner, M., McMillan, B., & Ferguson, E. (2008). Effects of daily
hassles and eating style on eating behavior. Health Psychology, 27, 520-531.
Oliver, G., & Wardle, J. (1999). Perceived effects of stress on food choice. Physiology &
Behavior, 66, 511-515. doi:10.1016/S0031-9384(98)00322-9
Pak, S.H.K., Olsen, L.K., & Mahoney, B.S. (2000). The relationships of health behaviors to
perceived stress, job satisfaction, and role modeling among health professionals in South
Korea. International Quarterly of Community Health Education, 19, 65-76.
doi:10.2190/XU7N-H2MR-7BHK-513A
Parasuraman, A., Zeithaml, V.A., & Berry, L.L. (1985). A conceptual model of service quality
and its implication for future research. Journal of Marketing, 49, 41-50.
doi:10.2307/1251430
Pham, M.T., & Avnet, T. (2009). Contingent reliance on the affect heuristic as a function of
regulatory focus. Organizational Behavior and Human Decision Processes, 108, 267278. doi:10.1016/j.obhdp.2008.10.001
Preacher, K.J., Curran, P.J., Bauer, D.J. (2006). Computational tools for probing interactions in
multiple linear regression, multilevel modeling, and latent curve analysis. Journal of
Educational and Behavioral Statistics, 31, 437-448. doi:10.3102/10769986031004437.
Preacher, K.J., Zyphur, M.J., & Zhang, Z. (2010). A general multilevel SEM framework for
assessing multilevel mediation. Psychological Methods, 15, 209-233.
doi:10.1037/a0020141
Pugh, S.D. (2001). Service with a smile: Emotional contagion in the service encounter. Academy
of Management Journal, 44, 1018-1027. doi:10.2307/3069445

!

74

Pugliesi, K. (1999). The consequences of emotional labor: Effects on work stress, job
satisfaction, and well-being. Motivation and Emotion, 23, 125-154.
doi:10.1023/A:1021329112679
Rafaeli, A., & Sutton, R.I. (1987). Expression of emotion as part of the work role. Academy of
Management Review, 12, 23-37.
Scott, B.A., & Barnes, C.M. (2011). A multilevel field investigation of emotional labor, affect,
work withdrawal, and gender. Academy of Management Journal, 54, 116-136.
doi:10.5465/AMJ.2011.59215086
Schaubroeck, J., & Jones, J.R. (2000). Antecedents of workplace emotional labor dimensions
and moderators of their effects on physical symptoms. Journal of Organizational
Behavior, 21, 163-183. doi:10.1002/(SICI)1099-1379(200003)21:2<163::AIDJOB37>3.3.CO;2-C
Scherbaum, C.A., & Ferreter, J.M. (2009). Estimating statistical power and required sample sizes
for organizational research using multilevel modeling. Organizational Research Methods,
12, 347-367. doi:10.1177/1094428107308906
Spillman, D. (1990). Survey of food and vitamin intake responses reported by university students
experiencing stress. Psychological Reports, 66, 499-502. doi:10.2466/pr0.1990.66.2.499
Steptoe, A., Lipzey, Z., & Wardle, J. (1998). Stress, hassles, and variations in alcohol
consumption, food choice and physical exercise: A diary study. British Journal of Health
Psychology, 3, 51-63. doi:10.1111/j.2044-8287.1998.tb00555.x
Straub, R.O. (2012). Health psychology: A biopsychosocial approach. New York: Worth
Publishers.
Totterdell, P. & Holman, D. (2003). Emotional regulation in customer service roles: Testing a
model of emotional labor. Journal of Occupational Health Psychology, 8, 55-73.
doi:10.1037/1076-8998.8.1.55
Wagner, D.T., Barnes, C.M., & Scott, B.A. (2013). Driving it home: How workplace emotional
labor harms employee home life. Personnel Psychology, 67, 487-516.
doi:10.1111/peps.12044
Wallace, J.C., Johnson, P.D., Frazier, M.L. (2009). An examination of the factorial, construct,
and predictive validity and utility of the regulatory focus at work scale. Journal of
Organizational Behavior, 30, 805-831.
Wang, M., Liu, S., Zhan, Y., & Shi, J. (2010). Daily work-family conflict and alcohol use:
Testing the cross-level moderation effects of peer drinking norms and social support.
Journal of Applied Psychology, 95, 377-386. doi:10.1037/a0018138

!

75

Watson, D., Clark, L.A., & Tellegen, A. (1988). Development and validation of brief measures
of positive and negative affect: The PANAS scales. Journal of Personality and Social
Psychology, 54, 1063-1070. doi:10.1037/0022-3514.54.6.1063
Zillmann, D., & Bryant, J. (1985). Affect, mood, and emotion as determinants of selective
exposure. Selective Exposure to Communication, 157-190.

!

76