UNIVERSIT DEGLI STUDI DI TORINO

DIPARTIMENTO DI PSICOLOGIA

DOTTORATO DI RICERCA IN SCIENZE UMANE

PSICOLOGIA DELLA SALUTE E DELLA QUALIT DELLA VITA
CICLO: XXV

TITOLO DELLA TESI

Workaholism: Definitions, measures, and dynamics

DOTTORANDA: Monica Molino

TUTOR: Prof.ssa Chiara Ghislieri

COORDINATORE DEL DOTTORATO: Prof.ssa Cristina Onesta Mosso

ANNI ACCADEMICI: 2010/2012

SETTORE SCIENTIFICO-DISCIPLINARE DI AFFERENZA: M-PSI/06

CONTENTS

INTRODUCTION................................................................................................................. p. 1
Defining workaholism

CHAPTER 1.......................................................................................................................... p. 6
Workaholism measures: What choice? The Italian version of DUWAS and BWAS

CHAPTER 2........................................................................................................................ p. 25
The role of workaholism in the job demands-resources model

CHAPTER 3........................................................................................................................ p. 44
Workaholism and recovery experiences: A diary-study

GENERAL CONCLUSION................................................................................................ p. 60

REFERENCES.................................................................................................................... p. 67

INTRODUCTION

Defining workaholism

Lieber und arbeiten, which means to love and to work. Sigmund Freud answered
with these words when he was asked what a normal person should be able to do well. Thus,
apparently, Freud equated work with one of the most important value in human life (Quick,
Murphy, Hurrel, & Orman, 1992). Todays societies, indeed, are characterized by doing
cultures: they admire success and accomplishment, and consider work as a central life value
shaping our identity, self-esteem, and sense of psychological well-being (Holland, 2008).
Therefore, work tends to define us to the point of willingness to sacrifice our health and our
relationships. When did work stop to be an important feature aimed at providing basic
necessities and offering satisfaction and become a dimension which spill over into the
personal life and well-being?
With the expansion of technology, the idea of a Monday through Friday 40-hour
workweek is fading away. With the appearance of Internet, laptops, smartphones and personal
digital assistants, individuals can always stay connected to their work, working overtime, and
working from home during evenings and weekends (Aziz & Tronzo, 2011; Derks & Bakker,
2010; van Beek, Hu, Schaufeli, Taris, & Schreurs, 2012). This over-commitment of energy
and time to work, typical of the last decades, has been used in the literature to describe the
notion of workaholism.
Currently, there is not an agreed-upon definition of workaholism, rather the academic
literature is characterized by lack of consensus regarding its meaning (Griffiths, 2005).
Nevertheless, scientific interest in this topic is growing since workaholism is considered as
one of the most common addictions that can impact different areas of human functioning at
the individual, family, organizational, and societal levels (Vodanovich & Piotrowski, 2006).

Although many writers have conceptualized workaholism as pathology, it is still not

generally accepted as a clinical condition (Molino, Ghislieri, & Colombo, 2012). In fact,
workaholism is not officially listed in the Diagnostic and Statistical Manual (DSM-IV-TR).
However, workaholism is considered a symptom of the obsessive-compulsive personality
disorder, which is characterized by perfectionism, inflexibility and preoccupation with work,
and by an excessive devotion to work and productivity to the exclusion of leisure activities
and friendships (American Psychiatric Association, 2000).
Workaholics are described as unhappy and obsessive workers who do not perform their
job well, and who create difficulties for their colleagues (Burke & Ng, 2007; Porter, 1996).
Compared to non-workaholics, workaholics experience more interpersonal conflict (Mudrack,
2006) and aggressive behaviours at work (Balducci, Cecchin, Fraccaroli, & Schaufeli, 2012),
are less satisfied with their jobs (Burke & MacDermid, 1999), neglect other aspects of their
lives, such as family or friend relationships and intimacy (Bakker, Demerouti, & Burke, 2009;
Minirth, Meier, Wichern, Brewer, & Skipper, 1981; Porter, 1996; Robinson, 1989), and have
high levels of job strain and health complaints (Burke, 2000).
Ng, Sorensen, and Feldman (2007) classified workaholics as workers who commit long
hours to work and enjoy the act of working but not the work itself. However, Porter (2001)
stated that joy is not a part of workaholism since it is considered as an addiction and Mudrack
(2006) argued that enjoyment is not a core component of workaholism, although some
workaholics may enjoy their work. Therefore some authors perceived work enjoyment as
being an independent psychological phenomenon, which can be discriminated from
workaholism, called work engagement (Schaufeli, Taris, & Bakker, 2008; Taris, Schaufeli, &
Shimazu, 2010). Whereas work engagement is associated with positive outcomes and it is
essentially considered desirable, workaholism is generally linked with negative outcomes and
it is considered an undesirable phenomenon (Schaufeli, Taris, & van Rhenen, 2008; Shimazu
& Schaufeli, 2009; van Wijhe, Peeters, Schaufeli, & Ouweneel, 2012).
Oates (1971) coined the term workaholism to describe an excessive and uncontrollable
need to work that permanently disturbs health, happiness and relationships. Later
conceptualizations defined the two core elements of workaholism: the tendency to work
excessively hard and spend a great deal of time in work activities, neglecting social or family
life; and being obsessed with work because of a compulsion and not because of external
factors, showing reluctance to disengage from work (McMillan & ODriscoll, 2006;
Schaufeli, Taris, & Bakker, 2008; Scott, Moore, & Miceli, 1997). In contrast with the
prevailing perspective which considers workaholism as a compulsion or a stable individual
2

characteristic, some authors have argued that the work environment and organizational
cultures may play a role in stimulating work addiction and workaholic behaviours (Burke,
2001; Fry & Cohen, 2009; Ng et al., 2007; van Wijhe, Schaufeli, & Peeters, 2010). Only a
few studies have used the later situational approach, but it represents an important research
direction in order to better understand the workaholism genesis and to find levers for
organizational interventions.
From an organizational perspective, initially, workaholism was viewed positively. For
instance, Machlowitz (1980) distinguished between fulfilled and unfulfilled workaholics;
Buelens and Poelmans (2004) described some of them as happy hard workers; and Spence
and Robbins (1992) assumed in their workaholic-triad also the dimension of enjoyment.
However, today there is near consensus amongst scholars and researchers that workaholism
leads to a behavioural addiction, with harmful consequences for individuals, thus excluding a
positive view of the phenomenon (Porter, 1996; Sussman, 2012).
Many scholars have conceptualized workaholism as a behavioural addiction with harmful
consequences for individuals (Porter, 1996; Robinson, 2000; Sussman, 2012) having several
attributes in common with alcoholism (e.g., withdrawal symptoms, identity issues, rigidity).
Based on this perspective, Andreassen, Griffiths, Hetland, and Pallesen (2012) developed a
new unidimensional scale (Bergen Work Addiction Scale; BWAS) for the assessment of
workaholism. Unlike the previous measures of workaholism present in the literature, among
which the Workaholism Battery (Spence & Robbins, 1992), the Work Addiction Risk Test
(WART; Robinson, 1999) and the Dutch Workaholism Scale (DUWAS; Schaufeli, Taris, &
Bakker, 2006) are the most often used, the BWAS is the first scale that considers each of the
seven core elements of all addictions (Griffiths, 1996; Leshner, 1997): salience, mood
modification, (reduced) tolerance, withdrawal, conflict, relapse, and health problems.
As in the case of other addictions, the debate whether workaholism is caused by the
environment or should be seen as a dispositional characteristic has not been resolved (Porter,
1996; Schippers, 1991). However, several scholars have suggested that work-related factors
could induce or reinforce workaholic behaviours (Burke, 2001; Fry & Cohen, 2009; van
Wijhe et al., 2010): too demanding tasks and requests from employers (Maume & Bellas,
2001), incentive systems for higher productivity, a work culture strongly oriented to loyalty
and results (Piotrowski & Vodanovich, 2006), and the example of managers and supervisors
who work hard as well as their reward for excessive work behaviours (van Wijhe et al., 2010).
Moreover, organizations often provide (unintended) on-going support for existing
workaholism (Porter, 2001) and reinforce the perception of workaholics of being special
3

persons whom organizations need (van Wijhe et al., 2010). For example, a recent study found
that there is a curvilinear link between organizational identification and workaholism: when
organizational identification becomes too strong, workaholism increases (Avanzi, van Dick,
Fraccaroli, & Sarchielli, 2012).
Unfortunately, today workaholism is considered the most rewarded addiction in our
culture (Spruell, 1987). Research on this topic provided evidence that a price is paid for
dysfunctional work addiction, at the both personal and organizational levels. Several studies
have found a positive relationship between workaholism and a) burnout, stress and exhaustion
(Andreassen, Ursin, & Eriksen, 2007; Burke, 2008; Schaufeli et al., 2008; Taris, Schaufeli, &
Verhoeven, 2005); b) work-family conflict (Bakker et al., 2009; Bonebright, Clay, &
Ankenmann, 2000; Schaufeli, Bakker, van der Heijden, & Prins, 2009; Taris et al., 2005); c)
poor relationships with others at work and at home (Matuska, 2010, Robinson, Carroll, &
Flowers, 2001; Robinson & Kelley, 1998; Robinson & Post, 1997); d) career dissatisfaction
and poor performance (Holland, 2008; Graves, Ruderman, Ohlott, & Weber, 2010); e) low
self-esteem, low life satisfaction and difficulties sleeping (Andreassen, Hetland, Molde, &
Pallesen, 2011; Burke, 2004; Burke & Matthiesen, 2004); f) intentions to leave the
organization (Burke, 2001; Kravina, Falco, Girardi, & De Carlo, 2010).

Based on the relevance of this topic, the general aim of this dissertation is to gain more
insight into the phenomenon of workaholism, considering its measures, its potential
antecedents in working contexts and its consequences for individuals and organizations.
These objectives have been pursued by means of three empirical studies presented in the three
chapters respectively.

Chapter 1 focuses on workaholism measures, since few attention has been given to the
phenomenon and to instruments to measure it for research purposes so far. Specifically, the
psychometric properties of two different measures, in their Italian version, have been
examined: the Dutch Work Addiction Scale (DUWAS; Schaufeli, Shimazu, & Taris, 2009;
Schaufeli, Taris, & Bakker, 2006; Schaufeli, Taris, & Bakker, 2008) and the Bergen Work
Addiction Scale (BWAS; Andreassen et al., 2012).

Chapter 2 tries to gain more insight in workaholism by investigating its role in the job
demands-resources (JD-R) model (Bakker & Demerouti, 2007), an overarching model
suitable for various working contexts and conditions. Specifically, the second study aims at
4

understanding whether job demands are related to workaholism, and whether job resources
can moderate this relationship. Moreover, it focuses on the direct and indirect effects of
workaholism on three important life domains: health, family, and work.

Chapter 3 uses a diary study to explore the little studied relationship of workaholism
with recovery experiences (Bakker, Demerouti, Oerlemans, & Sonnentag, 2012; van Wijhe et
al., 2012) and with exhaustion. The main goal of the third study is to understand if recovery
experiences introduced by Sonnentag & Fritz (2007) play a mediational role on the
relationship between workaholism and exhaustion considering within-person variations.

Finally, general conclusion integrates and discusses the key findings of the three studies
providing methodological considerations, suggestions for future research and practical
implications.

The current dissertation has been written with the supervision of the Prof. Arnold B.
Bakker, especially for studies presented in Chapter 2 and Chapter 3, on the basis of the work
carried out during the months spent at the Erasmus University of Rotterdam.

CHAPTER 1

Workaholism measures: What choice?

The Italian version of DUWAS and BWAS

Abstract

The increasing interest about workaholism is due to important changes in working culture and
habits and to its association with negative consequences for psychological well-being. Despite
that, the knowledge about this phenomenon as well as the presence of measures to asses it are
still limited. This study contributes to the understanding of workaholism and its
measurements examining the psychometric characteristics of the Italian version of two
measures of workaholism, the Dutch Work Addiction Scale (DUWAS) and the Bergen Work
Addiction Scale (BWAS). The research involved a convenience sample of 710 workers with
different professional profiles. Results confirmed the factor structure and psychometric
properties of both the scales, as well as their construct validity showing positive correlations
with other variables workaholism-related in literature. This study provides supports for the
workaholism evaluation in the Italian context and for a better understanding of its role in
working issues and health problems.

Keywords: workaholism; measures; BWAS; DUWAS; Italian validation.

INTRODUCTION

With the changes in the working world over the past decades, characterized particularly
by both the increase in weekly working hours and the tendency to work also outside the
traditional office and outside traditional work hours (McMillan, Brady, ODriscoll, & Marsh,
2002), workaholism has emerged as a prominent topic in practitioner literature (Ng, Sorensen,
& Feldman, 2007; Smith & Seymour, 2004). Despite its relevance in modern society and,
consequently, in scientific and clinical research, poor consensus exists about its meaning and
measurement so far.
The most relevant question seems to be when hard work stops to be virtuous and
becomes a problem. The most obvious characteristics of workaholics is that they devote an
excessive amount of time and energy to their work. Since, for some scholars, the norm to
assess workaholism is the number of hours worked (Burke & Fiksenbaum, 2009; Harpaz &
Snir, 2003; Snir & Zohar, 2008). However, conceiving workaholism exclusively in terms of
the number of working hours would be incomplete because it seems to address only the
behavioral component of the phenomenon neglecting its addictive nature (Molino, Ghislieri,
& Colombo, 2012; van Wijhe, Schaufeli, & Peeters, 2010).
The objective of this study is to improve the knowledge of workaholism providing more
empirical support for the evaluation of the construct in the Italian context. Starting from an
overview of the most common measures introduced in literature so far, the current study
investigates the properties, in terms of reliability and construct validity, of two measures
considered a good and useful support for researchers and practitioners in workaholism field:
the Dutch Work Addiction Scale (Schaufeli, Shimazu, & Taris, 2009; Schaufeli, Taris, &
Bakker, 2006; Schaufeli, Taris, & Bakker, 2008) and the Bergen Work Addiction Scale
(Andreassen, Griffiths, Hetland, & Pallesen, 2012).

Measures of workaholism

Currently few measures of workaholism have been reported in academic and scientific
literature and most of them are self-report type. There are three main tested measures that
provided several studies of reliability and validity (Andreassen et al., 2012; Molino et al.,
2012; Sussman, 2012): (1) the Work Addiction Risk Test (WART; Robinson, 1989); (2) the
Workaholism Battery (Spence & Robbins, 1992); and (3) the Dutch Work Addiction Scale
(DUWAS; Schaufeli et al., 2006; Schaufeli et al., 2009).
7

The WART (Robinson, 1989) was the first quantitative measure of work addiction; its
items were drawn from a list of symptoms and characteristics reported by clinicians working
with both clients and families experiencing workaholism. The WART contains 25 items
answered on a 4-point Likert scale (from 1 = never true to 4 = always true) and consists of
five dimensions: compulsive tendencies, control, impaired communications / self-absorption,
inability to delegate, and self-worth. However, further studies on the psychometric
characteristics of the WART demonstrated that the first three factors and their 15 items had
the best discriminant validity for discriminate workaholics from non-workaholics and that the
Compulsive Tendencies (CT) subscale was the most important to work out this distinction
(Flower & Robinson, 2002). Following these research indications, Taris, Schaufeli and
Verhoeven (2005) used the WART scale by creating a Dutch version and found a high
overlap between the full 25-item WART and the 8-item CT subscale, concluding that the
shorter CT subscale was adequately representative of workaholism.
The Workaholism Battery (Spence & Robbins, 1992) is considered the most frequently
used measure of workaholism. It is composed of 25 items rated on a 5-point Likert scale
(from 1 = strongly disagree to 5 = strongly agree) and distributed between three dimensions:
(1) work involvement, referred to the extent to which individuals constructively use their
time, both on and off the job, (2) work drive, referred to the individuals internal motivation
to work, and (3) work enjoyment, referred to the degree to which individuals derive
enjoyment from work.
Starting from the CT subscale of the WART (Robinson, 1999; Taris et al., 2005) and the
Workaholism Battery (Spence & Robbins, 2002), Schaufeli et al. (2006) developed the Dutch
Work Addiction Scale (DUWAS). Originally the scale was composed of 17 items, assessed
on a 4-point Likert scale (from 1 = almost never to 4 = almost always). In line with authors
conceptualization of workaholism (Schaufeli et al., 2006; Schaufeli et al., 2008), the DUWAS
consisted of two dimensions, working excessively (WE) and working compulsively (WC): the
former is referred to the workaholics tendency to allocate exceptionally much time to work
than to other life activities and to work beyond what is reasonably expected to meet
organizational or economic requests; the latter stated that, additionally, workaholics are
obsessed with their work and persistently think about work, even when they are not working.
The WE scale (9 items) was taken from the CT subscale (Taris et al., 2005) and was
relabelled because, in DUWAS authors opinion, most of its items referred to working hard,
without any reference to the underlying motivation. Whereas, the WC scale (8 items) was
taken from the Drive scale of the Workaholism Battery (Spence & Robbins, 1992) since it
8

explicitly refers to the compulsiveness of excessive work behavior. Later, a brief DUWAS
version has been validated with five items in each scale (Table 1). It showed good
psychometric properties in different studies (del Lbano, Llorens, Salanova, & Schaufeli,
2010; Schaufeli et al., 2008; Schaufeli et al., 2009), hence the shorter version of the DUWAS
can be considered an appropriate research instrument to measure separately the two
dimensions of workaholism, namely working excessively and working compulsively.
Even though different measures of workaholism with good psychometric characteristics
are present in literature, Andreassen et al. (2012) have reported the absence of measures of
workaholism closely linked to the core elements of any addiction, focusing on the addictive
nature of workaholism. Therefore they developed a new scale for the assessment of
workaholism, named Bergen Work Addiction Scale (BWAS). The BWAS consists of 7 items
(Table 1) rated on a 5-point Likert scale (from 1 = never to 5 = always). Each item represents
one of the core elements of any addiction. According to Griffiths (1996, 2005) all addictions,
chemical or non-chemical, comprise six components: (1) salience, which means that the
addictive activity dominates thinking and behavior; (2) mood modification, referred to the
modification of addictive persons mood; (3) tolerance, as increasing amounts of the activity
are necessary to achieve initial effects; (4) withdrawal, referred to occurrence of unpleasant
feelings when the activity is discontinued or suddenly reduced; (5) conflict, which emerges in
social relationships and other life activities because of the addictive behavior; and (6) relapse,
which means the tendency for reversion to earlier patterns of the activity after abstinence or
control. In addition, some authors (Andreassen et al., 2012; Griffiths, 2005; Leshner, 1997)
argued that the addictive behavior represents also different kind of health and/or other
problems for the person. Hence, the BWAS comprises all of these seven core elements of
addictions: item 1 represents salience; item 2 represents tolerance; item 3 represents mood
modification; item 4 represents relapse; item 5 represents withdrawal; item 6 represents
conflict; and item 7 represents problems. The BWAS showed relatively high content validity
in terms of the addiction field and an adequate factor structure, representing the first
unidimensional scale for the assessment of workaholism (Andreassen et al., 2012).
Basing on this overview of workaholism measures, two scales seem to be the most
important and relevant nowadays, namely the DUWAS and the BWAS. Indeed, they are short
measures with good psychometric properties, therefore they can be considered useful
instruments in research as well as in clinical contexts. To support their application also in the
Italian context, the aim of this study is to investigate the psychometric characteristic of the
Italian version of both DUWAS, comprising WE and WC subscales, and BWAS.
9

Hypothesis 1: both BWAS and DUWAS Italian versions show good psychometric
properties in terms of factor structure and internal validity.

To examine the convergent and discriminative validity of the two scales, the correlation
of their latent scores with each other is tested. We expect a high correlation between WE
subscale and WC subscale, as well as between them and BWAS, since they measure the same
construct. We also expect significantly different means of the three scales, as they consider
different dimensions of workaholism (the working excessively component, the working
compulsively component and the overall addictive component).

Hypothesis 2a: WE and WC subscales and BWAS Italian versions show high
correlations between each other.
Hypothesis 2b: WE and WC subscales and BWAS Italian versions report significantly
different means between each other.

Finally, to examine the construct validity of the two scales and to confirm the negative
nature of workaholism (del Lbano et al., 2010) this study tests the patterns of correlations
between DUWAS subscales, BWAS and other dimensions presented in workaholism-related
literature (e.g. Andreassen, Ursin, & Eriksen, 2007; Bakker, Demerouti, & Burke, 2009;
Burke, 2000; Taris et al., 2005; McMillan & ODriscoll, 2004; Aziz & Cunningham, 2008;
Schaufeli et al., 2008). Moreover, also the capacity of DUWAS subdimensions and BWAS to
distinguish among individuals on the basis of some demographic variables (gender and
professional profile) is tested. Finally, regarding BWAS, this study examines the cut-off
criterion provided by the authors to differentiate between workaholics and non-workaholics in
terms of the number of hours worked per week and health problems (Andreassen et al., 2012).

Hypothesis 3a: WE subscale, WC subscale and BWAS Italian versions show high
correlations with other constructs workaholism-related: exhaustion, disengagement,
job satisfaction, work-family conflict and job demands (workload, cognitive demands,
emotional demands, emotional dissonance, total weekly work hours).
Hypothesis 3b: WE subscale, WC subscale and BWAS Italian versions report
significantly different means between males and females and between employees and
self-employed workers.

10

Hypothesis 3c: the proposed cut-off score for the BWAS discriminates between
respondents on the weekly working hours and exhaustion: workaholics work
significantly more hours than non-workaholics; workaholics perceive more exhaustion
than non-workaholics.

Table 1. BWAS and DUWAS items

Original items

Italian translations
DUWAS Working Excessively

I seem to be in a hurry and racing against the clock.

Mi sembra di essere sempre di fretta, in corsa contro

il tempo.

I stay busy and keep many irons in the fire.

Sono molto occupato e metto troppa carne al fuoco.

I find myself doing two or three things at one time

such as eating lunch and writing a memo, while
talking on the phone.
I find myself continuing to work after my coworkers
have called it quits.

Mi ritrovo a fare
contemporaneamente.

I spend more time working than on socializing with

friends, on hobbies, or on leisure activities.

Trascorro pi tempo lavorando che stando con gli

amici o dedicandomi ai miei hobby.

4
5

due

tre

cose

Continuo a lavorare anche quando gli altri mi

dicono di smettere.

DUWAS Working Compulsively

6

Its important for me to work hard even when I

dont enjoy what Im doing

importante per me lavorare sodo anche quando

non mi piace quello che sto facendo

I often feel that theres something inside me that

drives me to work hard

Spesso sento che c qualcosa in me che mi spinge a

lavorare sodo

I feel obliged to work hard, even when its not

enjoyable

Mi sento obbligato a lavorare sodo, anche quando

non piacevole

I feel guilty when I take time off work

Mi sento in colpa quando mi prendo una pausa dal

lavoro

10

It is hard for me to relax when Im not working

difficile per me rilassarmi quando non sto

lavorando

Likert frequency scale from 1 (Almost) never to 4

(Almost) always

Scala di risposta Likert da 1 (Quasi) mai a 4

(Quasi) sempre

BWAS
How often during the last year have you

Con quale frequenza durante lultimo anno lei ha

Thought of how you could free up more time to

work?

Pensato a come avrebbe potuto riservare pi tempo

per il lavoro?

Spent much more time working than initially

intended?

Trascorso molto pi tempo al lavoro di quanto

inizialmente previsto?

Worked in order to reduce feelings of guilt, anxiety,

helplessness and depression?

Lavorato con lo scopo di ridurre i sentimenti di

colpa, di ansia, di incapacit e di depressione?

Been told by others others to cut down on work

without listening to them?

Percepito che gli altri Le dicevano di ridurre il

lavoro e Lei non li ha ascoltati?

11

Become stressed if you have been prohibited from

working?

Provato forte disagio perch le stato proibito di

lavorare?

Deprioritized hobbies, leisure activities, and exercise

because of your work?

Dato minore priorit agli hobby, alle attivit di

svago ed esercizio fisico a causa del Suo lavoro?

Worked so much that is has negatively influenced

your health?

Lavorato cos tanto che ci ha influenzato

negativamente la Sua salute?

Likert frequency scale from 1 Never to 5 Always

Scala di risposta Likert da 1 Mai a 5 Sempre

METHOD

Participants and procedures

The research involved a convenience sample of 710 Italian workers contacted throughout
snowball exercise. They filled out a self-report on-line questionnaire. Among the participants
378 were females (53% of the sample) and 330 were males (47% of the sample). Their mean
age was 41.52 (SD = 9.67); 62% were married or cohabited. In the sample, 78% had a
bachelors, masters degrees or a higher educational qualification.
Participant were from different occupational sectors (most of them from the private
service, 48%). In total, 46% were employees and 53% were self-employed (missing cases =
1%). Weekly working hours were, on average, 42.76 (SD = 10.69). Mean seniority on the job
was 13.69 years (SD = 9.73).

Measures

Workaholism was assessed by two different measures: (1) the 10-item DUWAS (del
Lbano et al., 2010; Schaufeli et al., 2008) and (2) the 7-item BWAS (Andreassen et al.,
2012). The DUWAS is distributed in two subdimensions, working excessively (measured by
items from 1 to 5) and working compulsively (measured by items from 6 to 10). As in the
original version, items were applied using a 4-point Likert scale (from 1 = almost never to 4 =
almost always). The five items of WC dimension were translated into Italian; subsequently a
bilingual translator performed a back translation of the items which was compared with their
original version (Beaton, Bombardier, Guillemin, & Ferraz, 2000). The Italian version of the
five items of WE dimension was taken from a previous work of validation (Molino et al.,
2012). As regards BWAS, each item measures one of the 7 components of addiction referred
12

to above (Andreassen et al., 2012; Griffiths, 1996). As in the original scale, items were rated
on a 5-point Likert scale (from 1 = never to 5 = always). The Italian version of the 7 items
was supplied by the authors.

Exhaustion and disengagement were measured by the 16-item Oldenburg Burnout

Inventory (OLBI; Demerouti, Mostert, & Bakker, 2010). Typical items are There are days
when I feel tired before I arrive at work for exhaustion and It happens more and more often
that I talk about my work in a negative way for disengagement (from 1 = strongly disagree
to 4 = strongly agree); Cronbachs alpha for this study was .76 for exhaustion and .79 for
disengagement.

Job satisfaction was assessed by 4 items taken by the COPSOQ II - Job satisfaction scale
(Pejtersen, Kristensen, Borg, & Bjorner, 2010). A typical item is How satisfied are you about
the way your abilities are used? (from 1 = very unsatisfied to 5 = very satisfied); Cronbachs
alpha for this study was .85.

Work-family conflict was measured with the Italian version (Colombo & Ghislieri, 2008)
of the 5-item scale taken by Netemeyer, Boles, and McMurrian (1996). Example item is The
demands of my work interfere with my home and family life (from 1 = never to 6 = always);
Cronbachs alpha for this study was .90.

Some general job demands were also measured in the questionnaire. Workload was
assessed by using the 4-item scale taken from Bakker, Demerouti, and Verbeke (2004). An
example item is Do you have too much work to do? (from 1 = never to 5 = always);
Cronbachs alpha for this study was .79. Cognitive demands were evaluated with a 4-item
scale of Bakker, Demerouti, Taris, Schaufeli, and Schreurs (2003a). A typical item of this
scale is Does your work demand enhanced care or precision? (from 1 = never to 5 =
always); Cronbachs alpha was .77. Emotional demands were measured with 3 items (Bakker,
Demerouti, & Schaufeli, 2003b), including Do you face emotionally charged situations in
your work? (from 1 = never to 5 = always); Cronbachs alpha was .86. Emotional dissonance
was assessed with the 3-item scale developed by Zapf, Vogt, Seifert, Mertini, and Isic (1999).
Sample item is How often during your work do you have to display emotions which do not
correspond to your inner feelings (from 1 = never to 6 = always); Cronbachs alpha was .86.

13

Statistical analysis

Descriptive analysis (M, SD, Asymmetry, Kurtosis) of each item of the DUWAS and
BWAS was performed on the whole sample. At a later stage, the psychometric characteristics
of the two scales were examined both through an exploratory factor analysis (EFA ULS
extraction) with PASW 18 and through a confirmatory factor analysis (CFA) performed by
Mplus 7 (Muthn & Muthn, 1998-2012). The CFA method of estimation was maximum
likelihood (ML) for the DUWAS scale and the maximum likelihood mean adjusted (MLM)
for the BWAS, as it showed a not-normal distribution on the basis of Asymmetry and
Kurtosis of its items. The overall sample (N = 710) was randomly split into two subsamples
homogeneous for some demographic characteristics (gender, marital status, professional
profile and occupational sector): the EFA was performed on the first subsample (N = 354) and
the CFA on the second subsample (N = 356). According to the literature (Bollen & Long,
1993) several goodness-of-fit criteria were considered: the 2 goodness-of-fit statistic; the
Root Mean Square Error of Approximation (RMSEA); the Comparative Fit Index (CFI); the
Tucker Lewis Index (TLI); the Standardized Root Mean Square Residual (SRMR).
As measures of reliability, Cronbachs alpha and the corrected item-total correlation
coefficients were calculated for the two subscales of the DUWAS and for the BWAS on the
whole sample. Moreover, paired sample t-test was calculated in order to verify a significant
difference between the mean of the two DUWAS subdimensions, WE and WC, and between
them and the mean of the BWAS. Analysis of variance (t-test for independent samples) has
been calculated based on some demographic variables (gender and professional profile) in
order to evaluate the capability of the workaholism scales to discriminate among different
groups. Finally, to investigate the concurrent validity of the DUWAS subdimensions and
BWAS also correlations between them and other constructs, indicated in literature
(Andreassen et al., 2012; del Lbano et al., 2010; Molino et al., 2012; Schaufeli et al., 2009)
as potentially workaholism related were tested: exhaustion, disengagement, job satisfaction,
work-family conflict, workload, cognitive demands, emotional demands, emotional
dissonance, work hours per week.
In the case of the BWAS, authors indication (Andreassen et al., 2012) have been
followed to set a cut-off score for categorization of workaholics. The criterion required at
least four as answer on at least 4 out of 7 items of the scale. As in the original validation,
the sample was split in workaholics and non-workaholics and analysis of variance was

14

calculated to test if they reported significantly different number of working hours per week
and scores on well-being variable (exhaustion in this study).

RESULTS

Descriptive statistics

Data screening analysis was conducted to check deviation from normality in all the
scales; Table 2 shows M, SD, Asymmetry and Kurtosis of each item. The DUWAS items
showed a relatively normal distribution, with all values of Asymmetry and Kurtosis
comprised in the range - 1.0 to + 1.0, except for item number 4, which showed a kurtosis
index of -1.04. Therefore, the DUWAS items could be analyzed by normal theory estimators
and not much distortion was to be expected (Muthn & Kaplan, 1985).

Table 2. Descriptive statistics of single items of BWAS and DUWAS scales

Items

SD

Kurtosis

Asymmetry

DUWAS
1

3.27

.83

-.25

-.83

2.84

.85

-.52

-.32

3.30

.78

.25

-.91

2.46

.99

-1.04

.08

3.01

.98

-.80

-.58

2.84

.90

-.67

-.35

2.78

.89

-.61

-.34

2.75

.92

-.73

-.31

2.01

1.03

-.85

.61

10

1.84

.98

-.51

.83

.75

BWAS
1

2.07

1.09

-.34

3.03

1.34

-.80

-.11

1.97

1.18

.01

1.02

2.05

1.13

-.15

.87

1.40

.81

5.32

2.32

3.10

1.29

-1.09

-.06

2.06

1.15

-.37

.82

15

The BWAS items did not show a normal distribution, as Asymmetry and Kurtosis exceeded
the range from - 1.0 to + 1.0, particularly in the case of item number 5 (Asymmetry value =
2.32; Kurtosis value = 5.32) which showed also the lowest mean and standard deviation.
Perhaps this item, referred to the withdrawal component of the addiction, could be less
applicable, than others, to the work situation. Nevertheless, we decided to keep it and to
investigate the measurement model of the overall scale developed by the authors in the later
analysis.

Exploratory factor analysis

Exploratory factor analysis (ULS extraction) was performed on sample 1 (N = 354). The
KaiserMeyerOlkin (KMO) measure of sampling adequacy and the Bartlett's test of
sphericity indicated that conducting factor analysis on the data was appropriate for both
DUWAS and BWAS (Dziuban & Shirkey, 1974): KMO was .81 and Bartlett's test was
1002.19 (df = 45; p < .001) for DUWAS; KMO was .80 and Bartlett's test was 595.46 (df =
21; p < .001) for BWAS.
The chosen factor solution for DUWAS (Table 3) resulted in two factors (Promax
rotation), in accordance with literature (del Lbano et al., 2010; Schaufeli et al., 2008). This
solution explained 41.22% of the variance. Both the factors showed acceptable saturations
corresponding to all items, with factor loadings above the conventional cut-off value of .40
(Hinkin, 1998). Items number 4 and 5 showed a moderately high loading on both factors,
anyhow higher for WE than for WC according with literature (del Lbano et al., 2010;
Schaufeli et al., 2008). The two factors correlated in a high and significant way (r = .48; p <
.001).
The chosen factor solution for BWAS (Table 4) resulted in one factor and explained
34.06% of the variance. All the factor loadings surpassed the conventional cut-off value of .40
(Hinkin, 1998), including item number 5 with a factor loading equal to .44.

16

Table 3. EFA solution (ULS extraction, Promax rotation) for DUWAS, 41.22% explained variance (N = 354)
Items

Working
Excessively

Working
Compulsively

.75

.27

.73

.32

.65

.32

.53

.48

.51

.45

.40

.76

.22

.70

.28

.65

.35

.44

10

.35

.43

Correlation between factors

Working Excessively

Working Compulsively

.48

Table 4. EFA solution (ULS extraction) for BWAS, 34.06% explained variance (N = 354)
Items

BWAS

.67

.65

.64

.62

.62

.44

.40

Confirmatory factor analysis

Confirmatory factor analysis was performed using Mplus 7 on the sample 2 (N = 356). In
both the scales, CFA confirmed results of EFA.
The solution of DUWAS fitted adequately to the data, confirming the two-factor
structure of the scale: 2 (34) = 117.60, p < .001; RMSEA = .08; CFI = .92; TLI = .90; SRMR
= .06. Figure 1 shows the standardized solution, with factor loadings ranging from .37 to .79.
This solution showed also a better fit to the data compared with the one-factor solution: 2
(35) = 248.17, p < .001; RMSEA = .13; CFI = .80; TLI = .74; SRMR = .08; 2 = 130.57, p <
.001.

17

Figure 1. CFA (ML estimation) standardized solution for DUWAS (N = 356)

Item 1

.42

Item 2

.40

Item 3

.52

Item 4

.62

Item 5

.65

Item 6

.52

Item 7

.53

Item 8

.37

Item 9

.82

Item 10

.87

.76
.77
1.00

Wk
Excess.

.69
.61
.59

.61

.69
.69
1.00

Wk
Comp.

.79
.43
.37

The one-factor solution of BWAS showed a satisfactory fit to the data: 2 (14) = 45.36, p
< .001; RMSEA = .07; CFI = .94; TLI = .90; SRMR = .04. Standardized factor loadings
ranged from .49 to .69, as Figure 2 shows. CFA on sample 2 also showed a satisfactory factor
loading for item number 5, equal to .54.

18

Figure 2. CFA (MLM estimation) standardized solution for BWAS (N = 356)

.49

Item 1

.76

Item 2

.65

Item 3

.55

Item 4

.53

Item 5

.71

Item 6

.61

Item 7

.59

.59
.67
1.00

BWAS

.69
.54
.63

.64

Reliability

Internal consistency of the measures of workaholism was good, since all values of
Cronbachs Alpha met the criterion of .70 (Nunnally & Bernstein, 1994): as Table 5 shows,
alpha coefficient was .79 for WE, .73 for WC and .79 for BWAS. The Alpha-if-item-weredeleted values revealed that any item decreased the Alpha value in all the cases. Moreover,
corrected item-total correlations ranged from .51 to .62 for WE, from .40 to .60 for WC and
from .40 to .59 for BWAS. Therefore, all scores were above the cut-off value of .40
(Nunnally, 1967).

Paired sample t-test and correlations between workaholism measures

Paired sample t-test showed a significant difference between the means of the two
subscales of DUWAS [t(709) = 22.05, p < .001]: the mean of WE (M = 2.98, SD = .65) was
significantly higher than the mean of WC (M = 2.44, SD = .65); they also showed a significant
positive correlation (r = .51, p < .001). Moreover, we compared the mean of BWAS with the
means of the two DUWAS subdimensions converting the original variables into new variables
with the same range (from 1 to 10), as we used two different Likert-scale to measure them in

19

the questionnaire. Paired sample t-test showed a significant difference between the means of
BWAS and both WE [t(709) = -54.83, p < .001] and WC [t(709) = -29.32, p < .001]: the mean
of both WE (M = 7.44, SD = 1.64) and WC (M = 6.11, SD = 1.63) were significantly higher
than the mean of BWAS (M = 4.48, SD = 1.48). The correlation between BWAS and both
WE (r = .58, p < .001) and WC (r = .55, p < .001) was positive and significant. Therefore, we
can conclude that the three measures showed convergent validity with each other, but they
effectively measure different components of workaholism.

Analysis of variance and correlations with other workaholism related dimensions

In order to test the concurrent validity of the DUWAS and BWAS Italian versions,
correlation between workaholism and other constructs potentially workaholism-related were
computed (Table 5). All the three workaholism dimensions significantly and positively
correlated with exhaustion and work-family conflict. BWAS and WE also positively
correlated with disengagement and negatively correlated with job satisfaction, albeit more
weakly. Additionally, the three measures of workaholism were positively related to the job
demands considered in this study. More specifically, correlation with workload was
particularly high for WE as well as the correlation with cognitive demands, whereas
correlation with emotional dissonance was higher for BWAS and WC; correlations with
emotional demands were higher for BWAS than for the two DUWAS subdimensions. Finally,
positive correlation with work hours per week was stronger for WE than for WC; also BWAS
showed a positive significant correlation with the total weekly hours worked, lower than WE
and higher than WC.
Differences in the perception of workaholism on the basis of some demographic
variables were evaluated through t-test for independent sample. Results showed higher level
of both WE [t(664) = 2.57, p < .05] and WC [t(706) = 2.40, p < .05] for females (M = 3.04,
SD = .62; M = 2.50, SD = .67) compared to males (M = 2.91, SD = .69; M = 2.38, SD = .63).
BWAS did not show differences on the basis of gender. Moreover, self-employed workers
perceived more workaholism than employees. Specifically, results showed higher level of WE
[t(699) = -2.11, p < .05] and BWAS [t(699) = -2.52, p < .05] for self-employed workers (M =
3.03, SD = .66; M = 2.31, SD = .74 respectively) compared to employees (M = 2.93, SD = .65;
M = 2.17, SD = .74 respectively).

20

Table 5. Correlations (Pearsons r), means, standard deviations and reliabilities of all variables
1
1. BWAS

10

11

12

(.79)

2. Working Excessively

.58**

(.79)

3. Working Compulsively .55** .51**

(.73)

4. Exhaustion

.45** .33** .35**

5. Disengagement

.13** .01

.10** .57**

6. Job satisfaction

-.15** .03

-.15** -.53** -.70** (.85)

7. Work-family conflict

.59** .61** .38** .43** .08*

8. Workload

.45** .61** .35** .29** .07

9. Cognitive demands

.22** .29** .22** .11** -.19** .13**

.28** .40**

10. Emotional demands

.22** .16** .10** .16** -.11** .03

.22** .10** .28**

11. Emotional dissonance

.34** .22** .30** .31** .27** -.29** .30** .27** .12** .25**

12. Work hours per week

.20** .27** .12** .04

-.04

.13**

.33** .30** .11** .02

2.24

2.98

2.44

2.33

2.24

3.37

3.38

3.57

4.10

3.18

3.05

42.76

SD

.74

.65

.65

.54

.60

.88

1.21

.77

.68

1.03

1.26

10.69

(.76)
(.79)

-.09*
-.03

(.90)
.51**

(.79)
(.77)
(.86)
(.86)
.03

Note. Cronbachs Alpha on the diagonal. * p < .05; ** p < .01.

Cut-off score analysis

On the basis of the criterion requiring a score of at least four as answer on at least 4 out
of 7 BWAS items, we split the whole sample in workaholics and non-workaholics: a total of
10.40% of participants were categorized as workaholics. Those categorized as workaholics
reported total weekly work hours to be 45.91 (SD = 12.76), significantly higher compared to
those categorized as non-workaholics (M = 42.39; SD = 10.38) [t(703) = 2.69, p < .005].
Furthermore, they showed significant differences also on exhaustion: workaholics had higher
scores (M = 2.75; SD = .58) than non-workaholics (M = 2.28; SD = .51) [t(708) = 7.31, p <
.001]. No differences emerged about disengagement, whereas job satisfaction was lower for
workaholics (M = 3.17; SD = 1.00) than for non-workaholics (M = 3.40; SD = .86) [t(702) = 2.15, p < .05].
Comparing the DUWAS subdimensions among workaholics and non-workaholics we
found significant differences in both of the cases: workaholics showed higher levels of WE
(M = 3.59; SD = .40) than non-workaholics (M = 2.91; SD = .64) [t(122.28) = 12.94, p < .001]
and they also showed higher level of WC (M = 3.05; SD = .64) than those non-workaholic
persons (M = 2.37; SD = .62) [t(708) = 8.93, p < .001].

21

DISCUSSION AND CONCLUSION

The aim of this study was to test the psychometric characteristics of the Italian version of
two workaholism measures presented in literature recently: the DUWAS (Schaufeli et al.,
2006; 2008; 2009) and the BWAS (Andreassen et al., 2012). These measures seem to be
particular relevant nowadays, since DUWAS provided several studies of reliability and
validity and permits to measure two different components of workaholism, the excessive and
the compulsive ones; whereas, BWAS represents the first unidimensional workaholism scale
which takes into account the addictive nature of the construct, assessing the seven core
elements of any addiction (Andreassen et al., 2012; Griffiths, 2005).
The present study showed the psychometric goodness of both the scales in their Italian
version, confirming Hypothesis 1. On the one hand, 10-item DUWAS presented the twofactor structure obtained in previous studies (del Lbano et la., 2010; Schaufeli et al., 2008)
with the first five items measuring working excessively factor and the last five items assessing
working compulsively factor. Both of the subdimensions showed acceptable internal
reliability. On the other hand, BWAS presented a clear unifactorial structure with satisfactory
reliability confirming its items ability to assess the seven core elements of the addiction
throughout an overall measure.
The convergent and discriminative validity of the two scales (Hypotheses 2a and 2b) was
confirmed. As expected, results showed relatively high correlations between the two DUWAS
subdimensions and also between BWAS and both WE and WC; therefore we can conclude
that they can be considered measures of the same construct, namely workaholism. At the
same time, this study demonstrated that the three measures effectively refer to different
dimensions of workaholism, since they showed significantly different means on the same
sample. Therefore, they provide the possibility to focus on alternatively the great deal of time
that the workaholics spend doing their work (working excessively), the workaholics
reluctance to disengage from work persistently and frequently thinking about work, even
when they are not working (working compulsively), the overall addictive component of
workaholic behaviours, as well as of any other addictions (BWAS).
Moreover, the construct validity of the two scales was investigated. First of all, results
confirmed the negative nature of workaholism (Andreassen et al., 2012; del Lbano et al.,
2010; Molino et al., 2012; Taris et al., 2005) since it was related with antecedents and
consequences indicated in literature as workaholism-related, providing support for Hypothesis
3a. Specifically, according with previous studies, both DUWAS subdimensions and BWAS
22

positively correlated with exhaustion (e.g. Andreassen et al., 2007; Schaufeli et al., 2008;
Schaufeli, Taris, & van Rhenen, 2008), with work-family conflict (Bakker et al., 2009;
Bonebright, Clay, & Ankenmann, 2000; Schaufeli et al., 2009, Taris et al., 2005) and with
some general job demands (Schaufeli et al., 2008; Schaufeli et al., 2009). Moreover, WE and
BWAS also positively correlated with disengagement and negatively correlated with job
satisfaction (Burke, 2000; del Lbano et al., 2010). Finally, work hours per week positively
correlated with all the three workaholism dimensions and, as expected (Schaufeli et al., 2008),
this correlation was stronger for WE than for WC.
Furthermore, the scales proved to be sensitive to some demographical differences
consistently with previous studies: females showed grater levels of workaholism than males
(Molino et al., 2012) as regards both WE and WC subdimensions; self-employed workers
perceived more workaholism than employees (Gorgievski, Bakker, & Schaufeli, 2010;
Schaufeli et al., 2008) in cases of both WE and BWAS. Therefore, the study pointed out the
ability of the considered measures to distinguish among individuals on the basis of some
demographic variables, confirming Hypothesis 3b.
Finally, cut-off analysis indicated by BWASs authors confirmed also in the Italian
version of the scale its ability to differentiate between workaholics and non-workaholics
(Andreassen et al., 2012): the suggested cut-off of "four" distinguished well between
participants in terms of both working hours and levels of exhaustion, confirming that
workaholics work more hours per week and report poorer psychological well-being than nonworkaholics, according with Hypothesis 3c of the study. Moreover workaholics showed also
lower levels of job satisfaction but not significantly lower levels of disengagement. Finally,
the significant differences on WE and WC confirmed the ability of the cut-off score analysis
indicated to differentiate between workaholics and non-workaholics.

Limitations

One of the limitations of this study concerns sample selection, since we used a
convenience sample contacted throughout snow-ball exercise. Therefore the sample may be
not representative of the Italian working population at large. Moreover, this study considered
only single-source self-report data, which means the possibility of common method bias
(Podsakoff, MacKenzie, Lee, & Podsakoff, 2003). Behavioural and organizational measures,
observation of work behaviour and objective indicators are needed in future studies in order to

23

minimize the potential effects of common method variance and to more completely
understand how well the two measures assess workaholic behaviour.
A further limitation of this study is that it tested a cut-off score to discriminate between
workaholics and non-workaholics only for the BWAS and not for the DUWAS, focusing
more on the dimension, among the two, less analyzed in literature so far. Indeed, several
studies (Schaufeli et al., 2008; Schaufeli et al., 2009) showed the ability of the DUWAS
subdimensions to distinguish among individuals on the basis of workaholism.
Nevertheless, future studies should examine the psychometric properties and construct
validity of both scales in greater details. For instance, the reliability of the measures should be
investigated by means of a test-retest procedure and the structural invariance of the scales
should be cross-validated on separate samples on the basis of different professions, working
contexts or demographic variables. It would be also necessary to evaluate the scales in more
complex and longitudinal research design, and integrate results with multidisciplinary
approaches and qualitative methods.

Practical implications

This study may represent an advance in knowledge and understanding about

workaholism, since it provides more empirical support to researches and also practitioners in
assessing it, especially in the Italian context where few attention is given to the phenomenon
and a lack of instruments is reported (Molino et al., 2012).
Results of the current study demonstrated that both 10-items DUWAS and 7-item BWAS
can be used in Italy among workers from different professions for assessing and monitoring
levels of workaholism, in terms of working excessively, working compulsively and more in
general work addiction. Indeed, both the scales confirmed the structure proposed by the
authors showing that they can be alternatively applied on the basis of specific research
purposes. Starting from authors conceptualization of workaholism, the DUWAS seem to be
more indicated in those studies where it is important to distinguish between the compulsive
and the behavioural components of the construct. Whereas the BWAS represents a
unidimensional scale useful especially for those researches and practitioners interested in the
addictive dimension of workaholism.
In general, both the scales, given their small size and good psychometric characteristics,
represent practical instruments for research purposes providing support in both detecting
workaholism and assessing its relationship with some antecedents and consequences.
24

CHAPTER 2

The role of workaholism in the job demands-resources model

Abstract

The present study among 617 Italian workers (employees and self-employed) investigated
possible antecedents and consequences of workaholism. Using the JD-R model, we
hypothesized that job demands would be positively related to workaholism, particularly when
job resources are low. In addition, we hypothesized that workaholism would be negatively
related (directly or indirectly) to three important life domains: health, family, and work. The
results of (moderated) structural equation modeling showed that workload, cognitive
demands, emotional demands, and customer-related social stressors were positively related to
workaholism and work-family conflict (partial mediation). As hypothesized, job resources
(job security and opportunities for development) buffered the relationship between job
demands and workaholism. Additionally, workaholism was indirectly related to exhaustion
and intention to change jobs through work-family conflict. These findings suggest that
workaholism is a function of a suboptimal work environment and predicts unfavourable
employee outcomes. Theoretical and practical implications of these findings are discussed.

Keywords: JD-R model; job demands; job resources; workaholism; work-family conflict.

25

INTRODUCTION

Nowadays, the term workaholism is widely used among lay people, and the scientific
interest in this topic is increasing. Nevertheless, our knowledge about workaholism is still
limited, because there is little consensus about its meaning and because few studies have dealt
with the processes involved in the workaholism genesis (Schaufeli, Bakker, van der Heijden,
& Prins, 2009). The literature seems to lack studies investigating the relationship between
workaholism and working conditions, although some authors have argued that organizations
are becoming workaholics themselves (Schaef & Fassel, 1988). Indeed, organizations
increasingly push their employees to work harder and longer to remain successful in the
global competition (Fry & Cohen, 2009). In this way, organizations seem to aim for shortterm results, neglecting possible detrimental long-term effects of working excessively, in
terms of reduced job and life satisfaction, impaired health and well-being, and, consequently,
decreased efficiency and productivity.
Based on an addiction approach of workaholism and considering that the literature has
centred mainly on its consequences, the present study aims to better understand the
antecedents and consequences of workaholism by investigating its role in the job demandsresources (JD-R) model (Bakker & Demerouti, 2007). Specifically, the main goal of the study
is to understand whether job demands are related to workaholism, and whether job resources
can moderate this relationship. An affirmative answer would highlight the potential role of
organizations in both fostering and preventing the occurrence of workaholism. Moreover, the
study aims to investigate the direct and indirect effects of workaholism on three important life
domains: health, family, and work.

Job demands-resources model, workaholism and other outcomes

According to the JD-R model (Bakker & Demerouti, 2007), every work environment has
its own unique characteristics that can be classified in two general categories, job demands
and job resources, providing an overarching model suitable for various working contexts. Job
demands represent physical, psychological, social or organizational characteristics of the job
that require sustained physical and/or psychological effort and are therefore associated with
physiological and/or psychological costs; demands will potentially evoke strain if they exceed
the employees adaptive capability (Bakker & Demerouti, 2007; Bakker, Hakanen,
Demerouti, & Xanthopoulou, 2007). Examples of job demands are work pressure,
26

emotionally demanding interactions with clients, and complex tasks that challenge employees
cognitively (Demerouti, Bakker, Nachreiner, & Schaufeli, 2001).
Job resources refer to physical, psychological, social or organizational job aspects that
may: be functional in achieving work-related goals; reduce job demands and the associated
costs; and stimulate personal growth and development (Demerouti et al., 2001; Bakker et al.,
2007). Job resources may be located at the organizational level (e.g., job security,
opportunities for professional development), at the interpersonal level (e.g., supervisor and
colleague support, feedback from the supervisor), at the level of organisation of work (e.g.,
role clarity, participation in decision making), and at the level of the task (e.g., skill variety,
autonomy, performance feedback; Tims & Bakker, 2010).
The main assumption of the JD-R model is that the risk of job strain is highest in
working environments where job demands are high and job resources are limited (Bakker &
Demerouti, 2007; Demerouti et al., 2001). Complementary to this effect, the buffer hypothesis
states that high job resources may offset the negative impact of job demands on employee
well-being, including burnout (Bakker, Demerouti, Taris, Schaufeli, & Schreurs, 2003a;
Xanthopoulou, Bakker, Demerouti, & Schaufeli, 2007). The central question of the present
study is whether workaholism is an addiction related to working conditions. More
specifically, we investigate whether job demands may influence workaholism and whether
certain specific job resources can buffer this relationship. Thus, the present study extends the
buffer hypothesis of the JD-R model already introduced for job strain and burnout (Bakker,
Demerouti, & Euwema, 2005; Demerouti & Bakker, 2011; Xanthopoulou et al., 2007) to
workaholism. These hypothesized interactions will be tested combining four job demands
(workload, cognitive demands, emotional demands, and customer related social stressors),
that have been found to be important in many occupational groups (Bakker & Demerouti,
2007), with two job resources (opportunities for professional development and job security).
Opportunities for professional development are an important job resource because they may
satisfy workers basic need for competence and contribute to their intrinsic motivation to
achieve results (Ryan & Deci, 2000). Adequate opportunities for professional development
ensure that workers are capable of dealing with their job demands and remain engaged
without feeling compelled to create more challenges at work or take over tasks to improve
their skills (Tims & Bakker, 2010). In this way, they can satisfy job requests without
increasing risks of workaholic behaviours. Likewise, job security may be considered a
resource able to moderate the impact of job demands on workaholism. Indeed, in the current
working world characterized by competitiveness and job-related uncertainty, perceived
27

economic insecurity could create a sense of urgency in work and promote workaholic
behaviours (Matuska, 2010) as a reaction to workers fear of losing their jobs, especially in
the Italian context.

Hypothesis 1a: Job demands (workload, cognitive demands, emotional demands and
customer related social stressors) are positively related to workaholism.
Hypothesis 1b: Job resources (opportunities for professional development and job
security) moderate the positive relationship between job demands and workaholism.
Specifically, the relationship between job demands and workaholism will be more
positive for employees who have low (vs. high) job resources.

This study also aims at improving our knowledge about the consequences of
workaholism for individual employees, their family and organizations at large. Regarding
consequences for individuals, several studies have found a positive relationship between
workaholism and burnout, a state of exhaustion and depletion of mental resources
(Andreassen, Ursin, & Eriksen, 2007; Burke, 2008; Schaufeli, Taris, & Bakker, 2008; Taris,
Schaufeli, & Verhoeven, 2005). When workaholics spend excessive amounts of energy and
effort at work, they might exhaust their energy back up and burn out (Bakker, Demerouti,
Oerlemans, & Sonnentag, 2012; Sonnentag & Zijlstra, 2006; van der Hulst & Geurts, 2001).
There is wide consensus that workaholics constitute a risk group for ill-health (Kanai, 2006),
as both the number of hours they work and their reluctance to disengage psychologically from
work rise effort expenditure and limit their opportunities for recovery, increasing the risk of
burnout.
Empirical research has shown that workaholism adversely impacts also the non-work
domain, increasing work-family conflict (WFC; Bakker et al., 2009; Bonebright, Clay, &
Ankenmann, 2000; Taris et al., 2005). Indeed, workaholics spend a lot of time on their work,
also in the evening and weekend, at the cost of other life activities and social relations. Their
ability to balance demands coming from work and personal life consequently decreases as
they are highly engaged in work. Moreover, from a personal resources perspective supported
by the Conservation of Resources theory (Hobfoll, 2002), workaholics tendency to devote
more resources (e.g., time, efforts) to work leaves them with fewer resources for their family
life.
Thus, on the basis of the previous hypothesis, namely that job demands have a positive
relationship with workaholism, and on the basis of the abovementioned negative
28

consequences of workaholism, the study investigates the relationship between workaholism

and some of its outcomes and hypothesizes a mediational role of workaholism between job
demands on the one hand, and both WFC and exhaustion on the other hand (see Figure 1).
The mediational effects hypothesized are partial in both of the cases, since we also assume a
positive relation between job demands and both WFC (cf. Demerouti, Geurts, & Kompier,
2004; Michel et al., 2011; Schaufeli et al., 2009) and exhaustion (Demerouti et al., 2001;
Xanthopoulou et al., 2007). As workaholics spend excessive time and energies working at the
expense of their personal and family life, they are expected to experience high levels of WFC
(Bakker et al., 2009; Schaufeli et al., 2009; Taris et al., 2005), which is considered one of the
most important predictors of burnout (Lee & Ashforth, 1996; Schaufeli et al., 2009).
Therefore, also a mediational role of WFC between workaholism and exhaustion is
hypothesized, consistent with a previous study (Schaufeli et al., 2009).

Hypothesis 2a: Workaholism partially mediates the relationship between job demands
on the one hand, and WFC and exhaustion on the other hand.
Hypothesis 2b: WFC partially mediates the relationship between workaholism and
exhaustion.

Finally, the interest of the study is to highlight the potential relationship of workaholism
also with intentions to change jobs. Such a relationship would signal a direct possible
consequence at the organizational level with implications for social and labour policies. Few
studies investigated the relationship between workaholism and intentions to leave the
organization, finding contradictory results (Burke, 2001; Kravina, Falco, Girardi, & De Carlo,
2010). Otherwise, in literature there is wide evidence about the effect of exhaustion on
intentions to leave (Bakker, Demerouti, & Schaufeli, 2003b; Hu, Schaufeli, & Taris, 2011;
Layne, Hohenshil, & Singh, 2004; Parasuraman, 1982): exhausted resources may lead to
burnout, which, in turn, can push workers to desire to change job. These intentions to change
may turn into absenteeism and turnover (van der Colff & Rothmann, 2009), thus representing
negative consequences and costs for organizations. In this study, therefore, we expect a
positive relationship between exhaustion and intentions to change jobs, and a mediational role
of exhaustion in the relationships between workaholism and WFC on the one hand, and
intentions to change jobs on the other hand.

29

Hypothesis 3: Exhaustion mediates the relationship between a) workaholism and b)

WFC, and intentions to change jobs.

Figure 1. The theoretical model.

WFC

+
+
+

Job
demands

+
-

Exhaustion

Intentions to
Change Jobs

+
Workaholism

Job
resources

METHOD

Participants and Procedure

The research involved a convenience sample of 617 Italian workers who filled out a selfreport on-line questionnaire. We asked people to send an email with the link to the
questionnaire to other colleagues or friends (snowball exercise sampling), therefore we are not
able to calculate the response rate.
The sample consisted of 343 females (55% of the sample) and 274 males (45% of the
sample). They were aged between 23 and 69 years old (M = 39.95; SD = 9.56). Among the
participants, 58% were married or cohabited; 47% had children. In the sample, 75% had a
bachelors, masters degrees or a higher educational qualification.
In total, 54% (331) were employees and 43% (266) were self-employed (missing cases =
3%). Participants were from different occupational sectors: most of them (42%) were from the
private service, 13% were from industry, 10% were from public health, 9% were from
education and research, 8% were from public service, 6% were from commerce, 6% were
from other sectors (missing cases = 6%). Weekly working hours were, on average, 42.05 (SD
= 10.84). Mean seniority on the job was 12.25 years (SD = 9.57).

30

Measures

Job demands. Workload was measured by using the 4-item scale taken from Bakker,
Demerouti and Verbeke (2004). An example item is Do you have too much work to do?
(from 1 = never to 5 = always); Cronbachs alpha for this study was .81. Cognitive demands
were evaluated with a 4-item scale of Bakker et al. (2003a). A typical item of this scale is
Does your work demand enhanced care or precision? (from 1 = never to 5 = always);
Cronbachs alpha was .78. Emotional demands were measured with 3 items (Bakker et al.,
2003b), including Do you face emotionally charged situations in your work? (from 1 =
never to 5 = always); Cronbachs alpha was .88. Customer related social stressors were
assessed with the 9-item scale developed by Dormann & Zapf (2004). Sample items are
Some customers think they are more important than others and Our customers demands
are often exorbitant (from 1 = strongly disagree to 6 = strongly agree); Cronbachs alpha
was .87.

Job resources. Opportunities for professional development were measured by the 4-item
scale of Bakker et al. (2003a), including My work offers me the opportunity to learn new
things (from 1 = totally disagree to 5 = totally agree); Cronbachs alpha was .88. Job
security was assessed by 3 items taken from the study of Kraimer, Wayne, Liden, and
Sparrowe (2005). A typical item is My job will be there as long as I want it (from 1 =
totally disagree to 7 = totally agree); Cronbachs alpha was .91.

Workaholism was measured with the 7-item Bergen Work Addiction Scale (BWAS),
which has high content validity in terms of the addiction nature of the construct (Andreassen
et al., 2012). Sample items are How often during last year have you thought of how you
could free up more time to work? and How often during last year have you become stressed
if you have been prohibited from working? (from 1 = never to 5 = always); Cronbachs alpha
was .78.

Work-family conflict was assessed with the Italian version (Colombo & Ghislieri, 2008)
of the 5-item scale taken by Netemeyer, Boles, and McMurrian (1996). Example items are
The demands of my work interfere with my home and family life and Things I want to do
at home do not get done because of the demands my job puts on me (from 1 = never to 6 =
always); Cronbachs alpha was .91.
31

Exhaustion was measured by 8 items of the Oldenburg Burnout Inventory (OLBI;

Demerouti, Mostert, & Bakker, 2010). Typical items are When I work, I usually feel
energized (reverse coded) and There are days when I feel tired before I arrive at work
(from 1 = strongly disagree to 4 = strongly agree); Cronbachs alpha was .76.

Intentions to change jobs were assessed with 2 items (Schaufeli & Bakker, 2004). An
example item of the two is I sometimes think about changing job (from 1 = totally disagree
to 5 = totally agree); Cronbachs alpha was .77.

Strategy of Analysis

The Mplus 7 software package (Muthn & Muthn, 1998-2012) was used to test the
study hypotheses through structural equation modeling (SEM) and moderated structural
equation modeling (MSEM). The method of estimation was maximum likelihood (ML).
According to the literature (Bollen & Long, 1993) several goodness-of-fit criteria were
considered: the 2 goodness-of-fit statistic; the Root Mean Square Error of Approximation
(RMSEA); the Comparative Fit Index (CFI); the Tucker Lewis Index (TLI); the Standardized
Root Mean Square Residual (SRMR); the Akaike's Information Criterion (AIC). Values of
both RMSEA and SRSM lower than .08, and CFI and TLI values greater than .90 indicate a
good fit.
The four job characteristics were modelled in a latent factor representing job demands,
which was treated as exogenous variable in the model. For reasons of parsimony, item
parceling for the endogenous variables was used, computing two parcels for each latent
construct in the model. Finally, bootstrapping was used to test the significance of the
mediation effects in more accurate confidence intervals than other methods (Shrout & Bolger,
2002).
To address the common method variance issue, a confirmatory factor analysis was
performed using the Harmans single-factor test (Podsakoff, MacKenzie, Lee, & Podsakoff,
2003). Results indicated that one single factor could not account for the variance in the data
[2 (860, N = 617) = 7550.05, p < .01, RMSEA = .11, CFI = .42, TLI = .40, SRMR = .12].
This indicates that common method variance was not a major problem in the present study. In
the following analyses, we controlled for gender, age, and mean seniority on the job.

32

To test the moderating effects (Hypothesis 1b), MSEM was preferred to hierarchical
regression analyses because the first one corrects for measurement errors and provides the fit
of the models under study. To apply MSEM, the procedure described by Mathieu,
Tanenbaum, and Salas (1992; in Cortina, Chen, & Dunlap, 2001) was used. For each
hypothesized interaction effect we tested a model that included three exogenous variables and
workaholism as endogenous variable (see Figure 2). Each exogenous variable had only one
indicator, that was the standardized score of the respective variable. The indicator of the
interaction factor was the multiplication of the indicators of the interacting variables. The path
from each latent exogenous variable to its indicator was fixed at the square root of the scale
reliability, whereas the error variance of each indicator was set equal to the product of its
variance and one minus its reliability. The reliability of the interaction term was calculated by
the formula reported in Cortina et al. (2011). The job demand variables and the job resource
variables were allowed to correlate, while the correlations between them and their interactions
were expected to be zero.

Figure 2. The study model to test the interaction hypothesis. All constrained paths and error variances are
marked with C. res. error = residual error.

33

RESULTS

Descriptive statistics

Table 1 includes the means, standard deviations, and correlations between the study
variables, as well as their internal consistencies.

Table 1. Item means, item standard deviations and correlations among the study variables.
M

SD

1. Workload

3.55

.80

(.81)

2. Cognitive demands

4.06

.70 .40**

3. Emotional demands

3.12

1.06 .10** .29**

4. Cust. rel. soc. stres.

3.81

1.04 .30** .16** .15**

5.OPD

3.57

.99 .01

.29** .18** -.07

6. Job security

4.26

1.74 .07

.15** .10*

7. Workaholism

2.23

.73 .43** .22** .21** .23** -.02

-.01

(.78)

8. WFC

3.36

1.24 .53** .30** .23** .28** .06

.05

.59**

9. Exhaustion

2.36

.55 .35** .14** .16** .24** -.36** -.19** .44** .45**

10. Int. change job

2.48

1.13 .12** -.18** -.13** .09*

10

(.78)
(.88)
(.87)

.06

(.88)
.34**

(.91)

(.91)

-.48** -.26** .14** .06

(.76)
.36**

(.77)

Note. Cronbachs Alpha on the diagonal. * p < .05; ** p < .01.

SEM analyses

Table 2 presents the results of alternative SEM models estimated to test the study
hypotheses. The hypothesized mediation model, in which workaholism was a partial mediator
between job demands and WFC and between job demands and exhaustion; WFC was a partial
mediator between workaholism and exhaustion; and exhaustion was a predictor of intentions
to change jobs, showed a good fit to the data (M1). Nevertheless, in this model, the path
coefficient from job demands to exhaustion was non-significant. Therefore, we calculated a
more parsimonious model without this path (M1-alternative), where workaholism was a fullnot
partialmediator between job demands and exhaustion. This model also showed a good fit to
the data and the chi-square difference test showed that M1-alternative fits as well to the data as M1
(2 = 2.36; n.s.).
Moreover, results showed that M1-alternative was significantly better compared with the
model in which all direct effects from job demands, workaholism, WFC and exhaustion to

34

intentions to change jobs were calculated (M2; no correlations between exogenous variables),
confirming the presence of mediating effects within the model (2 = 652.58; p < .001).
Finally, we tested also the saturated model (M3), where all direct and indirect effects
were modelled. Model 3 showed an acceptable fit to the data, but it was not significantly
better than M1-alternative (2 = 7.83; n.s.). Moreover, in M3 only the direct effect from
exhaustion to intentions to change jobs was significant; the other direct effects added in this
model (compared to M1-alternative) were non-significant (job demands exhaustion; job
demands intentions to change jobs; workaholism intentions to change jobs; WFC
intentions to change jobs). For this reason, we chose the more parsimonious M1-alternative,
which is graphically represented in Figure 3, as the best fitting model. In conclusion, the
results support Hypotheses 1b, 2a, 2b and 3.

Table 2. Results of SEM analysis.

2

df

CFI

TLI

RMSEA

SRMR

AIC

Comparison

228.48
46 < .01 .93 .90
.08
.06
20554.22
M 1.
230.84
47 < .01 .93 .90
.08
.06
20554.58
2.36
> .05
M1a.
M1aM1
883.42
49 < .01 .67 .56
.17
.22
21203.17
652.58 < .001
M 2.
M2M1a
223.01
43 < .01 .93 .89
.08
.05
20554.76
7.83
> .05
M 3.
M1aM3
Note. JD = job demands; WS = workaholism; WFC = work-family conflict; EX = exhaustion; IC = intentions to
change jobs.
M1. Hypothesized model: JD WS, WFC & EX; WS WFC & EX; WFC EX IC
M1a. Alternative, more parsimonious model: JD WS & WFC; WS WFC & EX; WFC EX IC
M2. Direct effects model: JD, WS, WFC, EX IC
M3. Saturated model: JD WS, WFC, EX & IC; WS WFC, EX & IC; WFC EX & IC; EX IC
Figure 3. The final SEM model (M1-alternative). Standardized solution; all paths are statistically significant at p <
.001. WFC = work-family conflict.

wfc1

wfc2
.90

.96

WFC
Workload

.43

.75

Cognitive
demands

.45

Emotional
demands

.28

.23

Job
demands

.19

Cust. rel.
soc. stress.

R2 = .59

.37

Exhaustion
R2 = .40

.41

.66
Workaholism
R2 = .43
.83

wsm1

.79

.45
exh1

.45

Intentions to
Change Jobs
R2 = .20
.92

.67

exh2

int1

.45

int2

.74

wsm2

35

Table 3. Indirect effects using bootstrapping (2000 replications).

Bootstrap
Indirect effects
Est.

S.E.

CI 95%

JD Workaholism Exhaustion

.29

.06

.000

(.16, .43)

JD WFC Exhaustion

.10

.04

.018

(.02, .18)

JD Workaholism WFC Exhaustion

.06

.02

.004

(.02, .10)

Workaholism Exhaustion Int. change

.20

.05

.000

(.10, .30)

WFC Exhaustion Int. change

.10

.04

.004

(.03, .18)

Workaholism WFC Exhaustion Int. change

.04

.02

.009

(.01, .08)

Note. All parameter estimates are presented as standardized coefcients. JD = job demands; WFC = work-family
conflict; Int. change = Intentions to change jobs; CI = confidence interval.

MSEM analysis

We used moderated structural equation modelling (MSEM) to test the hypothesis that job
resources mitigate the positive relationship between job demands and workaholism
(Hypothesis 1b). Table 4 shows the results. Five out of eight interaction effects of job
demands and job resources were statistically significant, thus our hypothesis was supported
for 63% of all cases. Both workload and cognitive demands interacted with both OPD and job
security to predict workaholism; moreover, emotional demands interacted with OPD, but not
with job security. Finally, the two job resources did not have an effect on the relationship
between customer related social stressors and workaholism.
In cases where the MSEM analyses resulted in a significant interaction effect, the chisquare difference test showed that the fit of the models with the path from the latent
interaction variable to the endogenous variables was significantly better than the models
without this path (see Table 4), thus further supporting the interaction effects outlined in our
hypothesis (Cortina et al., 2001).

36

Table 4. Results of MSEM analysis: interactions of job demands and job resources (N = 617).
Workaholism
Predictor

UPC (SE)

Workload

.37 (.03)

.53***

OPD

-.05 (.03)

-.07

Workload x OPD

-.07 (.03)

-.12**

R2

SPC

30%

Workload

.38 (.03)

.54***

Job security

-.06 (.04)

-.08

-.07 (.03)

-.13**

Workload x Job security

R

31%

Cognitive demands

.21 (.04)

.30***

OPD

-.13 (.04)

-.18***

-.08 (.03)

-.14**

Cogn. dem. x OPD

R

10%

Cognitive demands

.18 (.03)

.28***

Job security

-.05 (.04)

-.07

Cogn. dem. x Job sec.

-.07 (.03)

-.14**

R2

10%

Emotional demands

.18 (.03)

.27***

OPD

-.09 (.04)

-.13*

-.09 (.03)

-.17**

Emot. dem. x OPD

R

30%

Emotional demands

.16 (.03)

.26***

Job security

-.02 (.04)

-.03

-.05 (.03)

-.09

Emot. dem. x Job sec.

R

8%

Cust. rel. soc. stress.

.18 (.03)

.27***

OPD

-.02 (.04)

-.03

-.03 (.03)

-.05

Cust. soc. str. x OPD

R

9%

Cust. rel. soc. stress.

.18 (.03)

.28***

Job security

-.03 (.04)

-.04

.00 (.03)

.01

Cust. soc. str. x Job sec.

R

30%

Fit

CFI

TLI

RMSEA

SRMR

11.26

.99

.97

.05

.02

7.12

< .01

5.59

.99

.99

.03

.02

7.77

< .01

15.47

.98

.94

.07

.03

9.36

< .01

5.45

.99

.99

.02

.02

7.59

< .01

20.89

.96

.91

.08

.04

9.94

< .01

4.52

1.00

1.00

.02

.01

3.04

> .05

12.04

.98

.96

.06

.02

.09

> .05

6.47

.99

.99

.03

.02

.01

> .05

Note. The df of all models is 4. * p < .05; ** p < .01; *** p < .001. UPC = unstandardized path coefficient; SPC
= standardized path coefficient; OPD = opportunities for professional development; Cust. rel. soc. stress. =
customer related social stressors. 2 = comparison between models without the path from the latent interaction
variable to the endogenous variable and models with this path.

37

Dawson and Richters (2006) software was used to plot the five significant moderation
effects. Figure 4 confirms the direction of the moderation effects. In all five situations, job
resources mitigated the positive relationship between job demands and workaholism; in other
words, the positive relationship between job demands and workaholism is particularly high
under conditions of low (vs. high) job resources. These findings offer support for Hypothesis
1b.
Figure 4. A: The effect of opportunities for professional development on the relationship between workload and
workaholism. B: The effect of job security on the relationship between workload and workaholism. C: The effect
of opportunities for professional development on the relationship between cognitive demands and workaholism.
D: The effect of job security on the relationship between cognitive demands and workaholism. E: The effect of
opportunities for professional development on the relationship between emotional demands and workaholism.

38

DISCUSSION AND CONCLUSION

The central aim of this study was to understand the role of workaholism in the JD-R
model (Bakker & Demerouti, 2007). In contrast with other scholars that considered
workaholism as a stable individual characteristic (Balducci et al., 2012; Burke, Matthiesen, &
Pallesen, 2006; Scott et al., 1997), we assumed that workaholism may vary as a function of
the working context (Fry & Cohen, 2009; van Wijhe et al., 2010). This represents the main
originality of our study, since previous research has focused mostly on the consequences of
workaholism and not on its potential job antecedents.
Hypothesis 1a assumed that job demands are positively related to workaholism and the
final model resulting from our analyses supports the existence of this positive relationship.
Although results of this cross-sectional study cannot confirm causal relationships between
variables, they can provide some evidence about the emerging idea that the work environment
is linked with work addiction. This represents an original and important finding, as it
recognizes the potential role of organizations, employers and managers in fostering
workaholic behaviours (Fry & Cohen, 2009; van Wijhe, Schaufeli, & Peeters, 2010) and it
provides preliminary evidence for practical implications aimed at both preventing and
limiting workaholism in workplaces.
Based on the JD-R model, Hypothesis 1b took into consideration the role of some job
resources, applying their buffer effect to the positive relationship between job demands and
workaholism. Some authors, indeed, demonstrated that job resources could reduce the
negative impact of job demands on well-being (Bakker et al., 2005). Nevertheless, this is the
first study which considers the possibility to extend the buffer hypothesis of the JD-R model
also to workaholism, defined as a kind of employee well-being (Schaufeli, Taris, & van
Rhenen, 2008). Although the results did not confirm the hypothesis in all the combinations
between job demands and job resources included in this study, the percentage of interactions
found, equal to 63%, may be considered a substantial finding; moreover, all significant effects
were in the expected direction. It follows that, under highly stressful working conditions, the
risk of workaholism is lower if sufficient job resources are available (like in the case of other
outcomes considered in the JD-R model; Xanthopoulou et al., 2007). Specifically,
opportunities for professional development probably plays a moderating effect as they offer
workers knowledge and skills necessary to achieve job results and cope with stressful
working situations (Molino, Ghislieri, & Cortese, 2013; Salanova, Agut, & Peiro, 2005) in an
effective and not addictive way. Further, job security can buffer the relationship between job
39

demands and work addiction as it reduces uncertainty and peoples fear to lose their job
(Kraimer et al., 2005; Matuska, 2010). Future studies should further explore this relationship
in order to better understand if in presence of less uncertainty about their job workers could
feel more able to cope with demands, without working in an excessive way.
These results suggest the possibility to apply the JD-R model to workaholism,
considering it as a dependent variable. Further, they provide preliminary evidence of the
buffering role of opportunities for professional development and job security in the
relationship between general job demands and workaholism. These findings are important
since they lead to focus the attention, at the organizational and also socio-political level, on
the role that some resources can play in buffering the negative effect of job demands on
workaholism.
Furthermore, the study intended also to investigate, in an overall model, consequences of
workaholism on three important life domains: individual health, in terms of workers
exhaustion; family, in terms of difficulties in participating in both work and family roles; and
work, in terms of peoples intentions to change jobs with negative consequences for
employers and organizations.
We know from the literature that both job demands and workaholism can increase both
WFC (Bakker et al., 2009; Bonebright et al., 2000; Taris et al., 2005) and exhaustion (Bakker
et al., 2012; Sonnentag & Zijlstra, 2006; van der Hulst & Geurts, 2001), as they generally
prompt an over-commitment of time and energies to work, at the expense of personal and
family life. Hypothesis 2a assumed that workaholism plays a mediational role within the
framework of these relationships. On the one hand, results confirmed that the positive
relationship between job demands and WFC is partially mediated by work addiction. This
means that individuals functioning in family domain is influenced by demands coming from
work not only di per se, but also via the reinforcement of those addictive behaviours which
worsen the balance between work and family, in terms of both quantitative and qualitative
time dedicated to them (Brady, Vodanovich, & Rotunda, 2008). On the other hand, results
found a totalnot partialmediation of workaholism between job demands and exhaustion,
highlighting the role of workaholism in the well-studied health impairment process proposed
by the JD-R model (Bakker & Demerouti, 2007).
Finally, also Hypothesis 2b, which assumed a mediation of WFC between workaholism
and exhaustion, found confirmation. In our study this mediation was partial, consistent with
literature indicating a direct effect of workaholism on health impairment and burnout
(Andreassen et al., 2007; Burke & Matthiesen, 2004; Greenberg, 1987; Ng et al., 2007;
40

Porter, 1996; Taris et al., 2008). On the basis of these results, to handle WFC and health
problem issues, and their negative consequences for workers and organizations, scholars and
practitioners should deal excessive requests from work, but also consider the role of
workaholism and the risk of encouraging an excessive dedication to work.
The final hypothesis of the study took into account the relationship of workaholism with
intentions to change jobs, precursor of negative consequences for organizations (Kim & Lee,
2007) especially in terms of absenteeism and turnover (van der Colff & Rothmann, 2009).
Specifically, it hypothesized that the positive relationship of workaholism and WFC with
intentions to change jobs is mediated by exhaustion, and results confirmed this hypothesis.
Leading to an exhaustion of energies, and thus to symptoms of burnout which decrease the
quality of health and life, both workaholism and WFC were indirectly related to intentions to
change job, probably as a strategy to avoid working stress situations. On the basis of these
preliminary findings, future longitudinal studies should investigate the negative consequences
of workaholism for organizations over longer periods of time.

Limitations

Despite the original findings, the present study has certain limitations. The first one is its
cross-sectional nature, which exclude the possibility to draw any conclusions in terms of
causal effects in the relationships tested. Future longitudinal research or diary studies are
needed to replicate these findings and verify their causality, especially to test the mediational
role of workaholism in the JD-R model and its relationship with job demands and job
resources.
A second limitation is that all the data were self-reported, which means that relationships
among the variables could be inflated (Conway, 2002). Even though this study provided some
evidence that common method variance did not represent a critical factor (Podsakoff et al.,
2003), it would be interesting for future studies to consider also other- and objective ratings.
A third limitation of the study is that the sample was heterogeneous and possibly not
representative of the working population at large. But then, for the same reason, this study
emphasize the flexibility of the JD-R model and the possibility to apply it to different
occupational contexts and situations. Nevertheless, the findings should be replicated in
specific organizations and working places, also to identify more contextualized practical
implications.

41

Regarding the interaction effects tested in this study, we found evidence in 63% of the
cases, thus we can draw only cautious conclusions regarding the buffer hypothesis.
Nevertheless, these findings are important from a theoretical perspective since they can be
considered as a starting point to expand the JD-R model processes also to workaholism.
Future research should try to replicate these findings considering different kind of job
demands and, above all, job resources not only at the organizational level, but also at the
interpersonal, organization of work and task levels. Moreover, the study did not take into
account dispositional variables, although they can play an important role in generating
addictions (Eysenck, 1997). Future studies should integrate these findings considering also
the effect of personal resources, such as self-esteem (Ng et al., 2007) or self-efficacy (Kravina
et al., 2010; Xanthopoulou et al., 2007).
A final limitation of the study is that it has considered only negative outcomes; they
should be integrated with also positive variables, such as work engagement and job
satisfaction, respecting the call for positive psychology (Bakker & Schaufeli, 2008; Schaufeli
et al., 2009). Moreover, relationship between workaholism and recovery need to be deepened
in future, in line with recent findings (Bakker et al., 2012).

Practical implications

The first step to be taken is reaching awareness of the existence of a problem with so
serious implications. At the organizational level, it is the culture that is sometimes to be
changed, to be more work-family supportive (van Wijhe et al., 2010). Employers and
managers play a crucial role, as they can set a good example to work in a healthy way (Fry &
Cohen, 2009): programs for leadership development in the work context and attention to
employer-recruitment selection represent important interventions. Moreover, employees
should be exposed to challenging, but not exaggerated, job demands (Bakker et al., 2009) and
the work environment should guarantee adequate resources. Above all, organizations should
offer sufficient opportunities for professional development (e.g., training, learning
experiences, career advancement) to support employees in facing job demands. Furthermore,
in the current Italian context, the possibility to provide greater protection in terms of job
security seems to be highly relevant (Molino et al., 2013), also to decrease the intentions to
change jobs, and National labor policies should take this emergent issue into account.
Considering the individual level of intervention, firstly workers need education to
become aware of the existence of workaholism, its causes and, above all, its potential
42

consequences for their well-being and quality of personal and family life. The accessibility to
formal psychological counseling (Ishiyama & Kitayama, 1994) and to training programs
focused on time management, stress management and personal effectiveness (Schabracq,
2005) can be important instruments to prevent or dealing with workaholism. Finally,
employees should be aware of how important detaching from work and recovering are for
their well-being, thus rediscovering their personal hobbies and interests (Oates, 1971).

43

CHAPTER 3

Workaholism and recovery experiences: A diary-study

Abstract

Workaholics generally allocate an excessive amount of time and energies to their work at the
expense of having time for recovery from work. Nevertheless, a complete recovery is an
essential prerequisite for workers well-being. This study examines the role of three recovery
experiences, psychological detachment, relaxation and control, in the relationship between
workaholism and exhaustion. Data were collected among 96 participants, employees and selfemployed workers, who completed a general questionnaire and a diary booklet for five
consecutive working days. Multilevel analysis results showed that among the three recovery
experiences, only psychological detachment plays a mediational role in the relationship
between workaholism and exhaustion. These results represent important preliminary findings,
since few studies have investigated the relationship between workaholism and recovery so far.
In general, this insight leads to design practical interventions that stimulate psychological
detachment from work, particularly for workers with workaholic tendencies.

Keywords: workaholism; recovery experiences; exhaustion; diary study.

44

INTRODUCTION

Due to the increasing of work intensity in recent years, many employees have to face a
high work pace, fixed deadlines, and insufficient time to do their tasks (van Wijhe, Peeters,
Schaufeli, & Ouweneel, 2012) with the consequent need to work hard, also in the evenings
and weekends. However, some of them work in an excessive way not only for external
requirements or economical necessities, but because they feel they have to (van Beek, Hu,
Schaufeli, Taris, & Schreurs, 2012). For several scholars, indeed, it is not the number of hours
people work, but the attitudes and feelings towards work that distinguishes workaholics from
non-workaholics (Machlowitz, 1980).
Working excessively does not represent a problem as long as workers can recover from
the work-related efforts on a daily basis (Sonnentag, 2003), since sufficient recovery is an
essential prerequisite for well-being (Sonnentag, Binnewies, & Mojza, 2010). Nevertheless,
workaholics are so involved in their work that they find very difficult to detach from it; thus,
they seem to neglect their need for recovery (Taris, Schaufeli, & Verhoeven, 2005).
Workaholics spend much time and energy at work, leaving fewer resources to devote to their
family and to their activities of non-work life (Bakker, Demerouti, & Burke, 2009).
Consequently, they have relatively few opportunities to recover from their workdays and they
generally suffer from work-family conflict and exhaustion (Taris et al., 2005).
The aim of the present study is to examine the relationship between workaholism,
recovery experiences and exhaustion, using a within-person daily diary approach that covered
five consecutive working days. The vast majority of studies on workaholism have
investigated between-person differences throughout cross-sectional research, answering
questions about its relationship with work-family conflict, burnout or performance
(Andreassen, Ursin, & Eriksen, 2007; Taris, Schaufeli, & van Rhenen, 2008; Taris et al.,
2005). Few studies on workaholism have used the daily approach and have focused on its
relationship with recovery experiences so far (Bakker, Demerouti, Oerlemans, & Sonnentag,
2012; van Wijhe et al., 2012). Whereas the cross-sectional method considers the person as
unity of analysis, thus the analyses are based on interindividual variation, diary designs permit
to analyze short-term intraindividual fluctuations on a daily basis. For this reason, such a
method has been indicated as the most appropriate in work and organizational research,
particularly in the areas of health and stress, work-home interface, and recovery (Ohly,
Sonnentag, Niessen, & Zapf, 2010).

45

Although there is some evidence showing recovery experiences as moderators in the

stressor-outcome relationships (Moreno-Jimenez et al., 2009; Siltaloppi, Kinnunen, & Feldt,
2009), their mediational role has received less attention thus far (Kinnunen, Feldt, Siltaloppi,
& Sonnentag, 2011; Sonnentag, Kuttler, & Fritz, 2010). Therefore, the goal of this diary study
is to investigate whether recovery experiences may play a mediational role between
workaholism and exhaustion, attenuating the psychological costs for those persons with high
levels of work addiction.

Recovery experiences

Recovery refers to the process during which an individuals functioning that have been
called upon during a stressful experience returns to its prestressor levels, reducing strain
(Craig & Cooper, 1992; Meijman & Mulder, 1998). When recovery is not sufficient,
individuals have to put in extra effort at work to maintain a satisfactory performance level,
which may inflict strain and in the long period lead to health problems (Kinnunen et al., 2011,
Meijman & Mulder, 1998).
This study focuses on the recovery experiences introduced by Sonnentag and Fritz
(2007), namely psychological detachment, relaxation, mastery and control. They represent
personal strategies that have the potential to promote recovery. The authors referred to the
Effort-Recovery model (Meijman & Mulder, 1998) and to the Conservation of Resources
theory (Hobfoll, 1998) to develop an understanding of successful recovery experiences.
According to the Effort-Recovery model (Meijman & Mulder, 1998), effort expenditure at
work leads to load reactions such as fatigue or physiological activation. Under optimal
circumstances, which means that individual is no longer exposed to the work demands, load
reactions are released and recovery occurs. Consistent with this model, a precondition for
recovery is that the functional systems strained during work will not be called upon any
longer. However, if no adequate recovery takes place, stress-related load reactions do not
return to prestressor levels (Geurts & Sonnentag, 2006). Moreover, continuous exposure to
high work demands and incomplete recovery could induce an accumulation of load reactions
leading to chronic health impairment (Geurts & Sonnentag, 2006; van Wijhe et al., 2012).
The Conservation of Resources Theory (Hobfoll, 1998) assumes that stress occurs when
an individuals resources are threatened or lost. Therefore, to recover from stress, individuals
have to restore their resources and gain new resources. Stress recovery on a day-to-day basis
particularly refers to internal resources such as energy or positive mood. Thus, the Effort46

Recovery model and the Conservation of Resources theory suggest two complementary
processes by which recovery occurs. Firstly, refraining from work demands and avoiding
activities in the extra-work time that call upon the same functional systems, or internal
resources as those required at work, are important strategies. Secondly, gaining new internal
resources such as energy or positive mood will additionally help to restore endangered
resources (Sonnentag & Fritz, 2007). On the basis of this theorization of recovery processes,
psychological detachment and relaxation have their roots in the Effort-Recovery model
(Meijman & Mulder, 1998), whereas mastery and control in the Conservation of Resources
theory (Hobfoll, 1998).
Psychological detachment implies disengaging mentally, besides physically, from work
during off-job time. The term detachment has been used to describe an individuals sense of
being away from the work situation (Etzion, Eden, & Lapidot, 1998). It means that an
individual needs to stop thinking about work and not to be occupied by work-related affairs
and problems when he is not working. Psychological detachment from work goes beyond the
pure physical absence from the workplace during off-job time; rather, it implies leaving the
work behind oneself in psychological terms and switch off (Sonnentag & Bayer, 2005). Based
on the Effort-Recovery model, when no further demands are made on the functional systems,
recovery can takes place (Meijman & Mulder, 1998), otherwise the functional systems would
be continuously challenged and no full recovery could occur. Relaxation is a process
generally associated with leisure activities. It refers to feeling calm and peaceful, and it is
characterized by a state of low activation and increased positive affect (Sonnentag & Fritz,
2007). The state of relaxation may be either a result of deliberately chosen activities aiming at
the relaxation of body and mind, or it may also be achieved when performing other activities
such as reading a book, taking a walk or listening to music.
Mastery refers to pursuing off-job activities that distract from the job by providing
challenging experiences and opportunities to learn new skills in domains different from own
job. Although mastery experiences may impose additional demands, they challenge the
individual without overtaxing his or her capabilities since they help to build up new internal
resources such as skills, competencies, self-efficacy and positive mood (Bandura, 1997;
Hobfoll, 1998; Sonnentag & Fritz, 2007). Control applied to leisure time refers to the degree
to which an individual can decide which activity to pursue, when and how to pursue this
chosen activity in the leisure time. According to Sonnentag & Fritz, the experience of control
seems to be associated with positive reactions (Burger, 1989), and may increase self-efficacy
and feelings of competence. In addition, control during leisure time gives the individual the
47

opportunity to choose those specific activities that may be especially supportive for him or her
for the recovery process.

The workaholics tendency to devote more resources to their work than to their non-work
life leads them to neglect their activities outside the job. Addictive workers are unable to
reduce or control hard work, they continue to work despite social or health problems, and they
experience unpleasant withdrawal symptoms when away from work (Bakker et al., 2012;
Scott, Moore, & Miceli, 1997). These behaviours reduce workaholics resources when they are
not-working, in terms of both time and energy that could be invested in non-work activities
and psychological and mental resources.
The prolonged exposure to work demands during after working hours does not allow
workaholics to complete recovery before the next working day starts, with high risks for their
health (Bakker et al., 2012; van Wijhe et al., 2012). Thus, work-related activities and thoughts
carried out after work may impair daily recovery since workaholics allocate an excessive
amount of time and energies to their work at the expense of having time for recovery
experiences. On the basis of previous studies which found different results for mastery
compared with the other recovery experiences (e.g. Molino, Ghislieri, Cortese, & Bakker,
2013), in this study only three recovery experiences have been considered. Therefore we
hypothesize that workaholism is negatively related to psychological detachment, relaxation
and control at the within-level.

Hypothesis 1: day-level workaholism is negatively associated with a) day-level

psychological detachment, b) day-level relaxation and c) day-level control.

Empirical evidence suggests that recovery is helpful in restoring from job strain and is
negatively related to health complaints, depressive symptoms, exhaustion and sleep problems
(Fritz & Sonnentag, 2006; Sonnentag & Bayer, 2005; Sonnentag & Fritz, 2007; Sonnentag,
Mojza, Binnewies, & Scholl, 2008; Sonnentag, Kuttler, & Fritz, 2010). Therefore this study,
focusing on particularly the three recovery experiences, hypothesizes a negative relationship
of them with exhaustion.

Hypothesis 2: a) day-level psychological detachment, b) day-level relaxation and c)

day-level control are negatively related to day-level exhaustion.

48

Moreover, several studies showed that workaholism is related with poor health and wellbeing, such as subjective health complaints and exhaustion (Andreassen et al., 2007; Burke,
2008; Schaufeli, Taris, & Bakker, 2008; van der Hulst & Geurts, 2001). Indeed, some authors
indicated workaholics as a risk group for ill-health (Kanai, 2006); therefore, a positive
relationship of workaholism with exhaustion is expected.

Hypothesis 3: day-level workaholism is positively associated with day-level

exhaustion.

Figure 1. Hypothesized model

General level
Detachment

Day-level
Detachment

General level
Relaxation

Day-level
Relaxation

General level
Control

Day-level
Control

General level
Workaholism

Day-level
Workaholism

General level
Exhaustion

Day-level
Exhaustion
Between
Within

Day-level
Detachment

Day-level
Workaholism

Day-level
Exhaustion
Day-level
Relaxation

Day-level
Control

Finally, on the basis of the previous hypotheses, this study expects a mediational role of
recovery experiences between workaholism and exhaustion. It has been argued that lack of
recovery plays a crucial role in explaining why job stressors may produce poor well-being and
health problems (Demerouti, Bakker, Geurts, & Taris, 2009; Geurts & Sonnentag, 2006). This
49

study hypothesizes that particularly the three recovery experiences can mediate the positive
relationship between workaholism and exhaustion.

Hypothesis 4: the relationship between day-level workaholism and day-level

exhaustion is partially mediated by a) day-level psychological detachment, b) daylevel relaxation and c) day-level control.

METHOD

Participants and procedures

Participants were contacted throughout snowball exercise and they received a diary
booklet. They were instructed to fill in a general questionnaire, in which they provided some
demographic data and information on the general level of the measured variables, before
starting with the diaries. Then they were asked to fill in the diary for five consecutive working
days at the end of each day before go to bad.
A total of 96 usable diaries were returned. The total sample included 59 females (62%)
and 37 males (38%). Their mean age was 36.35 (SD = 10.42); 72% were married or cohabited
and 64% did not have children. In the sample, 47% had a bachelors, masters degrees or a
higher educational qualification, 40% finished the high school and the others had a lower
level of education.
Among participant, 79% were employees and 21% were self-employed workers; 25%
were working in the industrial sector, 19% in the commerce, 12% in the education and
research sector, 10% in private service, 9% in public services and 25% indicated another
sector. The mean working hours per week were 39.73 (SD = 8.81). Mean seniority on the job
was 10.45 years (SD = 10.13).

Measures

General questionnaire

General Workaholism was measured with the 7-item Bergen Work Addiction Scale
(BWAS; Andreassen, Griffiths, Hetland, & Pallesen, 2012). Sample item is How often
50

during last year have you thought of how you could free up more time to work? (from 1 =
never to 5 = always); Cronbachs alpha for this study was .81.

General Recovery Experiences were detected by the Recovery Experience Questionnaire

(Sonnentag & Fritz, 2007). We used a 9-item scale, 3 items for each dimension. Sample item
is: In the evening after work, I dont think about work at all for detachment; In the evening
after work, I do relaxing things for relaxation; In the evening after work, I determine for
myself how I will spend my time for control (from 1 = I do not agree at all to 5 = I fully
agree). Cronbachs Alpha was .93 for psychological detachment; .87 for relaxation; .72 for
control.

General Exhaustion was measured by the 8 items taken by the Oldenburg Burnout
Inventory (OLBI; Demerouti, Mostert, & Bakker, 2010). Typical item is There are days
when I feel tired before I arrive at work (from 1 = strongly disagree to 4 = strongly agree);
Cronbachs alpha for this study was .73 for exhaustion.

Diary survey data

The diary booklet consisted of five identical questionnaires, one for each day, from
Monday to Friday. Participants responded to all day-level measures on a seven-point scale
(from 1 = strongly disagree to 7 = strongly agree) and they were asked to thing about the past
day and not about their overall situation (all items of the diaries started with the world
Today, ...).

Day-level Workaholism was assessed with seven adapted items of the BWAS
(Andreassen et al., 2012). Sample items is Today I worked so much that it has negatively
influenced my health. Cronbachs alpha for this study was calculated separately for each day
and ranged from .79 to .87 (M = 83).
Day-level Recovery experiences were measured by 12 items of the Recovery Experience
Questionnaire (Sonnentag & Fritz, 2007), 3 items for each dimension. Example items are:
This evening, after work, I forgot about work for psychological detachment; This evening,
after work, I used the time to relax for relaxation; This evening, after work, I decided my
own schedule for control. Cronbachs alpha ranged from .94 to .96 (M = 95) for

51

psychological detachment; from .88 to .92 (M = 90) for relaxation; from .78 to .89 (M = 81)
for control.

Day-level Exhaustion was assessed using the 8 items taken by the OLBI (Demerouti et
al., 2010). Sample item is Today, I felt emotionally drained by my work. Cronbachs alpha
ranged from .67 to .87 (M = 75).

Statistical analysis

This study collected repeated measures data which can be considered as multilevel data,
with repeated measurements nested within persons (Hox, 2002). This leads to a two-level
model with 480 occasions at the lower level and 96 participants at the higher level. Multilevel
structural equation modelling (MSEM) was applied by using Mplus 7 (Muthn & Muthn,
1998-2012). The method of estimation was maximum likelihood (ML). According to the
literature (Bollen & Long, 1993) several goodness-of-fit criteria were considered: the 2
goodness-of-fit statistic; the Root Mean Square Error of Approximation (RMSEA); the
Comparative Fit Index (CFI); the Tucker Lewis Index (TLI); the Standardized Root Mean
Square Residual (SRMR).
At the between level of the MSEM, control variables included general level of
workaholism, general level of the three recovery experiences and general level of exhaustion.
Controlling for the effect of general-level variables to the respective day-level variables
permits to examine the relationships between day-level fluctuations after taking into account
individual baselines. When investigating the effects of predictors on specific outcomes
additional factors, in particular demographic factors, could impact on these variables.
However, none of the demographic characteristics were significant predictors of dependent
variables in this study, they were therefore excluded from further analysis.
We modelled paths between day-level variables at the within part of the model (Figure 1)
because this level refers to fluctuations over time (variations within persons). We modelled
paths from between-level variables to day-level variables at the between part of the model
because this level refers to individual baselines (variation between persons). We used grandmean centering for general-level explanatory variables (general level of workaholism,
psychological detachment, relaxation, control and exhaustion) to the sample mean (Ohly et
al., 2010). We did not center variables that were treated as explanatory variables and
outcomes. Therefore, the results can be interpreted as follows: when a day-level variable x is
52

related to a day-level variable y, it means that on days when respondents report the x variable
at higher levels than they did on average (within-level variation in variable x), they report
high levels of variables y. When a general-level variable x is related to a day-level variable y,
it means that when respondents report levels of the variable x higher than the mean score of
the sample (between-level variation in variable x), they report higher levels of variable y
(Ohly et al., 2010; Petrou, Demerouti, Peeters, Schaufeli, & Hetland, 2012).

RESULTS

Descriptive statistics

The means, standard deviations and correlations for all study variables are presented in
Table 1. All significant relationships between variables were in the expected direction. In
particular, significant correlation were found between day-level workaholism, day-level
psychological detachment, day-level relaxation, day-level control and day-level exhaustion,
except for the correlation between the last two day-level variables.
Table 1. Item means, item standard deviations and correlations among the study variables.
M
1. General workaholism 2.16

SD

.80

2. General psych.
detachment

3.04

1.15 -.25**

3. General relaxation

3.57

1.01 -.24** .30**

4. General control

3.50

.86 -.01

5.General exhaustion

2.36

.48 .40** -.16** -.21** -.16**

6. Day-level
workaholism

2.43

1.15 .57** -.24** -.11* .01

7. Day-level psych.
detachment

4.93

1.80 -.25** .49** .21** .20** -.14** -.33**

8. Day-level relaxation

4.63

1.67 -.14** .31** .37** .28** -.16** -.15** .44**

9. Day level control

4.48

1.54 -.15** .25** .23** .50** -.21** -.10* .38** .53**

10. Day-level exhaustion 2.72

10

.26** .41**

1.11 .39** -.17** -.12** .02

.20**

.38** .53** -.13** -.07

-.13**

Note. Day-level data was average across the 5 days;. * p < .05; ** p < .01.

53

Fluctuations over time

Before testing the hypotheses, the intraclass correlation (ICC) has been used to determine
whether the use of multilevel analysis was justified. The ICC reflects the amount of betweenperson variability compared to the amount of total variability. Results showed that 42% of the
variance in day-level workaholism, 75% of the variance in day-level psychological
detachment, 67% of the variance in day-level relaxation, 59% of the variance in day-level
control and 30% of the variance in day-level exhaustion was attributable to between-persons
variations. Consequently, significant amounts of variance are left to be explained by withinperson variations across the 5 days, justifying the multilevel approach (James, 1982).

MSEM analyses
The hypothesized MSEM model showed satisfactory fit to the data: 2 (20) = 38.89, p <
.01, RMSEA = .05, CFI = .96, TLI = .90, SRMR = .00 (within-level) and .07 (between-level).
The model is graphically represented in Figure 2.
At the between-level of the model all the paths from between-level variables to the
respective day-level variables were significant. Therefore, the examination of the
relationships between day-level fluctuations was possible taking into account individual
baselines.
At the within-level of the model a significant negative relationship between day-level
workaholism and day-level psychological detachment was found. Day-level workaholism did
not show a significant relationship with both day-level relaxation and day-level control. Thus,
results confirm Hypothesis 1a but they do not confirm Hypotheses 1b and 1c.
Moreover, among the three day-level recovery experiences only psychological
detachment was significantly and negatively associated with day-level exhaustion, supporting
Hypothesis 2a and rejecting Hypotheses 2b and 2c. Further, day-level workaholism showed a
significant positive relationship with day-level exhaustion, supporting Hypothesis 3. Finally,
significant positive correlations between the three recovery experiences have been found at
both the within- and between-level. This correlation was stronger between relaxation and
control, compared with their respective correlations with psychological detachment.
Therefore, day-level psychological detachment was the only one recovery experience
among the three showing a partial mediational role between day-level workaholism and daylevel exhaustion. The application of the Sobel test was used to examine the significance of
54

this mediating effect; results indicated that the partial mediation of day-level psychological
detachment between day-level workaholism and day-level exhaustion was significant (z =
1.68; p < .05), confirming Hypothesis 4a.

Figure 2. Results for multilevel structural equation modelling

.30
.26*
-.25*

General level
Detachment

.64

Day-level
Detachment

General level
Relaxation

.52

Day-level
Relaxation

.30*
.37
.53

.41

General level
Control

.58

Day-level
Control

General level
Workaholism

.55

Day-level
Workaholism

-.36

-.24*
-.21*

.61
.40

General level
Exhaustion

Day-level
Exhaustion

.48

Between
Within

Day-level
Detachment
-.17

-.12
.38

Day-level
Workaholism

.30

Day-level
Exhaustion
.29

Day-level
Relaxation
.46

Day-level
Control

Note. Standardized solution; solid arrows indicate significant relationships (p < .001; * p < .05), and broken
arrows refer to non-significant paths.

DISCUSSION AND CONCLUSION

The present diary study examined a model concerning the daily influences of
workaholism on recovery experiences and exhaustion. Although significant results have been
found only for psychological detachment, the study can be considered innovative as it

55

combined workaholism and recovery literatures, on a daily basis, detecting preliminary

findings to support future research and to identify practical implications.
It was hypothesized that workaholics spend additional hours working during the evening
at the expense of their recovery (Bakker et al., 2012; van Wijhe et al., 2012). The workaholics
tendency to devote more resources to their work than to their non-work life leads them to
neglect their activities outside the job and, consequently, their opportunities to recover from
work. Our results found confirmation for Hypothesis 1a, since they showed a significant
negative relation of daily workaholism with daily psychological detachment. Working
excessively, also in the off-job time, seem to interfere with the possibility to psychologically
detach from work. Workaholics, indeed, tend to work long hours not only for external
requirements or needs, but because they often feel they have not done enough yet (van Beek
et al., 2012; van Wijhe et al., 2012); thus, they are so involved in their job-related problems
and affairs that it seems particularly difficult for them stop to think about work. Hypothesis 1b
and 1c did not find confirmation from this study: the effects of workaholism on the
experiences of relaxation and control on leisure time were not significant. Therefore we can
conclude that workaholic behaviours interfere with recovery, especially in terms of switching
off and disengaging mentally from work during off-job time (Sonnentag & Fritz, 2007).
According to the Effort-Recovery model, adequate recovery from work is negatively
related to long-term health impairment (Meijman & Mulder, 1998; Sonnentag & Fritz, 2007)
and some empirical evidence present in literature supports this model (Fritz & Sonnentag,
2006; Sonnentag & Bayer, 2005; Sonnentag et al., 2008; Sonnentag et al., 2010). Therefore,
Hypothesis 2 suggested a negative relationship of the three recovery experiences with
exhaustion. In our model only Hypothesis 2a found confirmation: results showed a significant
negative effect of daily psychological detachment on daily exhaustion, but this relationship
was not significant in the cases of daily relaxation and daily control. Some authors indicated
psychological detachment as the core experience of the recovery process (Sonnentag & Fritz,
2007; Volman, Bakker, & Xanthopoulou, 2012) and our results confirm that it plays a crucial
role in this process, compared with other recovery experiences (Sonnentag & Bayer, 2005).
Future research, longitudinal or diary studies, should focus on the differences between the
recovery experiences considered so far in relationship with exhaustion.
On the basis of findings stated in several studies (Andreassen et al., 2007; Burke, 2008;
Schaufeli et al., 2008; van der Hulst & Geurts, 2001), Hypothesis 3 supposed a positive
relationship of daily workaholism with daily exhaustion. Results confirmed this hypothesis
and added to the literature since they demonstrated this effect on a daily basis, unlike previous
56

studies which examined this relationship only throughout cross-sectional design (Andreassen
et al., 2007; Schaufeli et al., 2009).
Finally, Hypothesis 4 suggested a mediational role of recovery experiences between
workaholism and exhaustion. This hypothesis found confirmation for only psychological
detachment, supporting the meaning of psychological detachment as a core experience in the
recovery process (Sonnentag & Fritz, 2007; Volman et al., 2011). On the basis of these
findings, we can conclude that the positive relationship between daily workaholism and daily
exhaustion is partially mediated by daily psychological detachment: workaholics seem to
have more difficulties, compared with non-workaholics, to switch off from work during offjob time, and this can be considered one of the causes of their increased health problems.
These results are consistent with a previous study of Sonnentag et al. (2008), which found that
psychological detachment from work is particularly important for employees highly engaged
in their work.
All in all, this study confirmed that workaholics tendency to spend excessive time on
work during the evening lead them to have fewer opportunities to recover from work (van
Wijhe et al., 2012). Particularly, they find it difficult to mentally abstain from work demands
during off-job time and this lack of complete detachment increases risks of exhaustion
(Sonnentag & Fritz, 2007). These results represent original and preliminary findings about the
relationship between workaholism and recovery, and about the mediational role of recovery
experiences in the stressor-outcome relationships, research topics few investigated thus far
(Bakker et al., 2012; Kinnunen et al., 2011). Moreover, a strong point of the current study is
that it examined this pattern of relationships on a daily basis, respecting the call for more
within-person studies on workaholism (van Wijhe et al., 2012).

Limitations

This study also has some limitations that should be mentioned. The first one is that data
relied on self-report, raising concerns about common method variance (Podsakoff,
MacKenzie, Lee, & Podsakoff, 2003). However, controlling for the general level of variables
partially precludes the possibility that the effects we found are attributed to general individual
tendencies instead of daily variables. Nevertheless, future research should consider also otherratings (e.g., colleagues or supervisors) and objective ratings, so as to avoid this problem.
Secondly, results mainly concerning relationships between variables considered in this
study are correlational in nature (Stone-Romero & Rosopa, 2008). The multilevel approach
57

provides better estimates for the variables by taking into account the individual baselines, but
inferences about causality and the sequence of effects are quite limited and could not be made
with confidence. Future studies should examine reversed and reciprocal effects of, for
instance, exhaustion and recovery on workaholism. Moreover, the study focused on a
convenience heterogeneous sample of employees and self-employed workers. Therefore, we
should be cautious with generalizing the results. Another limitation of this study is that it
focused specifically on experiences with potential for recovery during off-job time, although
research has indicated that recovery may occur during working time as well (Trougakos, Beal,
Green, & Weiss, 2008).

Practical implications

This study confirmed previous findings that indicated psychological detachment from
work during off-job time as a crucial experience for protecting individuals well-being
(Sonnentag & Fritz, 2007). Therefore, some practical implications can be identified on the
basis of our results. The most important implication for practice is that organizations,
particularly managers and supervisors, should encourage employees to spend their leisure
time in non-job-related activities, with special attention for people with tendencies to
excessive work.
Despite it may be difficult for employees to find a good interaction between all their
different priorities (their work, their family, their leisure activities), they need to be supported
in understanding the importance of engaging in different activities during off-job time and
mentally detach from work on a daily basis (Kinnunen et al., 2011; Volman et al., 2011). One
way to enhance psychological detachment is to keep ones work life separate, as much as
possible, from ones non-work life (Sonnentag et al., 2008). Organizations should encourage
and support this segmentation practices addressing the implicit norms of unlimited
availability and the use of communication devices (e.g., smartphones, laptops with e-mail
access) that make it difficult to mentally and physically detach from work during off-job
hours (Bakker et al., 2012). In general, organizations should question the necessity of longworking-hours culture to support employees finding a healthy work-life balance (Kinnunen et
al., 2011). Moreover, time management training can be considered an important intervention,
especially for employees with working excessively tendencies, to gain conscious control over
their time schedule by setting realistic goals and prioritizing tasks (van Wijhe et al., 2012).

58

This could help them understand when it is time to stop working, and also thinking about
work, at the end of the working days.

59

GENERAL CONCLUSION

In todays societies most people believe that work is healthy and desirable, and the ethic
is based on viewing working hard as virtuous (Burke, 2010). Moreover, work-addicted
employees may be considered an asset for the organizations, since they seem particularly
committed and involved in their job. It has been claimed that workaholism is a well-dressed
addiction and a key virtue (Robinson, 2001). Nevertheless, today workaholism is one of the
most common addictions with many negative consequences for individuals and organizations
(Spruell, 1987).
For these reasons, it seems particularly relevant in the field of Work and Organizational
Psychology to gain more insight in such a problem and its implications, since its causes are
often concealed and negated. Within the organizations, indeed, the ideal worker is those who
work hard and for extended hours; moreover, workaholics are often in denial of their problem.
Therefore, it seems particularly difficult to identify or treat workaholism (Sussman, 2012).
Thus, the primary challenge in dealing with this phenomenon is reaching awareness of its
existence and changing the perception that workaholism is a positive and even desirable
characteristic.
In light of these considerations, the central aim of this dissertation was to give a
significant contribution to the understanding of workaholism. Specifically, the goals of the
studies presented above were: a) to provide supports for future research in the Italian context,
examining the psychometric characteristics of two different measures of workaholism; b) to
examine, within the JD-R model, whether certain working conditions can have a positive
relationship with workaholism and whether job resources, specifically opportunities for
professional development and job security, can moderate this relationship; c) to investigate
consequences of workaholism on three important life domains, namely health, family and
work; d) to explore the relationship of workaholism with recovery experiences and
exhaustion, examining their mediational role in the health-impairment process.

Chapter 1 intended to contribute to the understanding and measurement of workaholism.

Specifically, it examined the properties, in terms of reliability and construct validity, of the
60

Italian version of two measures considered a valid support for researchers and practitioners
interested on workaholism: the Dutch Work Addiction Scale (DUWAS; Schaufeli, Shimazu,
& Taris, 2009; Schaufeli, Taris, & Bakker, 2006; Schaufeli, Taris & Bakker, 2008) and the
Bergen Work Addiction Scale (BWAS; Andreassen, Griffiths, Hetland, & Pallesen, 2012).
Exploratory and confirmatory factor analyses supported the proposed factor structure of
both the scales. Moreover, the construct validity of the two measures was confirmed
throughout investigation of their correlations with other variables indicated in literature as
workaholism-related, specifically: exhaustion (Andreassen, Ursin, & Eriksen, 2007;
Schaufeli, Taris, & van Rhenen, 2008); work-family conflict (Bakker, Demerouti, & Burke,
2009; Bonebright, Clay, & Ankenmann, 2000; Schaufeli et al., 2009, Taris, Schaufeli, &
Verhoeven, 2005); disengagement (Burke, 2000); job satisfaction (del Lbano, Llorens,
Salanova, & Schaufeli, 2010); and some general job-demands (Schaufeli et al., 2008;
Schaufeli et al., 2009). Therefore, also the negative nature of workaholism was confirmed on
the basis of its positive correlations with variables of poor health and negative correlations
with variables of well-being (Andreassen et al., 2012; del Lbano et al., 2010; Molino,
Ghislieri, & Colombo, 2012; Taris et al., 2005). Finally, the two scales proved to be sensitive
to some demographical differences, showing that females report higher levels of workaholism
than males (Molino et al., 2012), and self-employed-workers suffer from work addiction more
than employees (Giorgievski, Bakker, & Schaufeli, 2010; Schaufeli et al., 2008).
Taken together, these findings add to the understanding about workaholism since they
provide empirical support in assessing it in the Italian context, where a lack of studies on this
phenomenon is indicated (Molino et al., 2012). The main relevance of this study is that it
focused on two specific measures which can be used among workers from different
professions and can answer to different research purposes. The DUWAS, indeed, permits to
detach the two components of workaholism conceptualized by the authors, namely working
excessively and working compulsively; whereas the BWAS is an useful instrument focused
on the addictive dimension of workaholism, since it considers the seven components of any
addiction (salience; mood modification, tolerance, withdrawal, conflict, relapse, problems;
Griffiths, 1996).

Chapter 2 was aimed at investigating the role of workaholism in the job-demands

resources model (JD-R; Bakker & Demerouti, 2007), focusing on its potential antecedents in
working context and on its negative consequences. Specifically, the study presented in this
chapter tried to extend the buffer hypothesis of the JD-R model to workaholism, investigating
61

whether job demands are positively related to workaholism, and whether job resources may
moderate this relationship. Moreover, the study also aimed at improving the understanding
about the consequences of workaholism for individuals, their families and organizations at
large, considering direct and indirect relationships with exhaustion (Andreassen et al., 2007;
Burke, 2008; Schaufeli et al., 2008), work-family conflict (Bakker et al., 2009; Bonebright et
al., 2000; Taris et al., 2005) and intentions to change jobs (Burke, 2001; Kravina, Falco,
Girardi, & De Carlo, 2010).
Study results confirmed a positive relationship between job demands and workaholism;
moreover, they also showed that job resources, specifically opportunities for professional
development and job security, can buffer this positive relationship. These findings are
innovative and relevant as they highlight the potential role of organizations in both fostering
and preventing the occurrence of workaholism. Additionally, results confirmed the negative
consequences of workaholism on the three life domains considered. Specifically,
workaholism showed a direct positive relationship with both work-family conflict and
exhaustion, and an indirect positive relationship, fully mediated by exhaustion, with intentions
to change jobs.
These findings are relevant particularly since they consider the processes involved in
workaholism genesis and support a new and emerging point of view which considers
workaholism not as an individual stable trait but an addiction that could, as any other
addiction, be influenced and reinforced by the context. Therefore, the study is significant at a
theoretical, but also at a practical level, since it leads to consider how important is to address
this underestimated phenomenon at both personal and organizational levels.

Finally, Chapter 3 intended to investigate the relationship of workaholism with recovery

experiences and exhaustion by testing within-person variations throughout a diary study and a
multilevel research design. The study presented in this chapter had two main goals: a) to
examine the mediational role of three recovery experiences, psychological detachment,
relaxation and control (Sonnentag & Fritz, 2007), in the relationship between workaholism
and exhaustion; 2) to confirm the positive relationship of workaholism with exhaustion also
on a daily basis, since the majority of previous studies have investigated this relationship
throughout cross-sectional approaches (Andreassen et al., 2007; Burke, 2008; Schaufeli et al.,
2008; Taris et al., 2005).
Study results found a partial mediational role of one of the three recovery experiences
considered, namely daily psychological detachment, in the relationship between daily
62

workaholism and daily exhaustion. These preliminary findings are relevant for different
reasons. Firstly, they confirmed that the workaholics tendency to devote more resources to
their work than to their non-work life can lead them to reduce the possibilities to detach from
work (Bakker, Demerouti, Oerlemans, & Sonnentag, 2012; van Wijhe, Peeters, Schaufeli, &
Ouweneel, 2012). Secondly, it has been confirmed that psychological detachment represents
the core experience of the recovery process (Sonnentag & Fritz, 2007; Volman, Bakker, &
Xanthopoulou, 2011), showing a negative relationship with exhaustion on a daily basis.
Thirdly, the mediational role of daily psychological detachment between daily workaholism
and daily exhaustion was confirmed; therefore, mentally switching off from work during offjob time seems to be particular important since it may reduce the negative relationship of
workaholism with exhaustion.
Finally, findings of this study added to the literature since they confirmed the positive
relationship between workaholism and exhaustion also on a daily basis, respecting the call for
more diary studies on workaholism (van Wijhe et al., 2012). Therefore, they offer more
support for the conclusions of Chapter 2 of this dissertation, as they overcome one of its
limits and confirmed this positive relationship considering within-person variations.

Limitations and needs for future research

Despite the original findings presented so far, this dissertation has also some limitations,
as well as the state of research on workaholism nowadays (Sussman, 2012). Therefore studies
limitations and some needs for future research can be mentioned. The first limitation is that,
among the three studies presented above, two were cross-sectional designs, which excluded
the possibility to draw any conclusions in terms of causal effects in the relationships tested.
Only the third study was a diary study and used multilevel analysis to test between- and
within-person variations. Nevertheless, results mainly concerning variables considered are
correlational in nature (Stone-Romero & Rosopa, 2008). The multilevel approach provided
better estimates for the variables by taking into account the individual baselines, but
inferences about causality and the sequence of effects were still limited. Therefore, in the
future, more longitudinal research or diary studies are needed (van Wijhe et al., 2012),
especially to replicate previous findings and verify their causality. For this purpose, also
reversed and reciprocal effects between study variables should be considered.
The second limitation is that all studies used only single-source self-report data, which
raises questions about common method bias (Podsakoff, MacKenzie, Lee, & Podsakoff,
63

2003). This is considered a recurrent issue in research on Work and Organizational

Psychology (Spector, 1992). Nevertheless, the third study presented in this dissertation was
multilevel, thus it permitted to control the general level of variables and partially precluded
the possibility that the effects we found were attributed to general individual tendencies
instead of daily variables. In general, prevalence data using multiple measures of
workaholism are needed to assist in reaching an adequate estimate of the extent of the
phenomenon (Carroll & Robinson, 2000). Multiple types of data needed include, besides selfreport, collateral reports of supervisors, colleagues and family, as well as objective rates and
company health system utilization (Burke, 2000), in order to minimize the potential effects of
common method variance.
A third limitation of the three studies concerned sample selection, since convenience
samples contacted throughout snow-ball exercise were used. Consequently, samples were
heterogeneous and possibly not representative of the working population at large. On the
positive side, however, we could consider results found applicable to different occupational
contexts and professions, and not only to specific ones. Nevertheless, findings should be
replicated in specific organizations and working places, in order to identify more
contextualized conclusions and practical implications.
Finally, none study proposed qualitative methods or multidisciplinary approach, which
could be particularly helpful in order to deepen the knowledge of workaholism and to clarify
its meaning. Moreover, some more variables could be considered in the studies to extend their
general research purposes. For instance, the second study did not take into account some
dispositional variables and personal resources, although they can play an important role in
generating work addiction (Eysenck, 1997; Kravina et al., 2010; Ng, Sorensen, & Feldman,
2007; Xanthopoulou et al., 2007).

Implications for practice

All in all, this dissertation confirmed that workaholism is a complex and multidimensional problem with the potential to negatively impact the individual, family and
organizational domains. Costs for work addiction could be high for individual and family
well-being on the one hand, and for efficiency and productivity of organizations in the longterm on the other hand. Therefore, our findings highlight the importance of reaching
awareness of the existence of this often ignored or underestimated phenomenon. Individuals
and organizations must understand the problem, its causes and consider viable solutions.
64

Organizations should take appropriate actions to detect potentially workaholic

behaviours such as working at evenings and weekends or the assumption of extra-tasks. This
responsibility lies on managers and supervisors who can more easily than others witness,
detect and signal workaholic behaviours, and suggest and monitor alternative ones.
Employers and managers play a crucial role, as they can be a model and set a good example to
work in a healthy way (Fry & Cohen, 2009). Programmes for leadership development (Avolio
& Bass, 1999; Fry & Cohen, 2009; Yee, Lee, Yeung, & Cheng, 2011) in the work context and
attention to employer-recruitment selection represent important interventions. Leaders can
also actively help their collaborators to behave less dysfunctionally: encouraging them to
delegate some of their work, monitoring their job planning and priorities, giving them specific
times to take breaks and to leave work (Burke, 2010).
The second study particularly underlined the role of organizations in dealing with some
job characteristics thats to say job demands and resources: results, indeed, suggested the
potential role of job demands in increasing workaholism and the potential role of job
resources in buffering this positive relationship. Thus, the burden of finding a good balance
between productivity on the one hand, and individual and organizational short and long-term
costs on the other hand, is on organizations themselves. Particularly, employees should be
exposed to challenging, but not exaggerated, job demands (Bakker et al., 2009) and
organizations should guarantee adequate resources.
More in general, organizations could emphasize the importance of a work-life balance by
setting boundaries between work and leisure time. For instance, they may not encourage
employees to work harder than necessary, to work at home in the evenings and weekends and
can build the conditions in which work can be done in the normal working hours so that
employees can have sufficient time and opportunities to recover from work. Cartwright
(2000) suggested some actions and good practices for breaking out of the long-working-hours
culture, such as scheduling meetings only during core hours, taking regular breaks and not
working at lunch time, saying no to unrealistic deadlines, and monitoring working hours.
Employers should avoid to reward workaholic behaviours through praising, salary
increase, career promotion, but, on the contrary, praise those workers who are productive but
also keen on keeping a good balance between work and life (Bakker et al., 2009). The
employee reward system may be redesigned in such a way that working smart, rather than
working hard, is rewarded (van Wijhe, Schaufeli, & Peeters, 2010).
Considering the individual level, the first intervention needed concerned education in
order to better understand the problem and its potential consequences (Holland, 2008;
65

Sussman, 2012). Moreover, to be able to treat workaholism, clinical assessments might be

completed. Robinson (1998) suggested ten warnings that might be used by clinicians and
practitioners to identify workaholics: (1) hurrying and staying busy; (2) need to control; (3)
perfectionism; (4) difficulty with relationships; (5) work binges; (6) difficulty relaxing and
having fun; (7) brownouts or memory losses; (8) impatience and irritability; (9) selfinadequacy; and (10) self-neglect. Input from family and friends may be essential to validate
clinical assessment of workaholism, particularly when workers tend to deny the problem
(Sussman, 2012).
Further, psychological counseling or psychotherapy may be necessary to prevent or
dealing with workaholism (Ishiyama & Kitayama, 1994). Finally, on the basis of findings of
the third study, employees should be aware of how important mentally detaching from work
is for their recovery from job-strain (Kinnunen et al., 2011; Volman et al., 2011). One way to
enhance psychological detachment is to reach a successful segmentation between work and
non-work life (Sonnentag, Mojza, Binnewies, & Scholl, 2008) and organizations should
support employees addressing the implicit norms of unlimited availability and the use of
communication devices. An example is what the American consulting company Advisory
Board did in September 2012: they ordered their employees not to answer e-mails during the
evening and weekend in order to favour the balance between time at work and time at home,
and in order to show that a change in working culture is possible.

66

REFERENCES

American Psychiatric Association, (2000). Diagnostic and statistical manual of mental

disorders: text revision. 4th ed. Arlington, VA: American Psychiatric Press.
Andreassen, C. S., Griffiths, M. D., Hetland, J., & Pallesen, S. (2012). Development of a
work addiction scale. Scandinavian Journal of Psychology, 53, 265272.
Andreassen, C. S., Hetland, J., Molde, H., & Pallesen, S. (2011). Workaholism and
potential outcomes in well-being and health in a cross-occupational sample. Stress and
Health, 27, e209e214.
Andreassen, C. S., Ursin, H., & Eriksen, H. R. (2007). The relationship between strong
motivation to work (workaholism) and health. Psychology and Health, 22, 615629.
Avanzi, L., van Dick, R., Fraccaroli, F., & Sarchielli, G. (2012). The downside of
organizational identification: Relations between identification, workaholism and wellbeing. Work & Stress: An International Journal of Work, Health & Organisations, 26,
289307.
Avolio, B. J., & Bass, B. M. (1999). Re-examining the component of transformational and
transactional leadership using the multifactor leadership questionnaire. Journal of
Occupational and Organizational Psychology, 72, 441462.
Aziz, S., & Cunningham, J. (2008). Workaholism, work stress, work-life imbalance:
Exploring genders role. Gender in Management: An International Journal, 23,
555566.
Aziz, S., & Tronzo, C. L. (2011). Exploring the Relationship Between Workaholism Facets
and Personality Traits: A Replication in American Workers. The Psychological Record,
61, 269286.
Bakker, A. B., & Demerouti, E. (2007). The job demands-resources model: State of the art.
Journal of Managerial Psychology, 22, 309328.
Bakker, A. B., & Schaufeli, W. B. (2008). Positive organizational behavior: Engaged
employees in flourishing organizations. Journal of Organizational Behavior, 29,
147154.

67

Bakker, A. B., Demerouti, E., & Burke, R. (2009). Workaholism and relationship quality: A
spillover-crossover perspective. Journal of Occupational Health Psychology, 14,
2333.
Bakker, A. B., Demerouti, E., & Schaufeli, W. B. (2003b). Dual processes at work in a call
centre: An application of the job demandsresources model. European Journal of Work
and Organizational Psychology, 12, 393417.
Bakker, A. B., Demerouti, E., & Verbeke, W. (2004). Using the job demands-resources model
to predict burnout and performance. Human Resource Management, 43, 83104.
Bakker, A. B., Demerouti, E., & Euwema, M. C. (2005). Job resources buffer the impact of
job demands on burnout. Journal of Occupational Health Psychology, 10, 170180.
Bakker, A. B., Demerouti, E., Oerlemans, W., & Sonnentag, S. (2012). Workaholism and
daily recovery: A day reconstruction study of leisure activities. Journal of
Organizational Behavior, 34, 87107.
Bakker, A. B., Demerouti, E., Taris, T. W., Schaufeli, W. B., & Schreurs, P. J. G. (2003a). A
multigroup analysis of the job demands-resources model in four home care
organizations. International Journal of Stress Management, 10, 1638.
Bakker, A. B., Hakanen, J. J., Demerouti, E., & Xanthopoulou, D. (2007). Job resources boost
work engagement, particularly when job demands are high. Journal of Educational
Psychology, 99, 274284.
Balducci, C., Cecchin, M., Fraccaroli, F., & Schaufeli, W. B. (2012). Exploring the
relationship between workaholism and workplace aggressive behaviour: The role of
job-related emotion. Personality and Individual Differences, 53, 629634.
Bandura, A. (1997). Self-efficacy: The exercise of control. New York: Freeman.
Beaton, D. E., Bombardier, C., Guillemin, F., & Ferraz, M. B. (2000). Guidelines for the
process of cross-cultural adaptation of self-report measures. Spine, 25, 31863191.
Bollen, K. A., & Long, J. S. (1993). Testing structural equation models. Newbury Park, CA:
Sage.
Bonebright, C. A., Clay, D. L., & Ankenmann, R. D. (2000). The relationship of workaholism
with work-life conflict, life satisfaction, and purpose in life. Journal of Counseling
Psychology, 47, 469477.
Brady, R., Vodanovich, S. J., & Rotunda, R. (2008). The Impact of Workaholism on WorkFamily Conflict, Job Satisfaction, and Perception of Leisure Activities. The
Psychologist-Manager Journal, 11, 241263.

68

Buelens, M., & Poelmans, S. A. Y. (2004). Enriching the Spence and Robbins typology of
workaholism: Demographic, motivational and organizational correlates. Organizational
Change Management, 17, 459470.
Burger, J. M. (1989). Negative reactions to increases in perceived personal control. Journal of
Personality and Social Psychology, 56, 246256.
Burke, R. J. (2000). Workaholism in organizations: Psychological and physical well-being
consequences. Stress Medicine, 16, 1116.
Burke, R. J. (2001). Workaholism in Organizations: The Role of Organizational Values.
Personnel Review, 30, 637645.
Burke, R. J. (2004). Workaholism, self-esteem, and motives for money. Psychological
Reports, 94, 457463.
Burke, R. J. (2008). Work motivations, satisfactions, and health: Passion versus addiction. In
R. J. Burke & C. L. Cooper (Eds.), The long working hours culture. Causes,
consequences and choices (pp. 227251). Bingley, UK: Emerald.
Burke, R. J. (2010). Work hours, work intensity and work addiction: Weighing the costs. In
R. J. Burke & C. L. Cooper (Eds.), Risky Business. Psychological, Physical and
Financial costs of High Risk Behavior in Organizations (pp. 67106). Farnham UK:
Gower Publishing.
Burke, R. J., & Fiksenbaum, L. (2009). Work Motivations, Work Outcomes, and Health:
Passion Versus Addiction. Journal of Business Ethics, 84, 257263.
Burke, R. J., & MacDermid, G. (1999). Are workaholics job satisfied and successful in their
careers? Career Development International, 4, 277282.
Burke, R. J., & Matthiesen, S. B. (2004). Workaholism among Norwegian journalists:
Antecedents and consequences. Stress and Health, 20, 301308.
Burke, R. J., Matthiesen, S. B., & Pallesen, S. (2006). Personality correlates of workaholism.
Personality and Individual Differences, 40, 12231233.
Burke, R. J., & Ng, E. S. W. (2007). Workaholic behaviors: Do colleagues agree?
International Journal of Stress Management, 14, 312320.
Carroll, J. J., & Robinson, B. E. (2000). Depression and parentification among adults as
related to parental workaholism and alcoholism. The Family Journal: Counseling and
Therapy for Couples and Families, 8, 360367.
Cartwright, S. (2000). Taking the pulse of executive health in the UK. Academy of
Management Executive, 14, 1623.
69

Colombo, L., & Ghislieri, C. (2008). The work-to-family conflict: between theories and
measures. TPM - Testing, Psychometrics, Methodology in Applied Psychology, 15,
3555.
Conway, J. M. (2002). Method variance and method bias in industrial and organizational
psychology. In S.G. Rogelberg (Ed.), Handbook of research methods in organizational
and industrial psychology (pp. 344365). Malden, NJ: Blackwell Publishers.
Cortina, J. M., Chen, G., & Dunlap, W. P. (2001). Testing interaction effects in LISREL:
Examination and illustration of available procedures. Organizational Research
Methods, 4, 324360.
Craig, A., & Cooper, R. E. (1992). Symptoms of acute and chronic fatigue. In A. P. Smith &
D. M. Jones (Eds.), Handbook of human performance: Vol. 3. State and trait (pp. 289
339). San Diego, CA: Academic Press.
Dawson, J. F., & Richter, A. W. (2006). Interpreting interaction effects. Retrieved from
http://www.jeremydawson.co.uk/slopes.htm
Del Lbano, M., Llorens, S., Salanova, M., & Schaufeli, W. B (2010). Validity of a brief
workaholism scale. Psicothema, 22, 143150.
Demerouti, E., & Bakker, A. B. (2011). The Job DemandsResources model: Challenges for
future research. South African Journal of Industrial Psychology, 37, 19.
Demerouti, E., Bakker, A. B., Geurts, S. A., & Taris, T. W. (2009). Daily recovery from
workrelated effort during non-work time. In S. Sonnentag, P. Perrewe, & D. Ganster
(Eds.), Current perspectives on job-stress recovery (pp. 85123). Bingley, UK:
Emerald.
Demerouti E., Bakker, A. B., Nachreiner, F., & Schaufeli, W. B. (2001). The Job Demands Resources model of burnout. Journal of Applied Psychology, 86, 499512.
Demerouti, E., Geurts, S. A. E., & Kompier, M. A. J. (2004). Demands and resources as
antecedents of positive and negative interference between work and home. Equal
Opportunities International, 23, 635.
Demerouti, E., Mostert, K, & Bakker, A. B. (2010). Burnout and work engagement: A
thorough investigation of the independency of both constructs. Journal of Occupational
Health Psychology, 1, 209222.
Derks, D., & Bakker, A. B. (2010). The impact of e-mail communication on organizational
life. Cyberpsychology: Journal of Psychosocial Research on Cyberspace, 4, 4.

70

Dormann, C., & Zapf, D. (2004). Customer-related social stressors and burnout. Journal of
Occupational Health Psychology, 9, 6182.
Dziuban, C. D., Shirkey, E. C. (1974). When is a correlation matrix appropriate for factor
analysis? Some decision rules. Psychological Bulletin, 81, 358361.
Etzion, D., Eden, D., & Lapidot, Y. (1998). Relief from job stressors and burnout: Reserve
service as a respite. Journal of Applied Psychology, 83, 577585.
Eysenck, H. J. (1997). Addiction, Personality and Motivation. Human Psychopharmacology,
12, S79S87.
Flowers, C., & Robinson, B. E. (2002). A structural and discriminate analysis of the Work
Addiction Risk Test. Educational And Psychological Measurement, 62, 517526.
Fritz, C., & Sonnentag, S. (2006). Recovery, well-being, and performance-related outcomes:
The role of workload and vacation experiences. Journal of Applied Psychology, 91,
936945.
Fry, L. W., & Cohen, M. P. (2009). Spiritual Leadership as a Paradigm for Organizational
Transformation and Recovery from Extended Work Hours Cultures. Journal of
Business Ethics, 84, 265278.
Geurts, S. A. E., & Sonnentag, S. (2006). Recovery as an explanatory mechanism in the
relation between acute stress reactions and chronic health impairment. Scandinavian
Journal of Work, Environment and Health, 32, 482492.
Gorgievski, M., Bakker, A. B., & Schaufeli, W. B. (2010). Work engagement and
workaholism: Comparing the self-employed and salaried employees. Journal of Positive
Psychology, 5, 8396.
Graves, L. M., Ruderman, M. N., Ohlott, P. J., & Weber, T. J. (2010). Driven to work and
enjoyment of work: Effects on managers outcomes. Journal of Management, 38,
16551680.
Greenberg, J. (1987). Comprehensive stress management. Dubuque, IA: William C. Brown
Publishers.
Griffiths, M. D. (1996). Nicotine, tobacco and addiction. Nature, 386, 18.
Griffiths, M. D. (2005). Workaholism is still a useful construct. Addiction Research and
Theory, 13, 97100.
Harpaz, I., & Snir, R. (2003). Workaholism: Its definition and nature. Human Relations, 56,
291319.

71

Hinkin, T. R. (1998). A brief tutorial on the development of measures for use in survey
questionnaires. Organizational Research Methods, 1, 104121.
Hobfoll, S. E. (2002). Social and psychological resources and adaptation. Review of General
Psychology, 6, 307324.
Holland, D. W. (2008). Work Addiction: Costs and Solutions for Individuals, Relationships
and Organizations. Journal of Workplace Behavioral Health, 22, 115.
Hox, J. (2002). Multilevel Analysis: Techniques and Applications. London: Lawrence
Erlbaum Associates, Publishers.
Hu, Q., Schaufeli, W. B., & Taris, T. W. (2011). The Job DemandsResources model: An
analysis of additive and joint effects of demands and resources. Journal of Vocational
Behavior, 79, 181190.
Ishiyama, F. I., & Kitayama, A. (1994). Overwork and career-centered self-validation among
the Japanese: Psychosocial issues and counselling implications. International Journal
for the Advancement of Counselling, 17, 167182.
James, L. R. (1982). Aggregation bias in estimates of percentual agreement. Journal of
Applied Psychology, 67, 219229.
Kanai, A. (2006). Economic and employment conditions, karoshi (work to death) and the
trend of studies on workaholism in Japan. In R. J. Burke (Ed.), Research companion to
working time and work addiction (pp. 158172). Cheltenham: Edward Elgar.
Kim, S. E., & Lee, J. W. (2007). Is Mission Attachment an Effective Management Tool for
Employee Retention? An Empirical Analysis of a Nonprofit Human Services Agency.
Review of Public Personnel Administration, 27, 227248.
Kinnunen, U., Feldt, T., Siltaloppi, M., & Sonnentag, S. (2011). Job demands-resources
model in the context of recovery: Testing recovery experience as mediators. European
Journal of Work and Organizational Psychology, 20, 805832.
Kraimer, M. L., Wayne, S. J., Liden, R. C., & Sparrowe, R. T. (2005). The Role of Job
Security in Understanding the Relationship Between Employees' Perceptions of
Temporary Workers and Employees' Performance. Journal of Applied Psychology, 90,
389398.
Kravina, L., Falco, A., Girardi, D., & De Carlo, N. A. (2010). Workaholism among
management and workers in an Italian cooperative enterprise. TPM - Testing,
Psychometrics, Methodology in Applied Psychology, 17, 201216.

72

Layne, C. M., Hohenshil, T. H., & Singh, K. (2004). The Relationship of Occupational Stress,
Psychological Strain, and Coping Resources to the Turnover Intentions of
Rehabilitation Counselors. Rehabilitation Counseling Bulletin October, 48, 1930.
Lee, R. T., & Ashforth, B. E. (1996). A meta-analytic examination of the correlates of the
three dimensions of job burnout. Journal of Applied Psychology, 81, 123133.
Leshner, A. I. (1997). Addiction is a brain disease, and it matters. Science, 278, 4547.
Machlowitz M. M. (1980). Workaholics: Living with them, working with them. Reading, MA:
Addison-Wesley.
Mathieu, J. E., Tannenbaum, S. I., & Salas, E. (1992). Influences of individual and situational
characteristics on measures of training effectiveness. Academy of Management Journal,
35, 828847.
Matuska, K. M. (2010). Workaholism, life balance, and wellbeing: A comparative analysis.
Journal of Occupational Science, 17, 104111.
Maume, D. J., & Bellas, M. L. (2001). The Overworked American or the Time Bind?:
Assessing Competing Explanations for Time Spent in Paid Labor. American Behavioral
Scientist, 44, 11371156.
McMillan, L., ODriscoll, M., Marsh, N., & Brady, E. (2001). Understanding workaholism:
Data synthesis, theoretical critique, and future design strategies. International Journal
of Stress Management, 8, 6991.
McMillan, L. H. W., & ODriscoll, M. P. (2006). Exploring new frontiers to generate an
integrated definition of workaholism. In R. J. Burke (Ed.), Research companion to
working time and work addiction (pp. 89107). Cheltenham, UK: Edward Elgar.
Meijman, T. F., & Mulder, G. (1998). Psychological aspects of workload. In P.J.D. Drenth &
H. Thierry (Eds.), Handbook of work and organizational psychology: Vol. 2. Work
psychology (pp. 533). Hove, England: Psychology Press.
Michel, J. S., Kotrba, L. M., Mitchelson, J. K., Clark, M. A., & Baltes, B. B. (2011).
Antecedents

of

workfamily

conflict:

meta-analytic

review.

Journal

of

Organizational Behavior, 32, 689725.

Minirth, F., Meier, P., Wichern, F., Brewer, B., & Skipper, S. (1981). The workaholic and his
family: An inside look. Grand Rapids, MI: Baker Book House.
Molino, M., Ghislieri, C., & Colombo, L. (2012). Working excessively: theoretical and
methodological considerations. Giornale Italiano di Medicina del Lavoro ed
Ergonomia, 34(1 Suppl), A5 A10.
73

Molino, M., Ghislieri, C., & Cortese, C. G. (2013). When work enriches family-life: the
mediational role of professional development opportunities. Journal of Workplace
Learning, 25, in press.
Molino, M., Ghislieri, C., Cortese, C. G., & Bakker, A. B. (2013). Relationship between
recovery experiences and work-family conflict: a study on moderational effects. Paper
accepted for 16th congress of the European Association of Work and Organizational
Psychology, May 22nd-25th, Mnster, Germany.
Moreno-Jimnez, B., Mayo, M., Sanz-Vergel, A. I., Geurts, S., Rodriguez-Muoz, A., &
Garrosa, E. (2009). Effects of work-family conflict on employees well-being: The
moderating role of recovery strategies. Journal of Occupational Health Psychology, 14,
427440.
Mudrack, P. E. (2006). Understanding workaholism: The case of behavioral tendencies. In R.
Burke (Ed.), Work hours and work addiction (pp. 108128). Northampton, MA:
Edward Elgar.
Muthn, B., & Kaplan, D. (1985). A comparison of some methodologies for the factor
analysis of non-normal Likert variables. British Journal of Mathematical and Statistical
Psychology, 38, 171189.
Muthn, L. K., & Muthn, B. O. (1998-2012). Mplus users guide. Seventh Edition. Los
Angeles, CA: Muthn & Muthn.
Netemeyer, R. G., Boles, J. S., & McMurrian, R. (1996). Development and validation of
workfamily conflict and familywork conflict scales. Journal of Applied Psychology,
81, 400410.
Ng, T. W. H., Sorensen, K. L., & Feldman, D. C. (2007). Dimensions, antecedents, and
consequences of workaholism: a conceptual integration and extension. Journal of
Organizational Behavior, 28, 111136.
Nunnally, J. C. (1967). Psychometric Theory. New York: McGraw-Hill.
Nunnally, J. C., & Bernstein, I. H. (1994). Psychometric Theory, 3rd ed. New York:
McGraw-Hill.
Oates, W. (1971). Confession of a workaholic. New York: The World Publishing Company.
Ohly, S., Sonnentag, S., Niessen, C., & Zapf, D. (2010). Diary Studies in Organizational
Research. An Introduction and Some Practical Recommendations. Journal of Personnel
Psychology, 9, 7993.

74

Parasuraman, S. (1982). Predicting turnover intentions and turnover behavior: A multivariate

analysis. Journal of Vocational Behavior, 21, 111121.
Pejtersen, J. H., Kristensen, T. S., Borg, V., & Bjorner, J. B. (2010). The second version of
the Copenhagen Psychosocial Questionnaire. Scandinavian Journal of Public Health,
38, 824.
Petrou, P., Demerouti, E., Peeters, M. C. W., Schaufeli, W. B., & Hetland, J. (2012), Crafting
a job on a daily basis: Contextual correlates and the link to work engagement. Journal
of Organizational Behavior, 33, 11201141.
Piotrowski, C., & Vodanovich, S. J. (2006). The interface between workaholism and workfamily conflict. Organization Development Journal, 24, 8491.
Podsakoff, P. M., MacKenzie, S. B., Lee, J. Y., & Podsakoff, N. P. (2003). Common method
biases in behavioral research: A critical review of the literature and recommended
remedies. Journal of Applied Psychology, 88, 879903.
Porter, G. (1996). Organizational impact of workaholism: Suggestions for researching the
negative outcomes of excessive work. Journal of Occupational Health Psychology, 1,
7084.
Porter, G. (2001). Workaholic tendencies and the high potential for stress among co-workers.
International Journal of Stress Management, 8, 147164.
Preacher, K. J., & Hayes, A. F. (2008). Asymptotic and resampling strategies for assessing
and comparing indirect effects in multiple mediator models. Behavior Research
Methods, 40, 879891.
Quick, J., Murphy, L., Hurrel, J., & Orman, D. (1992). Stress and well being at work:
Assessments and interventions for occupational mental health. Washington, DC: APA.
Robinson, B. E. (1989). Work addiction. Dearfield Beach, FL: Health Communications.
Robinson, B. E. (1999). Spouses of workaholics: Clinical implications for psychotherapy.
Psychotherapy, 35, 260268.
Robinson, B. E. (2000). A typology of workaholics with implications for counselors. Journal
of Addictions and Offender Counseling, 21, 3448.
Robinson, B. E. (2001). Workaholism and family functioning: A profile of familiar
relationships, psychological outcomes, and research considerations. Contemporary
Family Research, 23, 123135.

75

Robinson, B. E., Carroll, J. J., & Flowers, C. (2001). Marital estrangement, positive affect,
and locus of control among spouses of workaholics and spoused of nonworkaholics: A
National study. The American Journal of Family Therapy, 29, 397410.
Robinson, B. E., & Kelley, L. (1998). Adult children of workaholics: Self-concept, anxiety,
depression, and locus of control. The American Journal of Family Therapy, 26,
223238.
Robinson, B. E., & Post, P. (1997). Risk of addiction to work and family functioning.
Psychological Reports, 81, 9195.
Ryan, R. M., & Deci, E. L. (2000). Self-determination theory and the facilitation of intrinsic
motivation, social development, and well-being. American Psychologist, 55, 6878.
Salanova, M., Agut, S., & Peiro, J. M. (2005). Linking organizational resources and work
engagement to employee performance and customer loyalty: the mediation of service
climate. Journal of Applied Psychology, 90, 12171227.
Schaef, K. S., & Fassel, D. (1988). The addictive organization. San Francisco: Harper & Row
Publishers.
Schaufeli, W. B., & Bakker. A. B. (2004). Job demands, job resources, and their relationship
with burnout and engagement: a multi-sample study. Journal of Organizational
Behavior, 25, 293315.
Schaufeli, W. B., Bakker, A. B., van der Heijden, F. M. M. A., & Prins, J. T. (2009).
Workaholism, burnout and well-being among junior doctors: the mediating role of role
conflict. Work and Stress, 23, 155172.
Schaufeli, W. B., Shimazu, A., & Taris, T. W. (2009). Being driven to work excessively hard.
The evaluation of a two-factor measure of workaholism in the Netherlands and Japan.
Cross-Culture Research, 43, 320348.
Schaufeli W. B., Taris, T. W., & Bakker, A. B. (2006). Dr. Jekyll and Mr. Hide: On the
differences between work engagement and workaholism. In R. Burke (Ed.), Research
companion to working time and work addiction (pp. 193217). Northampton, MA:
Edward Elgar.
Schaufeli, W. B., Taris, T. W., & Bakker, A. B. (2008). It takes two to tango: Workaholism is
working excessively and working compulsively. In R. J. Burke & C. L. Cooper (Eds.),
The long work hours culture: Causes, consequences and choices (pp. 203225). Bingly,
UK: Emerald.

76

Schaufeli, W. B., Taris, T. W., & van Rhenen, W. (2008). Workaholism, burnout, and work
engagement: Three of a kind or three different kinds of employee well-being? Applied
Psychology: An International Review, 57, 173203.
Schippers, G. M. (1991). Introduction. In G. M. Schippers, S. M. M. Lammers, & C. P. D. R.
Schaap (Eds.), Contributions to the psychology of addiction (pp. 714). Amsterdam:
Swets & Zeitlinger.
Scott, K. S., Moore, K. S., & Miceli, M. P. (1997). An exploration of the meaning and
consequences of workaholism. Human Relations, 50, 287314.
Shabracq, M. J. (2005). Well-being and Health: What HRM can do about it. In R. J. Burke &
C. L. Cooper (Eds.), Reinventing HRM, (pp. 187206). London & New York:
Routledge.
Shimazu, A., & Schaufeli, W. B. (2009). Is workaholism good or bad for employee wellbeing? The distinctiveness of workaholism and work engagement among Japanese
employees. Industrial Health, 47, 495502.
Shrout, P. E., & Bolger, N. (2002). Mediation in experimental and nonexperimental studies:
New procedures and recommendations. Psychological Methods, 7, 422445.
Siltaloppi, M., Kinnunen, U., & Feldt, T. (2009). Recovery experiences as moderators
between psychosocial work characteristics and occupational well-being. Work and
Stress, 23, 330348.
Smith, D. E., & Seymour, R. B. (2004). The nature of addiction. In R. H. Coombs (Ed.),
Handbook of addictive disorders: A practical guide to diagnosis and treatment (pp. 3
30). New Jersey: John Wiley & Sons, Inc.
Sonnentag, A. (2003). Recovery, work engagement and proactive behavior: A new look at the
interface between nonwork and work. Journal of Applied Psychology, 88, 518528.
Sonnentag, S., & Bayer, U. (2005). Switching off mentally: Predictors and consequences of
psychological detachment from work during off-ob time. Journal of Occupational
Health Psychology, 10, 393414.
Sonnentag, S., Binnewies, C., & Mojza, E. J. (2010). Staying well and engaged when
demands are high: The role of psychological detachment. Journal of Applied
Psychology, 95, 965976.
Sonnentag, S., & Fritz, C. (2007). The recovery experience questionnaire: Development and
validation of a measure for assessing recuperation and unwinding from work. Journal of
Occupational Health Psychology, 3, 204221.

77

Sonnentag, S., Kuttler, I., & Fritz, C. (2010). Job stressors, emotional exhaustion, and need
for recovery: A multi-source study on the benefits of psychological detachment. Journal
of Vocational Behavior, 76, 355365.
Sonnentag, S., Mojza, E. J., Binnewies, C., & Scholl, A. (2008). Being engaged at work and
detached at home: A week-level study on work engagement, psychological detachment
and affect. Work and Stress, 22, 257276.
Snir, R., & Zohar, D. (2008). Workaholism as discretionary time investment at work: An
experience-sampling study. Applied Psychology, 57, 109127.
Sonnentag, S., & Zijlstra, F. R. H. (2006). Job characteristics and off-job activities as
predictors of need for recovery, well-being, and fatigue. Journal of Applied Psychology,
91, 330350.
Spector, P. E. (1992). A consideration of the validity and meaning of selfreport measure of
job conditions. In C. L. Cooper & I. T. Robertson (Eds.), International review of
industrial and organizational psychology (pp. 123-151). Chichester: Wiley.
Spence, J. T., & Robbins, A. S. (1992). Workaholism: Definition, measurement, and
preliminary results. Journal of Personality Assessment, 58, 160178.
Spruell, G. (1987). Work fever. Training & Development Journal, 41, 4145.
Stone-Romero, E. F., & Rosopa, P. J. (2008). The Relative Validity of Inferences About
Mediation as a Function of Research Design Characteristics. Organizational Research
Methods, 11, 326352.
Sussman, S. (2012). Workaholism: A Review. Journal of Addiction Research & Therapy,
S6:001.
Taris, T. W., Geurts, S. A. E., Schaufeli, W. B., Blonk, R. W. B. & Lagerveld, S. E. (2008).
All day and all of the night: The relative contribution of two dimensions of
workaholism to well-being in self-employed workers. Work & Stress: An International
Journal of Work, Health & Organisations, 22, 153165.
Taris, T. W., Schaufeli, W. B., & Shimazu, A. (2010). The push and pull of work: About the
difference between workaholism and work engagement. In A. B. Bakker & M. P. Leiter
(Eds.), Work engagement: A handbook of essential theory and research (pp. 3953).
New York, NY: Psychology Press.
Taris, T. W., Schaufeli, W. B., & Verhoeven, L. C. (2005). Workaholism in the Netherlands:
Measurement and implications for job strain and work-nonwork conflict. Applied
Psychology: An International Review, 54, 3760.
78

Tims, M., & Bakker, A. B. (2010). Job crafting: Towards a new model of individual job
redesign. SA Journal of Industrial Psychology/SA Tydskrif vir Bedryfsielkunde, 36, Art.
#841, 9 pages.
Trougakos, J. P., Beal, D. J., Green, S. G., & Weiss, H. M. (2008). Making the Break Count:
An Episodic Examination of Recovery Activities, Emotional Experiences, and Positive
Affective Displays. Academy of Management, 51, 131146.
van Beek, I., Hu, Q., Schaufeli, W. B., Taris, T. W., & Schreurs, B. H. J. (2012). For Fun,
Love, or Money: What Drives Workaholic, Engaged, and Burned-Out Employees at
Work? Applied Psychology: An International Review, 61, 3055.
van der Colff, J. J., & Rothmann, S. (2009). Occupational stress, sense of coherence, coping,
burnout and work engagement of registered nurses in South Africa. SA Journal of
Industrial Psychology/SA Tydskrif vir Bedryfsielkunde, 35, Art. #423, 10 pages.
van der Hulst, M., & Geurts, S. (2001). Associations between overtime and psychological
health in high and low reward jobs. Work & Stress: An International Journal of Work,
Health & Organisations, 15, 227240.
van Wijhe, C., Schaufeli, W. B., & Peeters, M. C. W. (2010). Understanding and treating
workaholism: Setting the stage fos successful interventions. In R. J. Burke & C. L.
Cooper (Eds.), Risky Business. Psychological, Physical and Financial costs of High
Risk Behavior in Organizations (pp. 107134). Farnham UK: Gower Publishing.
van Wijhe, C., Peeters, M., Schaufeli, W. B., & Ouweneel, E. (2012). Rise and shine:
Recovery experiences of workaholic and nonworkaholic employees. European Journal
of Work and Organizational Psychology, 114.
Vodanovich, S. J., & Piotrowski, C. (2006). Workaholism: A critical but neglected factor in
O.D. Organization Development Journal, 24, 5560.
Volman, F. E., Bakker, A. B., & Xanthopoulou, D. (2012). Recovery at home and
performance at work: A diary study on selffamily facilitation. European Journal of
Work and Organizational Psychology, 117.
Xanthopoulou, D., Bakker, A. B., Dollard, M. F., Demerouti, E., Schaufeli, W. B., Taris, T.
W., & Schreurs, P. J. G. (2007). When do job demands particularly predict burnout?:
The moderating role of job resources. Journal of Managerial Psychology, 22, 766786.
Yee, R. W. Y., Lee, P. K. C., Yeung, A. C. L., & Cheng, T. C. E. (2011). The relationship
among leadership, goal orientation, and service quality in high-contact service

79

industries: An empirical study. International Journal of Production Economics, 141,

452464.
Zapf, D., Vogt, C., Seifert, C., Mertini, H., & Isic, A. (1999). Emotion work as a source of
stress: The concept and the development of an instrument. European Journal of Work
and Organizational Psychology, 8, 371400.

80

Un ringraziamento speciale a Chiara, che con il suo supporto riesce a trasformare

montagne e ostacoli in possibili traguardi.

Thanks to Arnold, because he opened to me the door of a new world, that I liked a lot.

Grazie alla mia famiglia che mi ha supportata fino alla fine, e alle persone che mi sono vicine.
Nonno Lillo questa tesi per te.

81