ArticlePDF Available

How do vacations affect workers’ health and well-being? Vacation (after-) effects and the role of vacation activities and experiences

Authors:

Abstract and Figures

Numerous studies have shown that job stress has detrimental consequences for workers’ health and well-being. A vacation as a prolonged period of recovery from work plays an important role in protecting employees’ well-being. In this dissertation, De Bloom presents the results of her meta-analysis and four studies that focused on three different types of vacations: short vacations in the Netherlands, 9-day active winter sports vacations and long summer vacations. Health and well-being improve during vacation, but these positive vacation effects fade out within the first week of work resumption. Vacation activities are only weakly associated with changes in well-being in all types of vacation whereas vacation experiences showed strong associations with changes in health and well-being. Especially pleasure derived from vacation activities, relaxation, psychological detachment from work and negative incidents during vacation seem to be related to advancements in health and well-being. This dissertation substantiates that a holiday serves as a powerful opportunity to recover from work. Despite the fact that vacation effects are short-lived, vacation memories may temporarily enhance mood and well-being and may act as buffer against future stressors. Vacations may also help people to mentally distance themselves from daily hassles and to put life in perspective which might engender psychological resilience. This research also suggests that a vacation period affects employee’ well-being not only by removing work strain, but also by allowing them to actively engage in pleasant and relaxing vacation activities. By establishing national vacation rights, policy may ensure that employees go on a vacation on a regular basis to recover from work and to preserve long term workability.
Content may be subject to copyright.
How do vacations affect workers’ health
and well-being?
Vacation (after-) effects and the role of vacation activities and experiences
Jessica de Bloom
Printed & lay-out by: Proefschriftmaken.nl || Printyourthesis.com
Published by: Uitgeverij BOXPress, Oisterwijk
Photograph by: I. Wahlen
ISBN 978-90-8891-442-3
How do vacations affect workers’ health
and well-being?
Vacation (after-) effects and the role of vacation activities and experiences
Een wetenschappelijke proeve op het gebied van de Sociale Wetenschappen
Proefschrift
ter verkrijging van de graad van doctor
aan de Radboud Universiteit Nijmegen
op gezag van de rector magnificus prof. mr. S.C.J.J. Kortmann,
volgens besluit van het College van Decanen
in het openbaar te verdedigen op 27 augustus 2012,
om 13.30 uur precies
door
Jessica de Bloom
geboren op 1 maart 1983 te Lingen (Duitsland)
Promotores:
prof. dr. S.A.E. Geurts
prof. dr. M.A.J. Kompier
Manuscriptcommissie:
prof. dr. J.M.A. Riksen-Walraven
dr. J.H.R. Maes
prof. dr. M.J.P.M. van Veldhoven (Universiteit van Tilburg)
To my parents
Life isn’t about waiting for the storm to pass;
it’s about learning to dance in the rain.
7
Contents
Prologue ...................................................................9
1. General Introduction ......................................................11
2. Do We Recover from Vacation? Meta-Analysis of Vacation Effects
on Health and Well-being ..................................................31
3. Effects of Vacation from Work on Health and Well-being: Lots of Fun, Quickly Gone ....57
4. How Does a Vacation from Work Affect Employee’ Health
and Well-being? .........................................................85
5. Effects of Short Vacations, Vacation Activities and Experiences
on Employee’ Health and Well-Being ........................................109
6. Vacation (After-) Effects on Employee’ Health and Well-being,
and the Role of Vacation Activities, Experiences and Sleep ......................137
7. General Discussion ......................................................165
Epilogue .................................................................195
Appendix .................................................................199
1. Qualitative study on vacation ...........................................199
2. Summary ...........................................................204
3. The Journey So Far: Curriculum Vitae ....................................218
9
Prologue
“Ladies and gentlemen, this is your captain speaking. On behalf of myself and the rest of the
crew, I like to welcome you onboard.
We expect a smooth journey today and our route will take us to several destinations around the
world: two stopovers are scheduled in a winter sports area and one in a Dutch holiday park. In
addition, we join a group of Dutch vacationers during their long summer vacations. Expected
flight time is 4 hours and 21 minutes. For safety instructions, please refer to our in-flight safety
lecture.
I will get back to you during your flight back home. In the meantime, sit back, relax and enjoy
your journey through vacation research.
And now, please fasten your seatbelt and raise your chair to the upright position. We are ready
for takeoff.”
11
Chapter 1
General Introduction
General Introduction
13
1
1.1. Introduction
In the pyramid of Djoser, archeologists discovered hieroglyphs engraved into the walls of the
monument roughly meaning ‘Hadnachte and his brother Panachti have been here to make an
excursion and enjoy Memphis’ (Hachtmann, 2008). This ‘graffiti’ dating from briefly before the
birth of Christ and other similar remains from ancient times are viewed as the first evidence for
touristic activities and vacations. Consequently, the etymologic origin of the word vacation also
dates back to the time of the Roman Empire and stems from the Latin word ‘vacatio’, which
means ‘being free from, being at leisure or having time for’. Apparently, the need to be free from
duties and the desire to engage in leisure activities has long been acknowledged.
However, since the effects of free time have often been called into question, time
off from work, and especially longer periods of leisure were, and to a certain extent still are, a
rare treasure for large parts of society. In particular the working class had to struggle hard in
order to be entitled vacation rights. And up till now, in many countries there are still no national
regulations which warrant free time and vacations for every employee.
Today, more than 2000 years after Hadnachte en Panachti went on their vacation to
Memphis, the world and the nature of work changed tremendously. Nonetheless, the ancient
definition of vacation covers our current understanding of vacation very well: a time free from
work, a time for leisure and a time for ourselves. Depending on the research field, definitions
of vacation tend to focus on slightly different aspects. Whereas in the field of tourism the focus
is on travel (World Tourism Organization, 1995), in occupational health the emphasis is on
suspension of work. As the focal point of this thesis is on vacation as a respite from work, we
define vacation as a prolonged period of absence from work granted to an employee, used
for rest, recreation or travel and lasting more than two days (see also Sluiter, Frings-Dresen,
Meijman, 2000; Merriam Webster, 2011).
Today, more than 2000 years after the concept of vacation emerged, more and more
employees are guaranteed the right on free time and vacation so that they can recover from
work and pursue activities of their own choice. However, even 2000 years after the discovery
of the pleasures of vacations, the evidence for positive effects of vacations on health and well-
being (H&W) remained anecdotal rather than scientifically proven for a long time. Only very
recently, vacations became a research topic in science.
Whilst vacations and recovery constitute a very young research field, the detrimental
effect of job stressors, which can be seen as the antipode of recovery, has been fairly well
established these days. Exposure to job stressors may directly elicit potentially harmful
physiological responses (e.g. Belkic, Landbergis, Schnall & Baker, 2004; Geurts & Sonnentag,
2006; Hjortskov et al., 2004) as well as indirectly via unhealthy life styles such as smoking,
14
Chapter 1
1
alcohol consumption, unhealthy diets, lack of exercise, and disturbed sleep (e.g., Ezoe &
Morimoto, 1994; Åkerstedt, 2006). Particularly when physiological responses, such as elevated
levels of blood pressure, heart rate, catecholamines and cortisol, prolong after demands and
stressors have ended, H&W are seriously at risk (e.g., Brosschot, Van Dijk & Thayer, 2007;
Mommersteeg, 2006; Vrijkotte, van Doornen & de Geus, 2000; Schnall, Schwartz, Landsbergis,
Warren & Pickering, 1998).
Consequently, recovery as an antagonist of work stress plays a crucial role in
protecting employees’ H&W (Geurts & Sonnentag, 2006). Sluiter et al. (2000) distinguished
four different types of recovery based on duration and time span after work: microrecovery
(first minutes after task performance), mesorecovery (ten minutes to one hour after task
performance), metarecovery (one hour to two days after work) and macrorecovery (more than
two days after work).
Recent diary studies have revealed that workers often recover insufficiently during
shorter respites like regular evening hours and weekends, for instance due to working
overtime or cognitive processes like having stressful thoughts about past or present stressors
(rumination) or future stressors (worrying) (Van Hooff, Geurts, Kompier & Taris, 2007; Geurts &
Sonnentag, 2006; Fritz & Sonnentag, 2005). The consequences of this inadequate recovery are
stress-related illnesses like burnout and severe sleep disturbances which are also prominent
determinants of long term sickness absence (Åkerstedt, Kecklund, Alfredsson & Selen, 2007;
Maslach, Schaufeli & Leiter, 2001, Geurts, Kompier & Gründemann, 2000).
Vacation as a relatively long period of rest is presumably a prime opportunity to
recover and a powerful weapon against work stress and its negative consequences. Indeed, a
longitudinal study by Gump and Matthews (2000) showed that not taking annual vacations was
associated with a higher risk of morbidity and mortality during a nine-year period.
In Western industrialized societies and especially in Europe, we may have the idea
that several paid vacations a year are a natural matter of fact for most employees (e.g. on
average, the Dutch go on vacation three times per year, NBTC NIPO, 2009). The brief history of
vacation and the worldwide comparison that we present in the next sections will prove differently.
Whilst the idea that free time for body and soul is important is actually very old, the concrete
implementation of this concept into everyday life of employees is rather young. Eventually, it
was not before the 1920’s that the first countries in Europe established the employees’ right to
go on a paid vacation (Hachtmann, 2007).
These days, there are still considerable differences regarding vacation rights. The
following sections of this introduction will therefore also focus on worldwide differences in
vacation legislation and its actual utilization.
General Introduction
15
1
After putting the concept of vacation in a historical and an international perspective,
we turn to scientific theories that may explain why vacations may affect employee’ H&W. The
presentation of our research questions and an outlook on the following chapters will bring this
introduction to a close.
1.2. Vacation in the past: a historical perspective
In the following, we will describe the most important developments regarding vacation across
history.
Classical Antiquity: Due to inscriptions carved into the pyramids of Djoser and the
famous writings of ancient historians like Herodotus (484- 425 BCE) and Pausanias (115- 180
CE), we know that parts of the aristocracy and merchants of the Roman Empire traveled far
distances and visited famous sights like the Great Sphinx or the pyramids of Giza. Great sport
events like the Olympic Games held in Greece also attracted many visitors from near and
far and visiting these happenings could therefore also be considered early types of tourism
(Hachtmann, 2007). However, these ancient travelers did not necessarily call their journey a
vacation, because sightseeing and enjoyment were de facto only by-products of business
travels which were either required for salesmen to earn a living or advantageous for aristocrats
to maintain important contacts in order to become or remain prominent leaders of society. The
excellent infrastructure of the Empire and the road network of more than 90,000 kilometers of
solid roads also enabled the temporal relocation of the Roman high society from the crowded,
hot capital town to the villa’s at the well-tempered seaside between Ostia and the Gulf of Naples
(Casson, 1994). Nevertheless, business often continued in the summer months. Therefore, it
may be debated whether this phenomenon could be considered vacationing, because the
upper class was eventually not freed from all work duties.
Middle Ages: After the collapse of the Roman Empire and innumerable accompanying
wars, the entire infrastructure, including the road system, was destroyed and traveling was
therefore nearly impossible in the Middle Ages. In these times, Europe was fallen apart and
diplomatic conflicts between sovereigns, separate monetary systems, tolls and continuing
wars made traveling very complicated and dangerous. Moreover, most monarchs did not start
to make a cadastral survey of their country before the 17
th
century. Accordingly, orientation
during a journey was hardly possible. During rides through dark forests, travelers were
often robbed or even murdered by the lawless that lived there. Because traveling normally
16
Chapter 1
1
comprised walking or, if you were more lucky and wealthy, riding on a horseback, going on a
voyage meant being exposed to cold, heat, rain and storm. Furthermore, boarding houses or
hotels rarely existed and were therefore hopelessly overcrowded, dirty and a breeding ground
for epidemics. Consequently, traveling was an exciting, dangerous affair and our medieval
ancestors would surely not hit upon the idea to go on a vacation just for relaxation and fun
(Ohler, 1986). Furthermore, the concept of free time and rest were certainly not well-developed
in these times and most people worked long hours for at least six days a week.
Modern Age: In the 15
th
century, carriages, which were already used by the Romans
but which fell into oblivion after the decline of the Roman Empire, were reinvented and evolved.
Little by little, these carriages became a popular means of transportation and a network of
stagecoaches came into being. Accordingly, traveling became somewhat more convenient,
faster and safer. However, traveling was in principle reserved for rich elites which could afford
a coach ride with an escort to guarantee safety (Hachtmann, 2007).
Within the times of romanticism, writers like Jean-Jacques Rousseau (1712- 1778)
glorified nature as a counterpart of the increasingly overpopulated and industrialized cities.
Nature became associated with freedom, beauty and purity and it also got more predictable
due to progresses within the exact sciences. As a consequence, short excursions into nature
became popular for the adherents of the middle class and the high society alike (Gräf & Pröve,
1997; Griep & Jäger, 1986).
The birthplace of modern tourism and vacation for all classes lies in Great Britain
because of two reasons. In the first place, the invention of the steam train which became
operative in England in 1840, stimulated travel across greater distances on a large scale. In the
second place, Thomas Cook (1808- 1892), a visionary pioneer of tourism, lived in England and
precipitated developments in tourism.
In 1841, the former carpenter, salesman and vehement teetotaller Thomas Cook
organised a train ride for a group of people who wanted to attend a protest demonstration
against alcohol consumption. Astonished by the great success of this ride, the simplicity and
the financial gain of such a tour for a big group of people, Cook decided to organize more
guided tours and used the train as means of transportation. In 1862, Cook also arranged
accommodations at the travel destination and established the first travel agency which offered
affordable package tours. Cook was also one of the first farsighted people who claimed that
everybody, ‘even’ the working classes, needed and deserved a vacation. Correspondingly, he
organized moonlight excursions starting on Saturday evening for the working class members
who worked six days a week and who were only free on Sunday (Hachtmann, 2007).
General Introduction
17
1
Beginning of the 20
th
century: At rough estimate, in 1900 just 0.1 percent of the
working class had the right to go on a short, unpaid vacation (Reulecke, 1976). Before 1900,
most workers had to work long hours for six days a week in order to earn a living. Only Sunday
and religious holidays were off. For the working class, there was simply no time or money to
go on vacation.
For white collar employees (e.g. salesman, civil servants, teachers) a couple of free
days off per year became standard sooner than for blue collar workers. However, for most of
these white collar workers, vacation was also unpaid in the beginning. Nevertheless, wealthy
white collar workers loved to go on a vacation. Yet, a vacation for fun and pleasure was societally
inappropriate and there had to be a justified reason to take a time out from work (Hachtmann,
2007). This was especially true for employees in the United States whose work culture was
strongly influenced by Calvinistic Christian values. Taking a holiday was seen as idleness and
morally untenable. For that reason, vacationing was often called a ‘cure’ and underpinned by
vague ‘medical’ reasons like the need for fresh air and clean water.
Of course, it did not take long before it became clear that not all vacationers in the
flourishing spa resorts at those times were recovering from illness. After all, these vacation
resorts offered amusement like stage plays, concerts, dance parties or even gambling
casinos. As countermovement, churches offered affordable summer camps without profane
temptations like drinking, gambling or indecent contact between men and women. For the first
time in history, even blue collar workers had the chance to go on a vacation, although many
workers could not yet get the necessary days off from work (Aron, 1999).
In the same period of time, public discussions about the (dis-) advantages of vacation
emerged. In a very interesting article in the New York Times Magazine (1919) entitled ‘How long
should a man’s vacation be’, famous businessman, widely known scientists and physicians
expressed their opinion on this issue. The key message from most of these people was that
vacationing is supposed to be beneficial for health, work ethic and work performance and
could therefore imply an economic advantage.
This kind of discussions indicated a revolution in thinking about work. Inhuman work
was increasingly criticized and workers established labour unions to fight for improved working
conditions. This rebellion again started in England to spread very rapidly across Europe and the
US. Essential improvements the unions contended for were safer work environments as well as
shorter work weeks and paid vacations. These aspirations proved successful. In 1919, Austria
was the first country worldwide to introduce a law which guaranteed workers the statutory right
on a paid vacation (Hachtmann, 2007). Other countries like Poland, Luxemburg and Denmark
followed soon. At present we even have European wide laws on vacation (see also paragraph
1.3 Vacation today: an international perspective).
18
Chapter 1
1
Before and after World War II: From 1925 on, holidays were misused as a means
to propagandize fascism in Europe. The Italian dictator Benito Mussolini (1883-1945) was
the first politician to set up the ‘Opera Nazionale Dopolavoro’, meaning ‘National club for
recreation’. This organization served as an example for Adolf Hitler’s (1989-1945) ‘Kraft durch
Freude’ organization and the Dutch ‘Gemeenschap Vreugde en Arbeid van het Nederlandsch
Arbeidsfront’ (both roughly meaning ‘Strength through joy’). The vacations and free time
activities propagandised by these associations should engender fascist worldviews by raising
patriotism, solidarity and a sense of belonging. An example par excellence for Nazi holiday
dreams coming true is ‘Prora’, a 5 km long, massive beach resort on the island of Rügen,
providing space for about 20 000 vacationers.
During World War II, vacationing came to a halt, because Europeans were in first
instance worried about fundamental human needs for food, shelter and protection. Boarding
houses and hotels which withstood the war were often used as temporary refuges for the
homeless.
During the reconstruction of the destroyed European cities after World War II and the
economic growth in the 1950’s, tourism also flared up. Labour unions took up their work again,
ascribed vacationing high priorities and set up special associations which organized vacations
for the working population. Thanks to organizations like the ‘Workers’ Travel Association’ in
Great Britain, ‘Tourisme et Travail’ in France or ‘Folk-Ferie’ in Norway, finally more and more
members of the working class were able to go on holidays (Hachtmann, 2007).
Gradually, commercial companies took over the touristic work from the labor
organizations. According to Keitz (1997), in 1950 about twenty percent of the European
population went on a vacation. Within ten years, this percentage rose to about fifty percent.
From the beginning, small and relatively wealthy countries like the Netherlands or Switzerland
were the countries with the highest percentages of inhabitants who go on vacation, either in the
home country or abroad (Hachtmann, 2007).
1.3. Vacation today:
an international perspective
From the historical overview we learned that touristic activities and vacationing already started
in the time of the Roman Empire and the Middle Ages and mostly implied business travels
or adventurous expeditions. The modern form of vacation which is predominantly aimed at
recovery, rest and recreation originated in Great Britain as a consequence of labour unions
actions and spread out throughout Europe and great parts of the world (Hachtmann, 2007).
General Introduction
19
1
Not surprisingly though, it was also in Europe where the first countries introduced national
regulations which guaranteed their workers paid vacation.
In 1993, the European Union established legal rights to at least four weeks of paid
vacation per year (EC of the European Parliament and the Council, 1993). This paid annual
leave may not be replaced by an allowance in lieu in order to make sure that employees
use vacation as recovery time instead of sending vacation days astray for money. Countries
belonging to the European Union have to comply with this directive and most countries also
introduced national rights over and above the EU statute for paid annual leave.
On average, Europeans have 25 days paid vacation per year. While employees in
some countries (e.g. Estonia, Cyprus) just receive the legal minimum of twenty days per year,
employees in most other countries have more paid vacations. Sweden (33 days), Denmark (30
days) and Germany (30 days) are the countries which endow their workers with the most days
of paid vacation per year (European Foundation for the Improvement of Living and Working
Conditions, 2008).
In line with the European directive, in the Netherlands, employers are required to
grant their employees at least four times their average work hours a week as vacation time
(Rijksoverheid, 2011). Accordingly, an employee who works fulltime (40 hours a week) should
at least get twenty days off (40*4 =160 hours or 20 days). The national average of paid annual
leave in the Netherlands is even higher: 26 days (European Foundation for the Improvement of
Living and Working Conditions, 2008).
Besides the minimum number of days off, every European country can determine
additional regulations regarding vacations. Ray and Schmitt (2008) studied these regulations
in different countries across the world. They found that, in some countries, older workers (e.g.
in Norway, Greece), employees under the age of eighteen (e.g. Switzerland, Germany) or shift
workers (e.g. in Austria) need to be granted more vacation days. Some countries also define
in which period of the year vacation should be approved (e.g. in Finland, Sweden) or how long
workers ought to be free consecutively (e.g. Denmark, France).
In the Netherlands, employers must follow the preferences of their employees
regarding the scheduling (timing, duration) of their vacation. This rule may only be strayed from
by way of exception (e.g. the public government determines the timing of vacations for teachers
and construction workers). The Netherlands are also one of the very few countries worldwide
which require employers to offer their employees a bonus pay on top of their regular wage for
vacation expenses (to the amount of eight percent of the gross annual salary). According to
Ray and Schmitt (2008) from all European countries, besides the Netherlands, only Sweden
and Austria require employers to pay employees at a premium rate while they are on vacation.
20
Chapter 1
1
In many countries outside the European Union, paid vacations are not required and
regulated by federal laws, let alone the financial compensation for vacation expenses. For
example, in the United States, employees do not have the right to take a paid annual leave
(Ray & Schmitt, 2007). Each state has its own legislation and, for the greatest part, commercial
companies can freely decide on the amount of days paid annual leave they accredit to their
workers. Accordingly, many companies use vacation as a stake in salary negotiations or as an
incentive for high performance and seniority. As a consequence, Americans have on average
only ten days vacation per year at their disposal (United States Department of Labour, 2010).
Moreover, the right to go on vacation at all and vacation time are inequitably distributed in
the US: low-wage employees are less likely (69%) than high-wage workers (88%) to have
paid vacations. In addition, lower-wage employees receive only seven days vacation per year
whereas higher-wage workers receive 14 days. The same applies to part-timers who are less
likely to have paid vacations (36%) than are full-timers (90%) (Ray & Schmitt, 2007).
Also in other western, industrialized countries, the number of vacation days required
by law is usually much smaller than in Europe. For example, in Canada employees have two
weeks vacation per year (Ray & Schmitt, 2007). In Asian countries employees are usually entitled
less days than in Europe as well. Japan and Hong Kong for example grant their employees ten
days vacation (The Japan Institute for Labour Policy and Training, 2002; Census and Statistics
Department of the Government of the Hong Kong Special Administrative Region, 2001).
Unfortunately, there is not yet a reliable overview of paid vacations rights across the
whole world and due to difficulties in obtaining the necessary data (conceptual differences
regarding vacation, language differences, lacking statistical data) this will presumably not
change in the near future.
Summing up, vacation rights and their utilization differ extremely across countries
(e.g. Europe versus America) and even within countries (e.g., in the United States, there may
be employees who have no vacation and high achieving seniors who have 25 days of paid
vacation). After looking at vacation from a historical and an international perspective, we will
now focus on the questions, why and how a vacation may contribute to H&W, which is central
to this thesis.
1.4. Mechanisms through which vacation may contribute to
recovery
Until now, many vacation studies remained rather mute about possible underlying mechanisms
through which vacations may contribute to recovery. Below, we will therefore present five
theories that can explain why vacations may be beneficial for H&W.
General Introduction
21
1
Having vacation may contribute to recovery from work through an active and a
passive mechanism. The passive mechanism reflects a direct release from daily exposure to
job demands. The active mechanism through which vacation may facilitate recovery covers the
engagement in valued, pleasant, self-chosen non-work activities and the opportunity to spend
quality time with close others.
The two most influential recovery theories, Effort-Recovery Theory (Meijman & Mulder,
1998) and Allostatic Load Theory (McEwen, 1998), merely presuppose the passive recovery
mechanism. Both theories share the assumption that removal of demands previously put on
the individual’s psychobiological systems is a necessary prerequisite for recovery to occur
(Sonnentag, 2001).
The active mechanism underlying vacation is best represented by three theories:
Conservation of Resources Theory (Hobfoll, 1998), Self-Determination Theory (Ryan & Deci,
2000) and Broaden- and -Build Theory (Fredrickson, 2001). The starting point of these theories
is the assumption that humans are ‘masters of their own fate’ who can actively and freely
pursue their own interests and intentionally strive for desirable outcomes. Vacation forms the
breeding ground for self-fulfillment and energy replenishment. Therefore, according to these
theories, recovery occurs because vacationers are able to engage in self-chosen, pleasant
activities and spend time with significant others.
Effort-Recovery Theory (Meijman & Mulder, 1998): The basic idea of Effort-Recovery
Theory is the necessity of mobilization of capacities and resources to meet the demands
of work. This effort expenditure at work has psycho-physiological costs or load reactions.
However, acute load reactions (e.g., fatigue), that are unavoidably associated with working, will
not have long term negative health consequences as long as workers recover sufficiently after
work. During time after work, effort is no longer expended and the psychobiological systems
that were activated during work time will return to baseline level. Recovery is correspondingly
seen as a period of rest in which employees are relieved from the demands that are otherwise
acting upon them. This absence of demands actually enables replenishment of resources.
However, recovery may be inadequate due to prolonged exposure to high (work) demands and/
or due to cognitive processes (e.g., worrying and rumination) that prolong physiological activity
even if employees are not directly exposed to demands during the recovery period (Geurts &
Sonnentag, 2006). When recovery is insufficient, employees will have to perform on the job
while being in a suboptimal state, which imposes an even higher demand on the recovery
process. The resulting accumulated negative load effects may have adverse consequences
on H&W.
22
Chapter 1
1
Allostatic Load Theory (McEwen, 1998): As a physiological theory of stress, Allostatic
Load Theory constitutes a model for the fluctuation of physiological systems within the body
to meet stressful demands. The underlying principle is to achieve stability through change
(Aronsson, Svensson, Gustafson, 2003). This regulation process is called allostasis. Repeated
or prolonged physiological activation may disturb an organism’s precarious homeostatic
(sympathetic-parasympathetic) balance. This disturbed balance and cumulative cost to the
body, called allostatic load, will manifest itself in chronic overactivity or inactivity of crucial
bodily systems (e.g., the immune system). Repeated or chronic stress leads to allostatic load
building up. Hence, allostatic load denotes the psychophysiological costs of chronic exposure
to stress. Therefore, complete unwinding from load effects built up at work is crucial for
preserving H&W (Sonnentag & Geurts, 2009).
Conservation of Resources Theory (Hobfoll, 1998): This theory claims that people
strive to obtain, protect and build resources that have specific importance to them. Strain
develops when these valued resources are threatened, lost, or not gained after having invested
in them. ‘Resources’ refer to a broad category including external objects and conditions such
as relationships, as well as personal characteristics and energies. For the aim of conceptual
clarity, in relation to vacation, we define ‘resources’ as time and attention devoted to highly
valued activities (e.g. hobbies, quality time with partner and family) that have the potential
to produce energy. Based on insights from human physiology, Marks (1977) stated that the
consumption of energy is necessary to stabilize the production of energy, and that particularly
the engagement in valued activities will produce energy. Along these lines, vacation may
constitute a possibility to replenish depleted resources and gain resources, because it is
an excellent occasion to engage in freely chosen and energizing activities such as the (re)
connection with family and friends.
Self-Determination Theory (Ryan and Deci, 2000): Regarding vacation effects,
autonomy and relatedness are the crucial elements of this theory. Autonomy and relatedness
are considered fundamental human needs, whereby satisfaction of these needs elicits positive
emotions, and the neglect of these needs leads to negative affect. Autonomy to initiate
behavior of one’s own choice refers to volition and the experience of self-determined behavior.
Relatedness refers to the feeling of being closely connected to others. Earlier research has
demonstrated that workers experienced higher positive and lower negative affect during off-
job time (i.e., weekends) than during work periods due to satisfaction of the workers’ need for
autonomy and relatedness (Ryan, Bernstein & Brown, 2010; Reis, Sheldon, Gable, Roscoe
& Ryan, 2000; Sheldon, Ryan & Reis, 1996). Following this reasoning, a vacation as a pre-
General Introduction
23
1
eminent opportunity to engage in activities of one’s own choice (autonomy) and to connect
to close others (relatedness) may fulfill the basic needs of autonomy and relatedness, which
should result in positive emotions and higher levels of H&W.
Broaden- and -Build Theory (Fredrickson, 2001): In this theory, positive emotions are
also considered crucial for H&W. According to Tugade and Fredrickson (2007), positive and
negative emotions have complementary adaptive functions and effects. Whereas negative
emotions evoke restricted and survival oriented behavior, positive emotions are supposed
to broaden people’s thought-action repertoires, thereby encouraging varied, novel and
exploratory thoughts and actions. The experience of positive emotions, such as pleasure, is
associated with the production of certain hormones in the brain’s ‘pleasure reward’ system (e.g.,
serotonin, dopamine) that may quickly down-regulate psycho-physiological stress responses
(Esch & Stefano, 2004). In an experiment on cardiovascular reactivity, Fredrickson, Mancuso,
Branigan and Tugade (2000) demonstrated how positive emotions can indeed rapidly undo the
unfavorable cardiovascular arousal induced by negative emotions. According to this theory,
positive emotions do not only have short term beneficial effects, but also have long term profits
by building enduring personal resources like intellectual growth, creativity, new skills, social
support, coping capacities and psychological resilience. These personal resources may also
function as buffers for future stressors.
Summing up, the positive effect of vacation on workers’ H&W will partly be determined by the
relief from work demands and job stressors. Moreover, autonomy to engage in behaviors of
one’s own choice, relatedness to friends and family, and the experience of positive emotions
associated with vacation is expected to boost the positive impact of vacation on H&W beyond
the sheer liberation from demands.
1.5. Central research questions
Despite its assumed great recovery potential, vacation as a prototypical recovery opportunity
(i.e. macrorecovery) is a neglected research topic so far. Initially, a vacation was simply seen
as a control occasion for the absence of stress (Eden, 2001). This means, the major interest
was not in vacation as such but rather in demonstrating that on-and off- job situations differ
in levels of psychological stress. Consequently, many studies on vacations applied a pre-post
design in examining the effects of vacation, and pre-post changes in well-being were attributed
24
Chapter 1
1
to the in-between vacation period. However, the sequential occurrence of phenomena does not
mean that they are causally related (Eden, 2001).
In order to estimate the true contribution of vacation to well-being, reliable and valid
on-vacation measurements are therefore strictly necessary. Accordingly, we defined a vacation
effect as the difference in H&W before and during vacation. Research question 1 is:
1. Vacation effect: Do health and well-being improve during vacation?
In case of an increase of H&W during vacation (i.e. a positive vacation effect), the next important
question is how long this effect lasts after work resumption. As vacation comes to an end and
positive effects are assumed to fade out after returning home and resuming work, the positive
effect still persisting after vacation is labeled a vacation after-effect. Research question 2 is:
2. Vacation after-effect: How long do vacation effects last after work
resumption?
Besides the mere absence of work, representing the passive recovery mechanism, vacation
may positively influence H&W because it enables vacationers to spend time on valued free-
time activities. In line with our reasoning regarding the active mechanism of recovery, a vacation
may provide vacationers with the opportunity to go through unique and pleasant experiences.
For that reason, vacation activities and the associated experiences may be core elements that
promote or impede recovery processes during and after vacation. Research question 3 is:
3. Activities & experiences: How do vacation activities and experiences relate to
changes in health and well-being during and after a
vacation period?
1.6. Thesis outline
In this dissertation, we present and discuss our research examining the effects of vacation and
the role of activities and experiences as possible determinants of these effects. We started our
research project by conducting a meta-analysis (Chapter 2) which also gave us the opportunity
to distil up till now unanswered questions regarding vacation (after-) effects and to develop a
solid research design to investigate developments in and determinants of H&W during and
after a vacation period.
General Introduction
25
1
After finishing our meta-analysis, we carried out three longitudinal field studies on
three different types of vacations: short vacations in the Netherlands (Chapter 5), 9-day active
winter sports vacations (Chapter 3 and Chapter 4) and long summer vacations (Chapter 6). We
will present the results of these studies in the chronological order in which we conducted the
data collection: 1) meta-analysis, 2) winter sports vacations, 3) short vacations and 4) summer
vacations.
Moreover, before conducting our first quantitative study, we carried out an explorative
survey with predominantly open questions in order to get a general idea about possible effects
of vacation on well-being and the determinants of these effects (for a brief summary of the
results, see Appendix 1). Although the convenience sample which completed the digital
questionnaire was rather small (N = 63) and highly educated, the answers helped us to
interpret certain findings of our empirical studies. Therefore, we will also refer to some findings
from this survey in the general discussion.
In Chapter 2, we present the findings of our meta-analysis on vacation. The main aim
of this study was to examine the results from earlier studies on vacation (after-) effects and the
role of vacation activities and experiences. A second aim was to learn about methodological
considerations for future research. Based on this information, we were able to set up our first
data collection on winter sports vacations.
We started collecting data on winter sports vacations, because this type of vacation is
more uniform for vacationers than other types of vacations. This means that vacation activities
(mostly skiing during the day, socializing in the evening) and the period of the year were roughly
comparable for all vacationers. In addition, the duration of the vacation (9 days on average)
and the time before and after departure (i.e. days off before and after vacation) were similar for
all holiday takers. In the first examination of the data on winter sports vacations (Chapter 3), we
focused on the effect of vacation on seven different H&W indicators during and after an active
winter sports vacation (N = 96). Later on, we combined these single indicators into one overall
indicator of H&W.
In Chapter 4, also based on our rich dataset on winter sports vacations, we tested
whether vacation (after-) effects are a universal phenomenon (i.e. whether increases in H&W
apply to all employees). But most importantly, we investigated the relation between vacation
activities and experiences, on the one hand, and changes in H&W during vacation, on the other.
We distinguished between various types of vacation activities: work-related, physical, social
and passive activities. Pleasure derived from activities and negative incidents constituted the
vacation experiences that we also examined in Chapter 4.
In Chapter 5, we present the results from our second empirical study on a very popular
type of vacations: long weekends and midweek vacations in a holiday park in the home country
26
Chapter 1
1
(4.5 days on average, N = 80). Vacation (after-) effects on H&W again formed the basis of this
study. Moreover, we studied the impact of vacation activities and experiences on changes
in H&W during vacation. We further extended this study by also exploring the relationship of
vacation activities and experiences with H&W changes after vacation. In addition, we expanded
the vacation experiences by including recovery experiences (i.e. psychological detachment
from work, relaxation experiences, mastery experiences and control), time spent on and quality
of conversations with the partner.
In our last empirical study, presented in Chapter 6, we investigated long summer
vacations which constitute, at least in the Netherlands, the most prototypical and long lasting
type of vacation (minimally 14 days, 23 days on average, N = 54) with also a lot of variation in
vacation type (e.g. backpacking, wellness, family visits) and activities. This study again deals
with vacation (after-) effects on H&W. In addition, we applied four on-vacation measurements
and zoomed in on the development of H&W levels during vacation. Moreover, we again inquired
into the role of vacation activities and experiences in H&W changes during and after vacation.
We further extended the study by including sleep time, sleep quality and the capacity to savor
positive experiences as possible determinants of the vacation (after-) effects.
We bring this dissertation to an end with a general discussion of our research
findings. Combining the findings from our meta-analysis and four studies on three different
types of vacations, also varying in length, enables us to shed light on the important question
whether vacation length matters for the strength and persistence of the vacation (after-) effects.
In addition, we will focus on the question whether vacation activities and experiences have
similar effects on changes in H&W during and after different types of vacations. We conclude
the discussion with suggestions for future work and practical implications.
1.7. References
Akerstedt, T. (2006). Psychosocial stress and impaired sleep. Scandinavian Journal of Work, Environment
& Health, 32, 493-501.
Akerstedt, T., Kecklund, G., Alfredsson, L., & Selen, J. (2007). Predicting long term sickness absence
from sleep and fatigue. European Sleep Research Society, 16, 341-345.
Aron, C.S. (1999). Working at play. A history of vacations in the United States. New York: Oxford University
Press.
Aronsson, G., Svensson, L., Gustafsson, K. (2003). Unwinding, recuperation, and health among
compulsory school and high school teachers in Sweden. International Journal of Stress
Management, 3, 217-234.
General Introduction
27
1
Belkic, K., Landbergis, P.A, Schnall, P.L., & Baker, D. (2004). Is job strain a major source of cardiovascular
disease risk? Scandinavian Journal of Work, Environment & Health, 30, 85-128.
Brosschot, J.F., Van Dijk, E., & Thayer, J.F. (2007). Daily worry is related to low heart rate variability during
waking and the subsequent nocturnal sleep period. International Journal of Psychophysiology,
63, 39-47.
Casson, L. (1994). Travel in the ancient world. Baltimore/London: The Johns Hopkins University Press.
Census and Statistics Department of the Government of the Hong Kong Special Administrative Region.
(2001). Benefits of employees under the employment ordinance. Retrieved June 21, 2011, from
http://www.statistics.gov.hk/publication/stat_report/social_data/B11301322002XXXXB0302.
pdf.
EC of the European Parliament and of the Council. (1993). Directive 93/104/EC of the European Parliament
and of the Council of 23 November 1993 concerning certain aspects of the organisation of
working time. Retrieved June 21, 2011, from http://eurlex.europa.eu/smartapi/cgi/sga_doc?
smartapi!celexapi!prod!CELEXnumdoc&lg=EN&numdoc=31993L0104&model=guichett.
Eden, D. (2001). Vacations and other respites: Studying stress on and off the job. In C. Cooper & I. T.
Robertson (Eds.), Well-being in organizations (pp. 305-330). West Sussex (UK): John Wiley
& Sons, Ltd.
Esch, T., & Stefano G. B. (2004). The neurobiology of pleasure, reward processes, addic tion and their
health implications. Neuroendocrinology Letters, 35, 235-251.
European Foundation for the improvement of living and working conditions. (2008). Working time
developments 2008. Retrieved June 21, 2011, from http://www.eurofound.europa.eu/eiro/
studies/tn0903039s/tn0903039s.htm#hd5.
Ezoe, S., & Morimoto, K. (1994). Behavioral lifestyle and mental health status of Japanese factory
workers. Preventive Medicine, 28, 98-105.
Fredrickson, B.L. (2001). The role of positive emotions in positive psychology: The Broaden- and -Build
Theory of positive emotions. American Psychologist, 56, 218-226.
Fredrickson, B. L., Mancuso, R. A., Branigan, C., & Tugade, M. M. (2000). The undoing effect of positive
emotions. Motivation and Emotion, 24, 237 - 258.
Fritz, C., & Sonnentag, S. (2005). Recovery, health, and job performance: Effects of weekend experiences.
Journal of Occupational Health Psychology, 10, 187-99.
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 & Health, 32, 482-492.
Geurts, S.A.E., Kompier, M.A.J., & Gründemann, R.G.W. (2000). Curing the Dutch disease? Sickness
absence and disability in the Netherlands. International Social Security Review, 53, 79-103.
Gräf, H.T., & Pröve, R. (1997). Wege ins Ungewisse. Reisen in der frühen Neuzeit 1500-1800. Frankfurt
a. M.: S. Fischer.
Griep, W., & Jäger, H.W. (1986). Reisen im 18. Jahrhundert. Neue Untersuchungen. Heidelberg: C. Winter
Universitätsverlag.
28
Chapter 1
1
Gump, B.B., & Matthews, K.A. (2000). Are vacations good for your health? The 9-year mortality experience
after the multiple risk factor intervention trial. Psychosomatic Medicine, 62, 608-612.
Hachtmann, R. (2007). Tourismus-Geschichte. Göttingen: Vandenhoek & Ruprecht.
Hjortskov, N., Rissen, D., Blangsted, A.K., Fallentin, N., Lundberg, U., & Sogaard, K. (2004). The effect
of mental stress on heart rate variability and blood pressure during computer work. European
Journal of Applied Physiology, 92, 84-89.
Hobfoll, S.E. (1989). Conservation of resources: a new attempt at conceptualizing stress. American
Psychologist, 44, 513-524.
How long should a man’s vacation be? (1910, July 31). The New York Times Magazine.
Keitz, C. (1997). Reisen als Leitbild. Die Entstehung des modernen Massentourismus in Deutschland.
München: Dtv.
Marks, S.R. (1977). Multiple roles and role strain: some notes on human energy, time and commitment.
American Sociological Review, 42, 921-936.
Maslach, C., Schaufeli, W.B., & Leiter, M.P. (2001) Job burnout. Annual Research of Psychology, 52,
397-422.
McEwen, B.S. (1998). Stress, adaptation, and disease: allostasis and allostatic load. Annals New York
Academy of Science, 840, 33-44.
Meijman, T. F., & Mulder, G. (1998). Psychological aspects of workload. In P. J. D. Drenth, H. Thierry &
C. J. de Wolff (Eds.), Handbook of work and organizational psychology (2nd edition). Work
psychology (Vol. 2, pp. 5-33). Hove: Psychology Press.
Mommersteeg, P.M.C. (2006). The psychophysiology of burnout. Enschede: Gildeprint Drukkerijen B.V.
NBTC NIPO. (2009). Kerncijfers over vakanties van Nederlanders 2009. Retrieved June 2011, from http://
www.nbtcniporesearch.nl/cvo/resultaten/presentaties/.
Ohler, N. (1986). Reisen im Mittelalter. München: Artemis.
Ray, R., & Schmitt, R. (2007). No vacation nation. Report from the Center for Economic and Policy
Research. Retrieved July 1, 2011, from http://www.cepr.net/index.php/publications/reports/
no-vacation-nation/.
Reis, H.T., Sheldon, K.M., Gable, S.L., Roscoe, J., & Ryan, R.M. (2000). Daily well-being: the role of
autonomy, competence, and relatedness. Personality and Social Psychology Bulletin, 26, 419-
435.
Reulecke, J. (1976). Vom blauen Montag zum Arbeiterurlaub. Vorgeschichte und Entstehung des
Erholungsurlaubs für Arbeiter vor dem Ersten Weltkrieg. In: Archiv für Sozialgeschichte,
Bd.XVI/1976, 205-248.
Rijksoverheid. (2011). Vakantiedagen en vakantiegeld. Retrieved July 14
th
2011, from http://www.
rijksoverheid.nl/onderwerpen/werktijden/vakantiedagen-en-vakantiegeld.
Ryan, R.M., & Deci, E.L. (2000). Self-Determination Theory and the facilitation of intrinsic motivation,
social development, and well-being. American Psychologist, 55, 68-78.
General Introduction
29
1
Ryan, R.M., Bernstein, J.H., & Brown, K.W. (2010). Weekends, work, and well-being: psychological need
satisfactions and day of the week effects on mood, vitality, and physical symptoms. Journal of
Social and Clinical Psychology, 29, 95-122.
Schnall, P.L., Schwartz, J.E., Landsbergis, P.A., Warren, K., Pickering, T.G. (1998). A longitudinal study of
job strain and ambulatory blood pressure: results from a three-year follow-up. Psychosomatic
Medicine, 60, 697-706.
Sheldon, K.M., Ryan, R., & Reis, H.T. (1996). What makes a good day? Competence and autonomy in
the day and in the person. Personality and Social Psychology Bulletin, 22, 1270-1279.
Sluiter, J.K., Frings-Dresen, M.H.W., Meijman, T.F.(2000). Reactivity and recovery from different types of
work measured by catecholamines and cortisol. A systematic literature overview. Occupational
Environmental Medicine, 57, 298-315.
Sonnentag, S. (2001). Work, recovery activities, and individual well-being: A diary study. Journal of
Occupational Health Psychology, 6, 196-210.
Sonnentag, S., & Geurts, S.A.E. (2009). Methodological issues in recovery research. In Sonnentag, S.,
Perrewe, P.L., & Ganster, D.C. (Eds.). Current perspectives on job-stress recovery (pp 1 - 36).
Bingley: Esmeralda Group Publishing Limited.
The Japan Institute for Labour Policy and Training. (2002). Research report No. 152, Research on annual
paid holidays. Retrieved June 21, 2011, from http://www.jil.go.jp/english/reports/documents/
jil-research/no152.pdf.
Tugade, M.M., & Fredrickson, B.L. (2007). Regulation of positive emotions: emotion regulation strategies
that promote resilience. Journal of Happiness Studies, 8, 311-333.
United States Department of Labour. (2010). Employee Benefits Survey 2010. Retrieved June 21, 2011,
from http://www.bls.gov/ncs/ebs/benefits/2010/ownership/civilian/table22a.pdf.
Vacation. (2011). Merriam-Webster.com. Britannica Company. Retrieved October 31, 2011, from http://
www.merriam-webster.com/dictionary/vacation.
Van Hooff, M.L.M., Geurts, S.A.E., Kompier, M.A.J., & Taris, T.W. (2007). Workdays, in-between workdays,
and the weekend: A diary study on effort and recovery. International Archives of Occupational
and Environmental Health, 80, 599-613.
Vrijkotte, T.G.M., van Doornen, L.J.P., de Geus, E.J.C. (2000). Effects of work stress on ambulatory blood
pressure, heart rate and heart rate variability. Hypertension, 35, 880-886.
World Tourism Organization. (1995). UNWTO technical manual: collection of tourism expenditure
statistics. Retrieved October 31, 2011, from http://pub.unwto.org/WebRoot/Store/Shops/
Infoshop/Products/1034/1034-1.pdf.
31
Chapter 2
Do We Recover from Vacation?
Meta-Analysis of Vacation Effects on Health
and Well-being
The aim of this meta-analysis is to investigate to what extent vacation has positive effects
on H&W, how long such effects endure after work resumption, and how specific vacation
activities and experiences affect these relationships.
Based on a systematic literature search (PsycInfo, Medline) and methodological exclusion
criteria, in a stepwise approach, 7 studies were selected and reviewed. Effect sizes
(Cohen’s d) were calculated i) for every outcome variable within every study, ii) for every
study by averaging the effect sizes per study, and iii) for homogeneous categories of
outcome variables (exhaustion, health complaints, life satisfaction).
The results suggest that vacation has positive effects on H&W (small effect, d = + 0.30),
but that these effects soon fade out after work resumption (small effect, d = - 0.27). Our
research further demonstrated that vacation activities and experiences have hardly been
studied. Therefore, their contribution to vacation effect and fade-out remains unclear.
Progresses in future vacation research will depend on strong research designs that
incorporate repeated measurements before, during and after vacation.
This chapter is based on:
De Bloom, J., Kompier, M., Geurts, S., De Weerth, C., Taris, T., & Sonnentag, S. (2009). Do we
recover from vacation? Meta-analysis of vacation effects on health and well-being. Journal of
Occupational Health, 51, 13-25.
Meta-analysis of Vacation Effects on Health and Well-being
33
2
2.1. Introduction
Time off is crucial for workers to recover from load effects built up at work. A core assumption
of Effort-Recovery Theory (Geurts & Sonnentag, 2006; Meijman & Mulder, 1998) and Allostatic
Load Theory (Clow, 2001; McEwen, 1998; Sterling & Eyer, 1990) is that initial normal load
reactions (e.g. accelerated heart rate and fatigue) can develop into more chronic load reactions
(e.g. prolonged fatigue, sleep complaints, high blood pressure) in cases of continued exposure
to workload and incomplete recovery during time after work (Geurts & Sonnentag, 2006). The
essence of recovery is a process of psycho-physiological unwinding after working, opposite
to the activation of the sympathetic-adrenal-medullary system and the hypothalamic-pituitary-
adrenal system during effort expenditure (work), particularly under demanding or stressful
conditions (Geurts & Sonnentag, 2006). Earlier research addressing rest breaks (Tucker,
2003), long work hours (Beckers et al., 2004; Härma 2006; Van der Hulst, 2003), and shift work
(Totterdell, Spelten, Smith, Barton & Folkard, 1995) has acknowledged the role of recovery from
work in preserving individual well-being, health and performance. Furthermore, over the years
labour unions have emphasized the importance of sufficient recovery time in their endeavours
for a shorter working week, rest breaks and vacation rights, and both national and international
working time legislations have been enacted to enable recovery possibilities for employees.
Recent studies have revealed that workers often recover insufficiently during time
off work due to, for instance, working overtime. This day-to-day incomplete recovery may
have serious adverse health consequences in the long run (Geurts & Sonnentag, 2006, for
an overview). Sluiter, Frings-Dresen and Meijman (2000) distinguished 4 different types of
recovery based on duration and time span after work: microrecovery (first minutes after task
performance), mesorecovery (10 minutes to 1 hour after task performance), metarecovery (1
hour to 2 days after work) and macrorecovery (more than 2 days after work).
Vacation as a form of macrorecovery is a prime candidate for helping workers
to recover more completely from work. Vacation is likely to be a more powerful recovery
opportunity than regular free evenings and weekends because of two mechanisms underlying
the recovery process. The first ‘passive’ mechanism reflects a direct release from daily job
demands: vacation is ideally a relatively long period of rest that is mostly spent in a different
and more relaxing environment that may help workers to detach psychologically from work
and from other daily demands and routines. The second ‘active’ mechanism reflects the active
engagement in potentially recovering activities: vacation is a pre-eminent opportunity to spend
time on valued non-work activities of one’s own choice, such as hobbies and family activities.
This article reviews the empirical literature with regard to the recovering impact of this
prototypical recovery possibility, i.e. a vacation from work. The term ‘vacation’ stems from the
34
Chapter 2
2
Latin word ‘vacatio’: ‘being free from work, being at leisure, having time for’. We hypothesize
that vacation, as a relatively long and uninterrupted period of respite from work may be a major
contributor to the recovery process, and therefore may be beneficial for H&W. Following a
vacation, employees return to work, and we are also interested in how long potential vacation
effects last, assuming that due to renewed exposure to work demands vacation effects will be
temporary and thus ‘fade out’.
From a work psychological point of view it is important not to treat a vacation as a
black box, but rather to find out whether vacation activities (e.g. sports or exercise) and vacation
experiences (e.g. vacation satisfaction) play a role in the relationship between vacation and
well-being.
In sum, this meta-analysis aims to answer 3 related research questions:
1) What empirical evidence exists for an improvement of H&W due to a vacation from work
(vacation effect)?
2) In the case of a positive effect of vacation, how long does it last (fade-out)?
3) a. Do vacation activities play a role in these potential relationships?
b. Do vacation experiences play a role in these potential relationships?
2.2. Method
A systematic literature search was carried out in 2 bibliographical databases: PsycInfo and
Medline. No publication year limits were set and the final search date was June 15
th
, 2008.
We used the following search terms within the fields ‘title’ or ‘keywords’:
1) vacation OR holiday (1702 hits), and
2) well-being OR health OR quality of life OR satisfaction OR stress OR burnout OR recovery
OR sleep OR mood OR affect (829,536 hits)
The combination of these 2 searches resulted in 125 hits (see Figure 2.1). In a first selection
round, the following exclusion criteria were used:
Language: non-English papers (minus 22)
Publication type: disserta tions, short communications, letters, non-empirical and/or
non-peer-reviewed papers (minus 38)
After application of these criteria 65 hits remained. All 65 abstracts were retrieved and read by
the first 3 authors. Exclusion criteria in this second round were:
Sample: papers not dealing with healthy, working sample (e.g. school
children, psychiatric patients): minus 14
Meta-analysis of Vacation Effects on Health and Well-being
35
2
Research purpose: studies irrelevant for the research questions, i.e. studies not
investigating vacation effects and/or fade-out on health and/or well-
being (e.g. weight gain during vacation, holiday shopping intentions,
sleepiness in drivers during summer vacation): minus 35
Design: studies not using an interrupted time series design with at least a
pre-test, i.e. before vacation and a post-test, i.e. after vacation per
subject as such studies do not permit the evaluation of a vacation
(e.g. only post-vacation measure during annual doctor visit): minus
5
Figure 2.1:
Systematic literature search on vacation and health or well-being from 125 to final 7 studies
Search terms in title or keywords:
vacation OR holiday
(1702 hits)
Search terms in title or keywords:
well-being OR health OR quality
of life OR satisfaction OR stress
OR burnout OR recovery OR
sleep OR mood OR affect
( 829536 hits)
125 hits
103 hits
65 hits
51 hits
16 hits
11hits
Final 7 hits
Language:
22 non-English papers
Sample:
14 papers not dealing with
healthy, working respondents
Research purpose:
35 studies irrelevant for research
questions
Publication type:
38 papers not published in peer-
reviewed journals
Detailed reading:
2 papers written by the same
authors, based on same sample
1 paper not comparing outcomes
within the same persons
1 paper with design not fitting
research purpose
AND
Design:
5 studies not using interrupted
time series design
36
Chapter 2
2
Based on these criteria, 54 articles were excluded, and 11 papers remained. Studies that
were referred to in the 11 selected papers were also examined but no additional, relevant
papers were detected. The first 3 authors of the present article studied the remaining 11 papers
and excluded 4 more papers. In 2 cases, papers were written by the same authors (Strauss-
Blasche, Ekmekcioglu & Marktl, 2000; Strauss-Blasche, Ekmekcioglu & Marktl, 2002; and
Hoopes & Lounsbury, 1989; Lounsbury & Hoopes, 1986), based on the same sample with the
second paper not offering extra information for our research purposes. Therefore the second
paper was excluded in both cases (Hoopes & Lounsbury, 1989; Strauss-Blasche et al., 2002).
A third paper was excluded (Etzion & Westman, 2001) because it investigated cross-over and
thus compared outcome variables before and after vacation in spouses instead of vacationers
themselves. Finally, a fourth paper had to be excluded (Eden, 1990) as it did not fit our research
purposes: the ‘vacation’ in this study was a compulsory off work period, ‘a brief interlude during
an acutely stressful computer crisis’ (Westman & Eden, 1997, p. 524). This resulted in a final
selection of 7 studies (see Table 2.1).
To mathematically quantify the empirical evidence for vacation effects in the 7 different
studies we calculated the effect size d for paired observations as described in Cohen (1988,
p. 46). Firstly, we calculated, within every study, effect sizes for all outcome variables in that
study. Secondly, we calculated average effect sizes for all studies by averaging all effect sizes
within each study.
Thirdly, in order to obtain a more detailed picture for specific homogeneous outcome
categories, we computed a mean d for those outcome variables that were used in 3 or more
different studies. This fine-grained analysis was performed for the following outcome categories:
exhaustion (4 studies), health complaints (3 studies), and life satisfaction (3 studies).
Following Cohen (1988) we distinguished small (0 to 0.5), medium (0.5 to 0.8)
and large effect sizes (> 0.8). Positive effect sizes indicate a beneficial effect of vacation
(improvement of H&W), whereas negative effects denote the opposite (decrease in health after
vacation compared to pre-vacation levels).
2.3. Results
Table 2.1 provides an overview of the 7 studies, by characterizing sample and design
characteristics, pre-vacation measurement(s), measurement(s) during vacation, post-vacation
measurement 1, and post-vacation measurement 2.
Meta-analysis of Vacation Effects on Health and Well-being
37
2
2.3.1. Sample and design characteristics
Number of participants: Sample sizes of the reviewed studies were mostly small.
Attrition from the pre-vacation to the first post-vacation measurement varied between 5%
(Etzion, 2003) and 59% (Strauss-Blasche et al., 2000). The loss of participants from the pre-
vacation to the post-vacation 2 measure varied between 5% (Etzion, 2003) and 86% (Strauss-
Blasche et al., 2000).
Sex, age and occupation: The distributions of sex, age and occupation were diverse
in the reviewed studies.
Control group: 5 studies (Fritz & Sonnentag, 2006; Lounsbury & Hoopes, 1986;
Strauss-Blasche et al., 2000; Westman & Eden, 1997; Westman & Etzion, 2001) did not include
a control group. Gilbert and Abdullah (2004) incorporated a non-vacationing control group
of 249 respondents (opposed to 355 holiday-takers) that reported a lower well-being than
the holiday-takers on pre-vacation. Etzion (2003) used a matched-pairs technique to create a
comparable control group of 55 respondents (age, marital status and job function). This control
group’s pre-vacation scores on exhaustion resembled the vacation group’s scores.
Duration: 3 of the 7 studies (Etzion, 2003; Fritz & Sonnentag, 2006; Gilbert & Abdullah,
2004) did not report the duration of the vacation of their respondents. The average duration of
the vacation in the other studies was 9 (Lounsbury & Hoopes, 1986), 10 (Westman & Etzion,
2001), and 14 days (Strauss-Blasche et al., 2000; Westman & Eden, 1997).
Timing: In 2 studies, the timing of the vacation was not reported (Fritz & Sonnentag,
2006; Gilbert & Abdullah, 2004). In 1 case, vacation time of the respondents was in spring
(Westman & Etzion, 2001). In the remaining 4 studies (Etzion, 2003; Lounsbury & Hoopes,
1986; Strauss-Blasche et al., 2000; Westman & Eden, 1997), participants went on vacation in
summertime.
Location: In 5 of the 7 studies (Etzion, 2003; Fritz & Sonnentag, 2006; Gilbert &
Abdullah, 2004; Lounsbury & Hoopes, 1986; Westman & Etzion, 2001), vacation location of the
respondents was not reported. In 2 studies, more than 75 percent of the participants stayed at
home during their vacation (76% in Strauss-Blasche et al., 2000 and 87% in Westman & Eden,
1997).
2.3.2. Pre-vacation measure
Timing of measurement: Two studies (Etzion, 2003; Gilbert & Abdullah, 2004) did not
report when the pre-vacation measure took place. Of the remaining 5 studies, Westman and
Eden (1997) was the only study that collected measures at 2 pre-vacation time points (6 weeks,
and 3 days before vacation; they found no evidence for major differences between these 2 time
points). In the study of Lounsbury and Hoopes (1986) data were collected 1 to 14 days (7 days
38
Chapter 2
2
Table 2.1:
Design characteristics of the 7 reviewed studies
Author(s), year of
publication
Sample and design
characteristics
Pre-vacation
measurement(s)
During vacation
measurement
Post-vacation
measurement 1
Post-vacation
measurement 2
Lounsbury & Hoopes,
1986
N pre-vac: 168
N post-vac 1: 128
N post-vac 2: -
Country: USA
: 28%
Age: 39 years
Occ: variety
Contr.group: No
Vacation features
Duration: 9 days on
average
Timing: summer
vacation
Location: NR
1-14 days before
vacation, median 7 days
Outcome variables
Life satisfaction
Job involvement/ central
life interest
Job involvement/valued
self
Organizational
commitment
Turnover intention
Job satisfaction
- 1st week after vacation,
median 3 days
Outcome variables
Same as pre-vacation
Activities:
Activities during
vacation
Experiences:
Overall vacation
satisfaction
Satisfaction with aspects
of vacation
-
Westman & Eden, 1997 N pre-vac 1: 88
N pre-vac 2: 76
N post-vac 1: 76
N post-vac 2: 76
Country: Israel
: 59%
Age: NR
Occ: Administrative
clerks
Contr.group: No
Vacation features
Duration: 14 days
Timing: summer
vacation
Location: 87 % at home,
13 % away from home
6 weeks before vacation
3 days before vacation
Outcome variables
Exhaustion (physical,
emotional, mental)
2
nd
week of vacation
Outcome variables
Same as pre-vacation
Activities:
-
Experiences:
-
3 days after vacation
Outcome variables
Same as pre-vacation
Activities:
-
Experiences:
Vacation satisfaction
3 weeks after vacation
(18 days after post-
vacation 1)
Outcome variables
Same as pre-vacation
Meta-analysis of Vacation Effects on Health and Well-being
39
2
Author(s), year of
publication
Sample and design
characteristics
Pre-vacation
measurement(s)
During vacation
measurement
Post-vacation
measurement 1
Post-vacation
measurement 2
Strauss-Blasche,
Ekmekcioglu & Marktl,
2000
N pre-vac: 130
N post-vac 1: 53
N post-vac 2: 18
Country: Austria
: 70%
Age: 34 years
Occ: 57% manual
workers, 43% white
collar workers
Contr.group: No
Vacation features
Duration: 14 days
Timing: summer
vacation
Location: 76 % at home,
24 % at holiday resort
10 days before vacation
Outcome variables
Life satisfaction
Physical complaints
Quality of sleep
Positive mood
Negative mood
- 3 days after vacation
Outcome variables
Same as pre-vacation
Activities:
Activities during
vacation
Experiences:
Recuperation
5 weeks after vacation
(32 days after post-
vacation 1)
Outcome variables
Same as pre-vacation
Westman & Etzion, 2001 N pre-vac: 126
N post-vac 1: 87
N post-vac 2: 87
Country: Israel
: 61%
Age: 41 years
Occ: blue collar
industrial workers
Contr.group: No
Vacation features
Duration: 10 days
Timing: Passover
vacation (spring)
Location: NR
10 days before vacation
Outcome variables
Exhaustion (physical,
emotional, mental)
Absenteeism for health
reasons
Absenteeism for other
reasons
(Company records, on
aggregate level, not
individually)
- 3 days after vacation
Outcome variables
Same as pre-vacation
Activities:
-
Experiences:
-
4 weeks after vacation
(25 days after post-
vacation 1)
Outcome variables
Same as pre-vacation
Table 2.1 (continued)
40
Chapter 2
2
Author(s), year of
publication
Sample and design
characteristics
Pre-vacation
measurement(s)
During vacation
measurement
Post-vacation
measurement 1
Post-vacation
measurement 2
Etzion, 2003 N pre-vac: 58
N post-vac 1: 55
N post-vac 2: 55
N control: 55
Country: Israel
: 49%
control: 49%
Age: 45 years
Age control: 43 years
Occ: employees at
industrial enterprise
Contr.group: Yes
Vacation features
Duration: NR, at least
1 week
Timing: summer vacation
Location: NR
NR
(“(…) before the
individual (…) went on
vacation”)
Outcome variables
Exhaustion (physical,
emotional, mental)
- NR
(“immediately after he/
she returned to work”)
Outcome variables
Same as pre-vacation
Activities:
-
Experiences:
Vacation satisfaction
Detachment from
workplace
3 weeks after vacation
Outcome variables
Same as pre-vacation
Gilbert & Abdullah, 2004 N pre-vac: NR
N post-vac 1: 355
N post-vac 2: -
N control: 249
Country: United
Kingdom
: 50%
control: 50%
Age: 16-24:14%; 25-34:22%;
35-44:17%, 45-54:21%; 55-
64:14%; 65-x:13%
Age control: 16-24:16%;
25-34:18%; 35-44:16%;
45-54:15%; 55-64:11%;
65-x:23%
Occ: variety
Occ control: variety
Contr.group: Yes
Vacation features
Duration: NR, at least 4
nights
Timing: whole year
round
Location: NR
NR
(“…3541 questionnaires
were distributed at 2
points in time, during a
12-month period”)
Outcome variables
Life satisfaction
Satisfaction with…
Friends
Family
Home
Relationships
Economic situation
Leisure
Neighborhood
Self
Services
Health
Nation
Positive affect
Negative affect
Current affect
Job satisfaction
- 2-6 months after first
questionnaire
(“(…) within a period
of 2-6 months after
completion of the first
questionnaire”)
Outcome variables
Same as pre-vacation
Activities:
-
Experiences:
-
-
Table 2.1 (continued)
Meta-analysis of Vacation Effects on Health and Well-being
41
2
Author(s), year of
publication
Sample and design
characteristics
Pre-vacation
measurement(s)
During vacation
measurement
Post-vacation
measurement 1
Post-vacation
measurement 2
Fritz & Sonnentag, 2006 N pre-vac: 414
N post-vac 1: 221
N post-vac 2: 221
Country: Germany
: 15%
Age: 46 years
Occ: Non-academic
University employees
Contr.group: No
Vacation features
Duration: NR, at least
1 week
Timing: NR
Location: NR
7 days before vacation
Outcome variables
Exhaustion
Disengagement
Health complaints
Work effort
Task performance
NR
(“The survey booklet
had to be filled in (…)
during vacation”)
Activities:
-
Experiences:
Relaxation experience
Mastery experience
Negative work reflection
Positive work reflection
Non-work hassles
(e.g. conflicts, financial
problems)
1-2 days after vacation
Outcome variables
Same as pre-vacation
2 weeks after vacation
(12-13 days after post-
vacation 1)
Outcome variables
Same as pre-vacation
Studies presented in order of publication date, NR = Not reported, - = Not measured, Occ = Occupation of participants, Contr.group = Control group
Table 2.1 (continued)
42
Chapter 2
2
on average) prior to vacation, and the 3 remaining studies fell into the same time range: 10
(Strauss-Blasche et al., 2000; Westman & Etzion, 2001), and 7 days (Fritz & Sonnentag, 2006).
Outcome variables: All studies measured the same H&W parameters before and
after vacation but the type of variables used varied: 4 studies measured exhaustion (Etzion,
2003; Fritz & Sonnentag, 2006; Westman & Eden, 1997; Westman & Etzion, 2001), 3 measured
health complaints (Fritz & Sonnentag, 2006; Gilbert & Abdullah, 2004; Strauss-Blasche et
al., 2000) and 3 measured general life satisfaction (Lounsbury & Hoopes, 1986; Strauss-
Blasche et al., 2000; Gilbert & Abdullah, 2004). Job satisfaction was measured in 2 different
studies (Lounsbury & Hoopes, 1986; Gilbert & Abdullah, 2004), and several parameters were
measured in only one study: e.g. negative mood (Strauss-Blasche et al., 2000), turnover
intention (Lounsbury & Hoopes, 1986), and self-reported work effort (Fritz & Sonnentag, 2006).
2.3.3. During vacation measure
Timing of measurement: Only 2 papers (Fritz & Sonnentag, 2006; Westman & Eden,
1997) incorporated a during vacation measurement. Westman and Eden (1997) scheduled
their during vacation measure in the second week of the vacation. Fritz and Sonnentag (2006)
did not report when their vacation measurement exactly took place.
Activities and experiences during vacation: Only Fritz and Sonnentag (2006) asked
their respondents about their vacation experiences when actually on vacation. They investigated
experiences during vacation in a detailed way, by gathering information on relaxation and
mastery experiences, positive and negative work reflection and non-work hassles.
2.3.4. Post-vacation measure 1
Timing of measurement: Etzion (2003) did not report when the first post-vacation
measure took place. Gilbert and Abdullah (2004) took their only post-vacation measure 2 to
6 months after the pre-vacation measure. The remaining 5 studies (Fritz & Sonnentag, 2006;
Lounsbury & Hoopes, 1986; Strauss-Blasche et al., 2000; Westman & Eden, 1997; Westman
& Etzion, 2001) scheduled their first post-vacation measure within the first week of returning to
work (3 days on average).
Activities during vacation: At post-vacation 1, i.e. retrospectively, 2 studies collected
information about vacation activities (Lounsbury & Hoopes, 1986; Strauss-Blasche et al., 2000).
Experiences during vacation: 3 papers (Fritz & Sonnentag, 2006; Gilbert & Abdullah,
2004; Westman & Etzion, 2001) gathered no information about vacation experiences. Three
of the remaining studies (Etzion, 2003; Lounsbury & Hoopes, 1986; Westman & Eden, 1997)
asked respondents about their vacation satisfaction in retrospect. Etzion (2003) and Strauss-
Meta-analysis of Vacation Effects on Health and Well-being
43
2
Blasche et al. (2000) included questions about recuperation during vacation and detachment
from the workplace respectively.
2.3.5. Post-vacation measure 2
Timing of measurement: Five studies (Etzion, 2003; Fritz & Sonnentag, 2006; Strauss-
Blasche et al., 2000; Westman & Eden, 1997; Westman & Etzion, 2001) adopted a second
post-vacation measure. Post-vacation 2 measures were collected 2 weeks after vacation (12-
13 days after post vac 1) in Fritz and Sonnentag (2006), 3 weeks (18 days after post-vacation
1) in Westman and Eden (1997) and Etzion (2003), and 4 weeks (25 days after post-vacation
1) in Westman & Etzion (2001). Strauss-Blasche et al. (2000) had the longest time interval: 5
weeks after vacation (32 days after post-vacation 1).
2.3.6. Research question 1: Vacation effect?
We calculated the pre-vacation – post-vacation 1 difference in health and well-being indicators
(‘vacation effect’) in all 7 studies. The time span between these 2 time points was unknown in
3 studies: there was no data available on vacation duration (Fritz & Sonnentag, 2006) or pre-
vacation time point, vacation duration and post-vacation 1 time point (Etzion, 2003; Gilbert &
Abdullah, 2004). The time span between the pre- and post-vacation 1 time points in the other
4 studies ranged between 19 (Lounsbury & Hoopes, 1986), 20 (Westman & Eden, 1997), 23
(Westman & Etzion, 2001) and 27 days (Strauss-Blasche et al., 2000).
First we calculated, within every study, effect sizes for every outcome variable in that
study. Then, we calculated general effect sizes for every study, i.e. averaged the number of
effect sizes within each study (Table 2.2).
The minimum number of outcome variables per study was 1 (Etzion, 2003; Westman
& Eden, 1997; Westman & Etzion, 2001) and the maximum number was 17 (Gilbert & Abdullah,
2004). Within the 7 papers 36 outcome variables were studied, hence 36 effect sizes were
calculated. Thirty of these were positive (improvement in well-being) and 6 negative (decrease
in well-being). The 6 negative effect sizes were small (mean d = -0.13) and of the positive effect
sizes, 25 were small and 5 were medium. Medium effect sizes were found for health complaints
(d = +0.71 in Fritz & Sonnentag, 2006 and +0.57 in Strauss-Blasche et al., 2000), exhaustion
(d = +0.65 in Westman & Eden, 1997), work effort (d = 0.52 in Fritz & Sonnentag, 2006) and
satisfaction with nation (d = 0.52 in Gilbert & Abdullah, 2004). The average effect sizes per
study varied from -0.04 (Lounsbury & Hoopes, 1986) to +0.65 (Westman & Eden, 1997). The
overall mean d across 7 studies was +0.30, indicating a small positive vacation effect.
In the 2 control group studies, Etzion (2003) found a small “pre-post vacation 1”
increase in exhaustion in the control group (d = -0.08). The “post-vacation 1” difference in
44
Chapter 2
2
Table 2.2:
Means, standard deviations and effect sizes for vacation effect on all outcome variables for each study
Study Outcome variable Mean
pre-vac
SD
pre-vac
Mean
post-
vac 1
SD
post-
vac 1
Cohen
d
Mean
d
Lounsbury &
Hoopes, 1986
Life satisfaction 24.87 5.68 23.83 6.36 + 0.28
- 0.04
Job involvement/
interest
21.17 4.88 22.10 4.61 - 0.34
Job involvement/
valued self
11.43 3.18 11.68 3.21 + 0.11
Organizational
commitment
10.51 2.56 10.65 2.69 + 0.07
Turnover intention 3.80 0.98 3.67 0.99 - 0.19
Job satisfaction 22.08 5.71 21.51 5.69 - 0.16
Westman &
Eden, 1997
Exhaustion 3.30 0.60 3.03 0.62 + 0.65
+ 0.65
Strauss-
Blasche,
Ekmekcioglu
& Marktl, 2000
(12)
Life satisfaction NR NR NR NR + 0.02
+ 0.37
Physical
complaints
NR NR NR NR + 0.57
Quality of sleep NR NR NR NR + 0.32
Positive mood NR NR NR NR + 0.46
Negative mood NR NR NR NR + 0.47
Westman &
Etzion, 2001
Exhaustion 2.89 0.65 2.70 0.99 + 0.25
+ 0.25
Etzion, 2003 Exhaustion 2.59 0.54 2.44 0.59 + 0.33 + 0.33
Gilbert &
Abdullah,
2004
Life satisfaction (1
item)
6.99 1.23 7.11 1.20 + 0.16
+ 0.23
Life satisfaction
(scale)
30.78 7.12 31.78 7.59 + 0.22
Positive affect 60.47 12.24 63.58 11.79 + 0.32
Negative affect 31.22 14.28 30.21 14.44 + 0.09
Current affect 29.30 22.77 33.29 23.13 + 0.21
Satisfaction friends 7.25 1.15 7.24 1.07 - 0.00
Satisfaction family 7.22 1.50 7.20 1.42 - 0.02
Satisfaction home 6.85 1.31 6.94 1.26 + 0.11
Satisfaction
relationships
6.80 1.06 7.02 1.02 + 0.35
Satisfaction econ.
situation
6.75 1.43 6.97 1.22 + 0.26
Satisfaction leisure 6.34 1.45 6.53 1.22 + 0.23
Satisfaction
neighborhood
6.29 1.36 6.49 1.30 + 0.24
Satisfaction self 6.22 1.22 6.55 1.20 + 0.44
Satisfaction
services
6.12 1.23 6.39 1.11 + 0.38
Satisfaction health 5.97 1.42 6.22 1.44 + 0.15
Satisfaction nation 4.75 1.19 5.15 1.32 + 0.52
Job satisfaction 6.42 1.29 6.67 1.19 + 0.33
Meta-analysis of Vacation Effects on Health and Well-being
45
2
exhaustion between vacationers and non-vacationers was small (d= +0.35), with non-
vacationers reporting more exhaustion. Gilbert and Abdullah (2004) found negative changes
for the control group on all outcome variables, indicative of deterioration in well-being (mean
d = -0.19). The difference between non-holiday and holiday takers at “post-vacation 1” was
small (mean d = +0.50), the former reported a lower well-being.
Next, a fine-grained analysis for the homogenous outcome categories exhaustion, life
satisfaction, and health complaints was conducted (Table 2.3). Effect sizes for the category
exhaustion (4 studies) varied from +0.25 (Westman & Etzion, 2001) to +0.65 (Westman &
Eden, 1997). The average d was +0.39, indicating a small vacation effect.
Concerning health complaints, effect sizes were +0.71 (Fritz & Sonnentag, 2006),
+0.15 (Gilbert & Abdullah, 2004) and +0.57 (Strauss-Blasche et al., 2000). The average effect
size was +0.48, indicating a small effect.
Finally, a small average effect size (d = +0.16) was found for the category life
satisfaction. Cohen’s d ranged between +0.02 (Strauss-Blasche et al., 2000), +0.19 (Gilbert &
Abdullah, 2004) and +0.28 (Lounsbury & Hoopes, 1986).
2.3.7. Research question 2: Fade-out?
The concept of ‘fade-out’ supposes the a priori existence of an effect. Vacation effects can
only disappear when they were present in the first place, i.e. at post-vacation 1. Our analysis
was thus based upon those 4 studies that employed 2 post-vacation measures, and found
a positive vacation effect (Etzion, 2003; Fritz & Sonnentag, 2006; Westman & Eden, 1997;
Westman & Etzion, 2001). Note that Strauss-Blasche et al. (2000) included a post-vacation 2
measure too, but they neither compared their outcome variables at this time point with those
Study Outcome variable Mean
pre-vac
SD
pre-vac
Mean
post-
vac 1
SD
post-
vac 1
Cohen
d
Mean
d
Fritz &
Sonnentag,
2006
Health complaints 1.94 0.47 1.59 0.35 + 0.71
+ 0.32
Exhaustion 2.18 0.55 2.05 0.55 + 0.32
Disengagement 2.10 0.53 2.06 0.53 + 0.10
Task performance 4.51 0.49 4.49 0.54 - 0.04
Work effort 2.90 1.14 2.26 1.15 + 0.52
Total + 0.30
Abbreviations: + = positive effect, improvement in health and/or well-being, - = negative effect, decrease
in health and/or well-being, Mean pre-vac = mean at pre-vacation, SD pre-vac = standard deviation at
pre-vacation, Mean post-vac 1 = mean at post-vacation 1, SD post-vac 1 = standard deviation at post-
vacation1, NR = Not reported in study
Table 2.2 (continued)
46
Chapter 2
2
at post-vacation 1, nor reported means and standard deviations at the different measurement
occasions.
In 4 studies that compared post-vacation 1 and 2 (Etzion, 2003; Fritz & Sonnentag,
2006; Westman & Eden, 1997; Westman & Etzion, 2001) effect sizes could be calculated for
exhaustion. In addition, in the study of Fritz and Sonnentag (2006) also 3 other effect sizes
could be calculated. Single outcome effect sizes per study were -0.02 (Etzion 2003), -0.08 (Fritz
& Sonnentag, 2006), -0.20 (Westman & Etzion, 2001) and -0.76 (Westman & Eden, 1997). In
the study of Fritz and Sonnentag (2006) effect sizes ranged from +0.06 to -0.34.
From the total of 7 different outcome variables, 1 had a positive sign, 1 was 0, and 5
had a negative sign meaning that in most cases well-being decreased between post-vacation
1 and 2. The only positive effect size was negligibly small (d = +0.06). Within the 5 negative
effect sizes, 4 were small and 1 medium. This medium effect size was found for exhaustion (d
= -0.76 in Westman & Eden, 1997).
The overall mean d across 4 studies was -0.27, indicating a small fade-out effect.
Table 2.4 further shows the time span between the 2 post-vacation measures that varied
between approximately 2 to 4 weeks. As there were only 2 post-vacation measures in all 4
Table 2.3:
Effect sizes for vacation effect in homogeneous outcome variables used in 3 or more different studies
Outcome variables Study Cohen
d
Mean Cohen d
corrected for
more than 1
indicator per
study
Mean
Cohen
d
Exhaustion (4 studies) + 0.39
Exhaustion Westman & Eden, 1997 + 0.65 + 0.65
+ 0.25
+ 0.33
+ 0.32
Exhaustion Westman & Etzion, 2001 + 0.25
Exhaustion Etzion, 2003 + 0.33
Exhaustion Fritz & Sonnentag, 2006 + 0.32
Health complaints (3 studies) + 0.48
Physical complaints Strauss-Blasche et al., 2000 + 0.57 + 0.57
+ 0.15
+ 0.71
Satisfaction with health Gilbert & Abdullah, 2004 + 0.15
Health complaints Fritz & Sonnentag, 2006 + 0.71
Life satisfaction (3 studies) + 0.16
Life satisfaction Lounsbury & Hoopes, 1986 + 0.28 + 0.28
+ 0.02
+ 0.19
Life satisfaction Strauss-Blasche et al., 2000 + 0.02
Life satisfaction Gilbert & Abdullah, 2004 + 0.16
+ 0.22
Life satisfaction
Abbreviations: + = positive effect, improvement in health and/or well-being, - = negative effect, decrease
in health and/or well-being
Meta-analysis of Vacation Effects on Health and Well-being
47
2
studies and the minimum fade-out interval was 12-13 days after vacation (Fritz & Sonnentag,
2006), it was impossible to study the specific course of fade-out and to determine when fade-
out began and when pre-vacation base levels were reached again.
Only Etzion (2003) compared scores on 2 measures in a non-vacation group taken at
the same time as post-vacation 1 and 2 in the vacation group. She found a small positive effect
(d = +0.16) meaning that exhaustion decreased in the control group in the time between the
second and the third measurement occasion. The difference between vacationers and non-
vacationers was +0.19 on “post-vacation 2”, meaning that non-vacationers were slightly more
exhausted than their vacation taking fellows.
Again, we performed a fine-grained analysis of homogeneous outcome variables,
measured in 3 or more different studies. Only exhaustion met this criterion (4 studies). The
average effect size was small (d = (-0.02) + (-0.76) + (-0.20) + 0.06)/4 = - 0.23).
Table 2.4:
Means, standard deviations and effect sizes for fade-out on all outcome variables for each study
Study Outcome
variable
Time span
post 1-post 2
Mean
post-
vac 1
SD
post-
vac 1
Mean
post-
vac 2
SD
post-
vac 2
Cohen
d
Mean
d
Westman &
Eden, 1997
Exhaustion 18 days 3.03 0.62 3.35 0.62 - 0.76
- 0.76
Westman
& Etzion,
2001
Exhaustion 25 days 2.70 0.99 2.92 0.94 - 0.20
- 0.20
Etzion,
2003
Exhaustion 21 days
(post-vacation
1 immediatly
after returning
to work)
2.44 0.59 2.45 0.66 - 0.02
- 0.02
Fritz &
Sonnentag,
2006
Health
complaints
12-13 days 1.59 0.35 1.71 0.42 - 0.34
- 0.08
Exhaustion 2.05 0.55 2.03 0.56 +0.06
Disengagement 2.06 0.53 2.06 0.54 0.00
Work effort 2.26 1.15 2.31 1.15 - 0.05
Total - 0.27
Abbreviations: time span post 1-post 2 = time span between post-vacation 1 and post-vacation 2, + =
positive effect, improvement in health and/or well-being, - = negative effect, decrease in health and/or well-
being, Mean post-vac 1 = mean at post-vacation 1, SD post-vac 1 = standard deviation at post-vacation 1,
Mean post-vac 2 = mean at post-vacation 2, SD post-vac 2 = standard deviation at post-vacation 2
48
Chapter 2
2
2.3.8. Research question 3a: Activities on vacation?
Only 2 of 7 studies collected data during vacation. However, neither study (Fritz & Sonnentag,
2006; Westman & Eden, 1997) collected information about what vacationers actually did during
their holiday. Two other studies (Lounsbury& Hoopes, 1986; Strauss-Blasche et al., 2000)
collected information on vacation activities in retrospect, i.e. at post-vacation 1. These studies
reported percentages that were spent on certain activities (e.g. traveling, reading, sightseeing)
but the authors did not relate these percentages to the outcome variables. This means that
research question 3a could not be addressed.
2.3.9. Research question 3b: Experiences on vacation?
One study (Fritz & Sonnentag, 2006) collected information on vacation experiences during the
vacation itself. Four other studies (Etzion 2003; Lounsbury & Hoopes, 1986; Strauss-Blasche
et al., 2000; Westman & Eden, 1997) collected information on vacation experiences at post-
vacation 1 when respondents had already resumed working.
Vacation satisfaction was measured in 3 studies (Etzion, 2003; Lounsbury & Hoopes,
1986; Westman & Eden, 1997) and appeared to be positively related to job satisfaction and
life satisfaction (Lounsbury & Hoopes, 1986) and negatively to exhaustion (Westman & Eden,
1997), whereas Etzion (2003) found no such relationship with exhaustion. Etzion (2003) also
retrospectively collected information on detachment from work during the vacation and did
not find a relationship with post-vacation exhaustion, whereas Strauss-Blasche et al. (2000)
found that well-being at post-vacation was higher among those respondents who reported
sufficient recuperation during vacation compared to those who indicated that recuperation
during vacation was insufficient.
In the only ‘during vacation study’ Fritz and Sonnentag (2006) tested the effect of
vacation experiences on health indicators after vacation. Positive (e.g. relaxation) as well
as negative experiences (e.g. negative work reflection) were related to almost all outcome
variables. Within these experiences, negative work reflection seemed to play a major role:
respondents engaging in negative work reflection during vacation reported also lower well-
being on post-vacation 1.
In sum, only 1 study (Fritz & Sonnentag, 2006) measured vacation experiences when
employees actually were on holidays. This study found evidence in support of a temporal
relation between vacation experiences and outcome variables: positive experiences were
related to improved well-being after vacation whereas negative experiences had the opposite
effect. Of the 4 studies that collected information on vacation experiences after returning to
work (mostly vacation satisfaction), 2 studies reported positive cross-sectional associations
Meta-analysis of Vacation Effects on Health and Well-being
49
2
between vacation satisfaction and outcome variables (Lounsbury & Hoopes, 1986; Westman
& Eden, 1997), whereas 1 study (Etzion, 2003) did not.
2.4. Discussion
The aim of this meta-analysis was to find out if vacation has a positive impact on H&W, how
long such beneficial effects would last, and whether vacation activities and experiences are
related to these outcomes. In a stepwise approach 7 studies were identified that could shed
light on these questions.
2.4.1. Vacation effect
There is evidence for a small effect of vacation on H&W. Average d was + 0.30, indicating
that well-being improved slightly following a vacation. In accordance with Effort-Recovery
Theory (Meijman & Mulder, 1998), the vacation effect was more prominent among outcome
variables that were closer to the core of the concept ‘H&W’, than among more distal variables.
Thus, health complaints and exhaustion as proximal health indicators improved more than life
satisfaction as a more distal indicator.
As only 4 studies reported the duration of the vacation, the relation between the
magnitude of effects and vacation length could not be established. Future research should
address this relation, eventually pointing to an “optimum point of recovery”. Subsequently,
such knowledge could be applied to develop guidelines for the scheduling and duration of
vacations.
2.4.2. Fade-out
There was also evidence for the post-vacation disappearance of vacation effects 2 to 4 weeks
post-vacation. The average d was -0.27. Regrettably the available information was too limited
to evaluate the precise course of fade-out and hence the duration of vacation effects. It seems
that (entire or partial) fade-out took place within 2 to 4 weeks post-vacation, but since the
second post-vacation measure was scheduled at least 2 weeks after vacation in all 4 studies,
we were not able to determine when beneficial effects on different variables exactly started to
diminish and were erased. Simple and frequent measures from the day of return until 8 weeks
after vacation would contribute to a better understanding of the course of fade-out.
Another interesting question is which factors might prolong vacation effects and
delay fade-out (Eden, 2001). Methods borrowed from cognitive therapy (e.g. brief daily writing
about positive vacation experiences) could be useful for this purpose.
50
Chapter 2
2
2.4.3. Vacation activities and experiences
The impact of vacation activities and experiences on vacation effects remains unclear hitherto.
Vacation activities as moderators of vacation effects have not been studied yet, while they may
be important behavioural determinants of positive and negative vacation outcomes.
The few results regarding vacation experiences suggest that vacation satisfaction
as well as negative work reflection do play a major role in influencing vacation outcomes in a
positive or negative way respectively. But until now, most reports on vacation experiences were
potentially biased because data were collected after returning home. To overcome this problem,
researchers need to include measurement occasions during vacation and ask respondents
about vacation expectations, activities (e.g. active versus passive, voluntary versus involuntary
activities), uplifts, hassles and (dis)satisfaction.
Surprisingly, there was even very limited information on vacation features like timing
and location available. Actually, basic information such as (average) vacation duration was
not reported in 3 cases. Most studies dealt with summer vacations. Furthermore, it remains
unclear in 5 studies whether participants stayed at home during their vacation or whether they
left their house and went away (Etzion, 2003; Fritz & Sonnentag, 2006; Gilbert & Abdullah 2004;
Lounsbury & Hoopes, 1986; Westman & Etzion, 2001). As spending time at a holiday resort
may well differ from spending time in one’s regular surroundings, future vacation research
should report vacation timing and location, to interpret findings in this light and to compare
different vacation features.
2.4.4. Methodological considerations
An intriguing issue in vacation research is the question of causality, i.e. were differences in
outcome variables before and after vacation indeed due to vacation? In many cases there
were plausible rival hypotheses, e.g., that pre-post vacation changes in work demands may
account for pre-post differences in health outcomes. Eden (2001, p. 178) called this tendency
of attributing changes in outcomes to vacation the “post hoc ergo propter hoc inference
fallacy”. Only an intensified repeated measure strategy can overcome this problem of limited
internal validity in the future.
Another frequent problem of earlier studies is the small number of respondents and
the accompanying attrition, possibly due to difficult recruitment and low compliance. This
might be counteracted by close collaborations with travel agencies, attractive rewards for
participants and devoted respondent care. The use of different kinds of attractive new media
(e.g. palm pilots, online surveys, mobile phones) could also support participant compliance
and prevent attrition.
Meta-analysis of Vacation Effects on Health and Well-being
51
2
The absence of a control group in most of the studies is also problematic. This
deficiency may partly be due to the fact that randomization into experimental and control groups
is difficult, if not impossible in vacation research. For instance, holiday and non-holiday takers
will differ anyway because non-vacationers may have many reasons for not going on vacation
like illness, lack of funds or abundance of work. The use of an internal referencing strategy
instead of a control group might be a better way to strengthen internal validity (Haccoun &
Hamtiaux, 1994). In this approach, additional variables are included that are similar to the
outcome variables but that are theoretically not expected to change because of a treatment
(i.e. vacation in our example). If these control variables do not change, whilst ‘real’ outcome
variables do, this is interpreted as empirical support for a true vacation effect. An example for
such a variable is teamwork competency.
A final shortcoming is the use of only self-reports in vacation research. With most
reviewed authors we agree that the use of other ‘objective’ measures like performance ratings
and physiological measures would be desirable.
2.4.5. Suggestions for future vacation research
Vacation research will profit from better designs, which boils down to the principle of repeated
measurements. Vacation research necessarily requires research on vacation: the assessment
of vacation activities and experiences during vacation itself. A suitable framework for structuring
diverse measurement occasions around a vacation period was developed by Westman and
Eden (1997) and consists of 2 pre-, 1 inter- and 2 post-vacation measurements. Its application
may well contribute to the comparability of future vacation research findings.
As discussed above, resolutions for earlier methodological problems, the detailed
investigation of the fade-out process by means of brief daily measures, studies on optimal
vacation duration, frequency and timing, and the design and evaluation of interventions to
prolong positive vacation effects, deserve a place on the vacation research agenda.
In general, neuroendocrine and cardiovascular measures are quite difficult and
costly to apply in field settings. Applications in vacation research may even be more difficult
as participants are out of sight of the researcher for a relatively long period and daytime
activity cannot be controlled for. However, as chronic incomplete recovery may manifest itself
in a disturbed balance of sympathetic and parasympathetic activity, also during sleep (e.g.
Akerstedt, 2006; Brosschot, Van Dijk & Thayer, 2007; Dahlgren, Kecklund & Akerstedt, 2006;
Hall et al., 2004; Rau & Triemer, 2004), a possibility for collecting physiological measures during
a vacation period would be, for instance, during night time. During sleep, parasympathetic
activation with its main restorative function should be dominant. High blood pressure levels,
high heart rate, low heart rate variability and high levels of catecholamine in morning urine
52
Chapter 2
2
would be strong markers of high sympathetic and low parasympathetic activation, and thus,
indicative of disturbed restorative functions and incomplete recovery.
Typically, moderators of vacation effects have hardly been studied. Still, vacation
research will benefit from the inclusion of moderators in the work context (e.g. job stressors, job
type), the non-work context (e.g. culture, relational problems, economic hardship) and person
characteristics (e.g. self-efficacy, workaholism). Moreover, different vacation features (duration,
timing and location) should be investigated and reported accurately to compare the effect of
different vacation types on outcome variables.
In conclusion, much has been learned from previous vacation studies. The general
picture that emerges from these pioneering studies is that vacation positively, though weakly,
impacts well-being. However, these positive effects do not last long. Future vacation research
may benefit from multiple measurements: before and after vacation, but especially during
vacation.
Meta-analysis of Vacation Effects on Health and Well-being
53
2
2.5. References
Akerstedt, T. (2006). Psychosocial stress and impaired sleep. Scandinavian Journal of Environmental
Health, 32, 493-501.
Beckers, D., van der Linden, D., Smulders, P. G. W., Kompier, M., Van Veldhoven, M. J. P. M., & Van
Yperen, N. W. (2004). Working overtime hours: Relations with fatigue, work motivation, and the
quality of work. Journal of Occupational and Environmental Medicine, 46, 1282-1289.
Brosschot, J. F., Van Dijk, E., & Thayer, J. F. (2007). Daily worry is related to low heart rate variability during
waking and the subsequent nocturnal sleep period. International Journal of Psychophysiology,
63, 39-47.
Clow, A. (2001). The physiology of stress. In F. Jonses, & F. Bright (Eds.), Stress, myth, theory and
research (pp.47-61). Harlow: Prentice Hall.
Cohen, J. (1988). Statistical power analysis for the behavioral sciences. Hillsdale, NJ: Erlbaum.
Dahlgren, A., Kecklund, G., & Akerstedt, T. (2006). Overtime work and its effects on sleep, sleepiness,
cortisol and blood pressure in an experimental field study. Scandinavian Journal of Work and
Environmental Health, 32, 318-327.
Eden, D. (1990). Acute and chronic job stress, strain, and vacation relief. Organizational Behaviour and
Human Decision Processes, 45, 175-193.
Eden, D. (2001). Vacations and other respites: studying stress on and off the job. In C. Cooper, & I. T.
Robertson (Eds.), Well-being in organizations (pp. 305-330). West Sussex: John Wiley & Sons,
Ltd.
Etzion, D. (2003). Annual vacation: duration of relief from job stressors and burnout. Anxiety, Stress, and
Coping, 16, 213-226.
Etzion, D., & Westman, M. (2001). Job stress, vacation, and the crossover of strain between spouses-
stopping the vicious cycle. Man and Work, 11, 106-118.
Fritz, C., & Sonnentag, S. (2006). Recovery, well-being, and performance-related outcomes: the role of
workload and vacation experiences. Journal of Applied Psychology, 91, 936-945.
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 & Health, 32, 482-492.
Gilbert, D., & Abdullah, J. (2004). Holidaytaking and the sense of well-being. Annals of Tourism Research,
31, 103-121.
Haccoun, R.R., & Hamtiaux, T. (1994). Optimizing knowledge tests for inferring learning acquisition
levels in single group training evaluation designs: the internal referencing strategy. Personnel
Psychology, 47, 593-604.
Hall, M., Vasko, R., Buysse, D., Ombao, H., Chen, Q, Cashmere, J. D., Kupfer, D., & Thayer, J. F. (2004).
Acute stress affects heart rate variability during sleep. Psychosomatic Medicine, 66, 56-62.
Härma, M. (2006). Workhours in relation to work stress, recovery and health. Scandinavian Journal of
Environmental Health, 32, 502-514.
54
Chapter 2
2
Hoopes, L. L., & Lounsbury, J. W. (1989). An investigation of life satisfaction following a vacation: a
domain-specific approach. Journal of Community Psychology, 17, 129-140.
Lounsbury, J. W., & Hoopes, L. L. (1986). A vacation form work: changes in work and nonwork outcomes.
Journal of Applied Psychology, 71, 392-401.
McEwen, B. S. (1998). Stress, adaptation, and disease: allostasis and allostatic load. Annals of the New
York Academy of Sciences, 840, 33-44.
Meijman, T. F., & Mulder, G. (1998). Psychological aspects of workload. In P. J. D. Drenth, H. Thierry, &
C. J. de Wolff (Eds.), Handbook of Work and Organizational Psychology (2nd ed., pp. 5-33).
Hove: Psychology Press.
Rau, R., & Triemer, A. (2004). Overtime in relation to blood pressure and mood during work, leisure, and
night time. Social Indicators Research, 67, 51-73.
Sluiter, J. K., Frings-Dresen, M. H. W., & Meijman, T. F.(2000). Reactivity and recovery from different
types of work measured by catecholamines and cortisol. A systematic literature overview.
Occupational Environmental Medicine, 57, 298-315.
Sterling, P., & Eyer, J. (1990). Allostasis: a new paradigm to explain arousal pathology. In S. Fisher &
J. Reason (Eds.), Handbook of Life Stress, Cognition, and Health (pp. 629-649). Chichester:
Wiley.
Strauss-Blasche, G., Ekmekcioglu, C., & Marktl, W. (2000). Does vacation enable recuperation? Changes
in well-being associated with time away from work. Occupational Medicine, 50, 167-172.
Strauss-Blasche, G., Ekmekcioglu, C., & Marktl, W. (2002). Moderating effects of vacation on reactions
to work and domestic stress. Leisure Sciences, 24, 237-249.
Totterdell, P., Spelten, E., Smith, L. , Barton, J., & Folkard, S. (1995). Recovery from work shifts: how long
does it take? Journal of Applied Psychology, 80, 43-57.
Tucker, P. (2003). The impact of rest breaks upon accident risk, fatigue and performance: a review. Work
& Stress, 17, 123-137.
Van der Hulst, M. (2003). Long work hours and health. Scandinavian Journal of Work, Environment and
Health, 29, 171-188.
Westman, M., & Eden, D. (1997). Effects of a respite from work on burnout: vacation relief and fade-out.
Journal of Applied Psychology, 82, 516-527.
Westman, M., & Etzion, D. (2001). The impact of vacation and job stress on burnout and absenteeism.
Psychology and Health, 16, 595-606.
57
Chapter 3
Effects of Vacation from Work on Health and
Well-being: Lots of Fun, Quickly Gone
Although vacation from work provides a valuable opportunity for recovery, few studies
have met the requirements for assessing its effects. These requirements include taking
measurements well ahead of the vacation, during vacation, and at several points in time
afterwards. Our study on vacation (after-) effects focused on two related questions: 1) Do
H&W of working individuals improve during a vacation?, and 2) How long does a vacation
effect last after resumption of work?
In a longitudinal study covering 7 weeks, 96 Dutch workers reported their H&W levels 2
weeks before a winter sports vacation, during vacation, and 1 week, 2 weeks and 4 weeks
after vacation on 7 indicators.
Participants’ H&W improved during vacation on 5 indicators: health status, mood, tension,
energy level and satisfaction. However, during the first week of work resumption, H&W had
generally returned to pre-vacation levels.
In conclusion, a winter sports vacation is associated with improvements in self-
reported H&W among working individuals. However, these effects fade out rapidly after
work resumption. We propose a framework for future vacation research and suggest
investigating the role of vacation type, duration, and possibilities to prolong vacation relief.
This chapter is based on:
De Bloom, J, Geurts, S.A.E., Taris, T.W., Sonnentag, S, De Weerth, C., & Kompier, M.A.J (2010).
Effects of vacation from work on health and well-being: Lots of fun, quickly gone. Work &
Stress, 24, 196-216.
Effects of Vacation on Health and Well-being
59
3
3.1. Introduction
Research in the field of occupational health has consistently demonstrated the adverse impact
of stress in the workplace on individuals’ H&W (e.g. Belkic, Landbergis, Schnall & Baker, 2004;
Ferrie, Westerlund, Virtanen, Vahtera & Kivimaki, 2008). This harmful effect is, in part, brought
about by physiological stress responses that continue or recur during nonwork time when job
stressors are no longer present (e.g. Brosschot, van Dijk & Thayer, 2007; Hjortskov, Rissen,
Blangsted, Fallentin, Lundberg & Sogaard, 2004). These prolonged physiological stress
responses can be amplified by ruminating thoughts about past and potential future stressors
(Geurts & Sonnentag, 2006) and may disturb the person’s homeostatic balance (‘allostasis’,
McEwen, 1998), that is, the balance between the sympathetic nervous system being dominant
during effort expenditure (e.g. in response to stressors) and the parasympathetic nervous
system being in control during rest and relaxation (e.g. recovery).
Accordingly, recovery during nonwork time plays a crucial role in protecting
employees against the adverse effects of exposure to job stressors. According to Geurts and
Sonnentag (2006), the essence of recovery is that “[...] the psychophysiological systems that
were activated during work will return to and stabilize at a baseline level, that is, a level that
appears in a situation in which no special demands are made on the individual” (p. 483). The
most influential theories on recovery, Effort-Recovery Theory (Meijman & Mulder, 1998) and
Allostatic Load Theory (McEwen, 1998), share the idea that removal of demands previously put
on the individual’s psychobiological systems is a prerequisite for recovery to occur.
Recovery after work may occur regularly between workdays (e.g. during evening
hours and during weekends) and during longer periods of off-job time such as vacations,
constituting meta- and macro-recovery, respectively (Sluiter, Frings-Dresen, Meijman & Van
der Beek, 2000). Recent diary studies have revealed that workers often recover insufficiently
during regular evening hours and weekends, for instance due to working overtime (Fritz &
Sonnentag, 2005; Van Hooff, Geurts, Kompier & Taris, 2007). This day-to-day incomplete
recovery constitutes a high risk for serious health impairment in the long term (Van Hooff,
Geurts, Kompier, Taris, Houtman & van den Heuvel, 2005).
Vacation as a longer and relatively uninterrupted period of absence from work is a
prime candidate for helping workers to recover more completely from work. Vacation may
contribute to recovery from work through a rather passive mechanism of liberation from
demands, as well as through the active engagement in valued and positively experienced free-
time activities of one’s own choice (e.g. family activities and hobbies).
According to Fredrickson’s Broaden- and -Build Theory (2001), positive emotions
produce flourishing by widening people’s thought-action repertoires and by building enduring
60
Chapter 3
3
resources (e.g. intellectual, physical, social and psychological). Positive emotions (e.g. joy,
contentment and love) experienced during vacation may not only strengthen the social bond
with partners, family members and/or friends, they may also break habitual thought patterns and
lead to unusual, creative, fresh ideas to solve long-lasting (job-related) problems. Therefore,
a vacation may help to build up enduring personal resources that may function as a buffer for
future threats.
In the current study, we therefore aim to answer two central research questions:
Question 1: Do H&W of working individuals improve during a winter sports vacation (i.e.,
vacation effect)?
Question 2: Once a vacation effect has occurred, how long does it last after resumption
of work (i.e., vacation after-effect)?
Although vacation is probably the most powerful prototypical respite occasion for
working individuals, as yet surprisingly few researchers have addressed its impact on recovery
from work. A recent meta-analysis of vacation research (De Bloom, Kompier, Geurts, De
Weerth, Taris & Sonnentag, 2009) identified only seven studies that met a set of minimum
methodological requirements for studying the effects of vacation on H&W. The results of these
seven studies suggest that vacation has positive, although weak effects on H&W, and that
these effects fade out quickly after returning home. However, the evidence is still inconclusive,
not only because of the small number of vacation studies, but also due to suboptimal research
designs often applied (De Bloom et al., 2009). We believe that an adequate study design to
investigate the impact of vacation on employees’ H&W comes down to five major criteria. In the
following sections we will discuss each of them in more detail.
3.1.1. A proper pre-vacation baseline
A number of studies included in the meta-analysis scheduled their pre-vacation measurements
shortly before participants went on vacation (De Bloom et al., 2009). However, research showed
that the time before a trip can be stressful (DeFrank, Konopaske & Ivancevich, 2000). In a
similar vein, Westman (2004, 2005) stated that pre-vacation activities like planning the vacation,
travelling to the vacation destination and coordinating work tasks for the period of absence
may also cause pre-vacation stress. Accordingly, it is plausible that measurement occasions
immediately before vacation are confounded by either “vacation preparation stress” or working
to deadlines before leaving (“working ahead-stress”). But it may also be that vacationers look
forward to the vacation, inducing enhanced H&W. In both cases, it is unreasonable to expect
that levels of H&W in the week before vacation represent baseline levels of a regular working
week. Therefore, in the current study, all comparisons to investigate vacation effects were
anchored by a baseline during a regular working week, two weeks before vacation.
Effects of Vacation on Health and Well-being
61
3
3.1.2. An on-vacation measurement occasion
A concern in some earlier vacation studies regards the absence of H&W measurements
during the vacation period itself (for notable exceptions see Eden, 1990; Fritz & Sonnentag,
2006; Westman & Eden, 1997). In most of the earlier vacation studies, pre- and post-vacation
measurements were compared, and changed levels of H&W were attributed to the unmeasured
intervention, that is, the vacation (De Bloom et al., 2009). However, attributing a change in
H&W to the vacation is a fallacy of the “post hoc ergo propter hoc”-type (“after this, therefore
because of this”), because the sequential occurrence of phenomena does not mean that there
is a causal relation between these phenomena (Eden, 2001).
The reason for the dominantly chosen pre-post comparisons to determine a vacation
effect is presumably that obtaining data while people are on holiday is difficult (Eden, 2001).
Some researchers have even described the logistics of locating people during vacation as
“nightmarish” (Eden, 1990, p.182). Furthermore, respondents might possibly not appreciate
being examined during their highly valued holidays (i.e. “holy days”), and traditional research
materials like paper-pencil questionnaires are hard to use in a vacation setting.
However, investigating a vacation effect by only comparing pre-vacation and post-
vacation measurements is inadequate because post-vacation measurements are biased by
work resumption and fade-out may already have set in. Every measurement occasion after
vacation will therefore reflect an after-effect of vacation and probably underestimate the
genuine vacation effect. Accordingly, the use of a pre-post vacation design does not allow us to
disentangle vacation- and vacation after-effects and can lead to erroneous conclusions about
the effect of vacation on H&W. As a consequence, it is essential to obtain information about
H&W during vacation in order to draw such conclusions. In the current study, we included two
on-vacation measurement occasions and defined a “genuine” vacation effect as a significant
change in H&W levels during vacation compared to pre-vacation baseline levels.
3.1.3. Multiple post-vacation measurement occasions
Insufficient attention has been paid to the fade-out process of vacation effects, once they have
occurred. As a consequence, it remains largely unknown when fade-out sets in, what its exact
course is and when positive after-effects of vacation have completely vanished (De Bloom et
al., 2009). Vacation effects are by definition temporary, as any positive effect of vacation will
fade out sooner or later, for instance, due to the renewed exposure to work demands. Because
previous research suggests that vacation effects fade out rapidly (De Bloom et al., 2009), it is
necessary to measure levels of H&W immediately after vacation.
In addition, only a few vacation studies have employed more than one post-vacation
measurement occasion and, if they have done so, the time lag between the two post-vacation
62
Chapter 3
3
occasions has varied widely. In most cases, the first post-vacation occasion has been
scheduled in the first week after work resumption and the second post-vacation occasion at
least two weeks later (Etzion, 2003; Westman & Eden, 1997; Westman & Etzion, 2001). As a
result, there was no information available on H&W during the second week after vacation. To
close this time gap, we collected data not only in the first week but also in the second work
week after vacation.
Occasionally, previous studies have found longer lasting vacation effects (e.g.
Westman & Eden, 1997). Moreover, De Lange and colleagues suggest that longitudinal studies
should apply many follow-up measures that are both evenly and unevenly spaced (De Lange,
Taris, Kompier, Houtman & Bongers, 2003). Therefore, we also included a third post-vacation
occasion four weeks after work resumption.
3.1.4. Minimalism and simple comparisons
Vacation research is complex, because it necessarily involves a repeated-measures design.
Comparisons between measurement occasions to investigate vacation effects and their
duration should be as straightforward, logical and simple as possible. In our view, the essence
of vacation research can be reduced to the vacation effect and its potential after-effects.
A vacation effect reflects the difference in H&W levels between the pre-vacation
measurement occasion (baseline) and the on-vacation measurement occasion(s). A
comparison of the post-vacation measurement occasions with the on-vacation measurement
occasion reveals whether there may be short term, mid-term and long term after-effects of a
vacation period. To determine when vacation effects have diminished completely (i.e. baseline
levels are attained again) it makes sense to also compare post-vacation measurements
with pre-vacation baseline levels. Therefore, in our study, after-effects were investigated by
comparing H&W levels after vacation with both on-vacation levels and pre-vacation baseline
levels.
3.1.5. Equal and exact timing of measurement occasions for every participant
Whilst earlier vacation studies had “… no precedent for ideal timing …” of measurements
(Westman & Eden, 1997, p.519) and were often rather vague in reporting when exactly
measurements took place, we could base the timing of our measurement occasions on earlier
findings (see reasoning above) and link every occasion to an identical point in time before,
during and after vacation for every single participant. Even the time of the day was kept as
constant as possible.
In our study, pre-vacation baseline levels (Pre) were measured two weeks before
vacation. The on-vacation levels (Inter) were measured during vacation itself, on the second
Effects of Vacation on Health and Well-being
63
3
day after arrival and on the second-last day before departure. The post-vacation levels were
measured during the first (Post 1), the second (Post 2) and the fourth week (Post 3) after
returning home and resuming work. Figure 3.1 presents the research design employed in this
study.
A vacation effect is present when H&W levels during vacation are higher than pre-
vacation levels (Pre versus Inter). The existence of a short term after-effect can be detected by
comparing the on-vacation measurement occasion with the first post-vacation measurement
occasion (Inter versus Post 1). In case of an improvement in H&W from Pre to Inter and no
significant differences between Inter and Post 1, vacation effects apparently persist which is
supportive of a short term after-effect.
If post-vacation levels are lower than on-vacation levels, these post-vacation levels
will be compared with pre-vacation levels to determine when baseline levels are reached again.
In the case of significant differences between the pre-vacation and the first post-vacation
measures (Pre versus Post 1), vacation effects apparently endure (supportive of a short term
after-effect).
The existence of a mid-term after-effect will become evident by comparing the second
post-vacation occasion with the on-vacation measurement occasion (Inter versus Post 2) as
well as with the pre-vacation levels (Pre versus Post 2). A significant difference between the pre-
vacation and the second post-vacation levels would be supportive of a mid-term after-effect.
If participants’ H&W levels on the second post-vacation occasion are still higher than
baseline levels (indicating that the vacation effect still persists), we proceed with a final set of
After-effects
Vacation
effect
2nd week
Vacation
M = 9 days
SD = 2days
Baseline
2 weeks
1 week 3rd week
Pre Inter Post 1 Post 2 Post 3
4th week
1st week
After work resumption Before vacation
During
vacation
Figure 3.1:
Research design for the current study
64
Chapter 3
3
comparisons (Inter versus Post 3 and Pre versus Post 3) to decide if vacation has long term
after-effects.
3.2. Method
3.2.1. Data collection procedure
We carried out a longitudinal field study on winter sports vacations because this type of holiday
normally covers one week and vacationers usually have no more than one or two days off before
departure and after return. As a result, vacation duration and the time before and after vacation
were roughly comparable for all participants. The same is true for the vacation activities that
people typically engage in: winter sports activities during the day (Nordic skiing, alpine skiing,
snowboarding, sledding, skating) and socializing (après-ski) in the evening. Consequently,
winter sports holidays represent a type of vacation that is more uniform with respect to activities
and duration than, for instance, summer vacations and therefore well suited for our research
purposes.
Our study covered a time span of seven weeks around the vacation period, including
the vacation itself and took place between February 15
th
and April 15
th
2008. On all measurement
occasions during working periods i.e., two weeks before vacation (Pre), and the first (Post 1),
second (Post 2) and fourth week (Post 3) after returning home, the participants received an
e-mail with a link to a digital diary twice a week. Participants were asked to complete the diary
just before bedtime on a fulltime working day. To make sure that participants would not forget
to complete the digital diary in the evening, they additionally received a reminder text message
(SMS) on their cell phone earlier that day.
In order to take on-vacation measures of H&W, the participants were provided with
cell phones with international pre-paid SIM cards to take with them on holiday. They were asked
to return the cell phones after returning home in a pre-stamped envelope. While on holiday,
every participant was called on this cell phone and interviewed by one of the researchers on the
second day after arrival and on the second-last day before departure between five and seven
pm (Inter measurement occasion).
Before the cycle of data collection started, participants received a card with an
overview of their personal measurement occasions during the seven-week period. After the
whole cycle of data collection, respondents were thanked for their participation, were given the
opportunity to comment on the research procedure and received information about the time
when the results were expected to be published in the academic literature and on our website.
Effects of Vacation on Health and Well-being
65
3
To encourage participation and to reduce missing data, we announced a lottery prize
among all participants: a one-week winter sports holiday for the next winter sports season.
Chances of winning were higher for participants who returned all questionnaires than for
participants who missed measurement occasions. In May, the winner was drawn by lot and
made public. Moreover, every participant received 10 Euro as pre-paid talk credit on his or her
vacation-phone.
3.2.2. Missing data: prevention and treatment
Missing data constitute a major problem in longitudinal designs (Taris, 2000) and effective
strategies to prevent and deal with missing data were applied. First of all, because we assumed
that especially well-informed participants would comply with our intensive data collection
procedure, we devoted much attention to instructing them on the research procedure.
Second, we scheduled two measurement occasions within each week. In order
to obtain a reliable indicator of the week-level of H&W, the two within-week measures of a
particular H&W indicator were averaged. This approach also served to prevent missing data in
case of a single non-answered prompt during a workweek. In that case the other measurement
in that week (if available) was treated as the week average.
Third, for data collection, we used electronic mail and SMS to remind the participants
to fill in the questionnaires at the correct moment in time. Because we used digital diaries, we
could recognize un-answered prompts immediately, and a detailed non-completion script was
applied for the digital diaries as well as the telephone surveys. These strategies reduced the
amount of missing data.
Finally, in anticipation of possible technical problems with the mobile vacation
phones, a sealed envelope containing a paper-and-pencil questionnaire with the interview
questions was sent to the participants before departure as backup. When all attempts to reach
a participant by phone failed, we sent a SMS that allowed participants to open the envelope
and to fill in the questionnaire. Nine measurements during vacation were in fact paper-and-
pencil questionnaires returned in a pre-stamped envelope.
In order to guarantee the reliability and comparability of the measurements, we
excluded data from the digital diary (a) if participants filled in the questionnaire on non-work
days instead of on fulltime working days, and (b) if participants completed the questionnaire
between 6 am and 6 pm instead of just before bedtime.
Considering the ten measurements per individual, 83 respondents replied to at least
eight single measurements (digital diaries and telephone interviews during vacation). Based
on a maximum of 960 possible single measurements in this study (10 measurements in 96
persons), the overall completion rate was 87% (834 measurements). The combination of the ten
66
Chapter 3
3
measurements (two measurements a week) into five occasions resulted in even more reliable
week-indicators and high completion rates: 100% (N = 96) on Pre, 98% (N = 94) on Inter, 90%
(N = 86) on Post 1 and 96% (N = 92) on Post 2 and Post 3. For 83 of the 96 participants data
sets were complete (no missing data on any of the five occasions).
3.2.3. Participants
To recruit participants in the Netherlands, we distributed information via travel agencies, winter
sports websites, shops for skiing-equipment, winter sports journals and newspaper ads.
Additionally, we visited a winter sports fair and contacted ski-clubs (i.e. sporting clubs for skiers
who jointly exercise for their upcoming winter sports holiday).
As a result of the recruitment procedure, 176 persons indicated that they were
interested in taking part in this study. After administering detailed information about the
research procedure and promising confidentiality, these 176 persons received a phone call
from one of the researchers. During this call, possible questions about the research scheme
were answered and the participants were screened for participation prerequisites: participants
(i) had to work at least 24 hours per week (18 exclusions), (ii) go on winter sports vacation for
at least 1 week between February 15th and April 15th, 2008 (22 exclusions), and (iii) enrol in
the study on time (17 exclusions). Persons working extremely irregular schedules were also
excluded (4 exclusions). Moreover, a small number of interested persons did not want to be
called during vacation (4 exclusions), did not use electronic mail (5 exclusions) or found the
research procedure too burdensome (3 exclusions). Another seven persons were excluded
because they did not go on vacation after all due to sickness. All in all, of the 176 people who
were initially interested, 108 met the inclusion criteria. Of those 108, 96 actually took part in the
study, resulting in a 89% response rate.
The majority of this Dutch sample was male (65%), mean age was 44 years (SD =
10 years) and as regards education 5% of the sample was lower (no secondary education,
lower secondary or junior secondary education), 40% medium (senior general secondary and
university preparation education) and 55% highly educated (higher professional and higher
education). The majority of the respondents were employed (82%) whilst 18% were self-
employed. The participants worked in a variety of sectors: 23% worked in the commercial
sector, 20% were higher educated specialists (e.g. engineers, ICT-workers), 14% worked in the
service sector, 12% in health care, 11% were administrative employees, 7% were craftsmen or
worked in the production industry, 4% were teachers, and the remaining 9% worked in other
sectors.
The participants worked at least 24 contractual hours per week and weekly work
hours (including overtime) varied from 24 to 60 hours. Average working time was 38 hours
Effects of Vacation on Health and Well-being
67
3
per week (SD = 8 hours). Forty-seven percent of the participants supervised other persons,
whereas 53% had no supervisory tasks. In terms of their personal living situation, the majority
of the respondents (57%) was married and lived with at least one child, 29% were married and
lived without children, 9% were unmarried and lived alone, 2% were single parents and 2% lived
in their parents’ house.
The mean vacation duration was 9 days (SD = 2 days, range: 7-19 days). Vacation
destinations were typical winter sports areas, with the top-three destinations being Austria
(70%), France (15%) and Switzerland (6%). Most of the respondents were experienced skiers:
every participant had been on a skiing vacation at least one time before, and the average
number of previous skiing vacations was 22 (SD = 15 times).
3.2.4. Measures
In order to be able to give a detailed account of H&W, we incorporated a range of different H&W
indicators. To prevent non-response we minimized the effort required from the participants and
maximized user-friendliness by reducing the number of digital diary questions as much as
possible. Therefore, we employed seven single-item measures to tap the seven main indicators
of H&W: sleep quality, health status, mood, fatigue, tension, energy level and satisfaction.
Single-item measures often have a high face validity, and participants value their
directness and lack of redundant and repeated comparable items. Accordingly, multiple item
measures may be validly replaced by single-item measures and still be psychometrically
acceptable if the underlying constructs are sufficiently one-dimensional and unambiguous to
the participants (e.g. Elo, Leppänen & Jahkola, 2003; Van Hooff, Geurts, Taris & Kompier,
2007).
For simplicity, we adapted response scales based on the well-known basic Dutch
grade notation system ranging from 1 (extremely low/negative) to 10 (extremely high/positive)
and anchored the first and the last grade. The exact wording of each single-item measure and
the anchors can be found in Table 3.1.
3.2.5. Statistical approach
The data were analyzed in a 5 (Occasion: five occasions) × 7 (health and wellbeing: seven
H&W indicators) multivariate analysis of variance (MANOVA) with repeated measures on both
Occasion (the independent variable or factor) and H&W (our criterion variables). Subsequently
follow-up univariate ANOVAs were performed for each of the seven H&W indicators separately
(cf. DeShon & Morris, 2003).
The vacation effect (Question 1) was examined by computing Fisher’s Least Significant
Difference (LSD) test for Pre versus Inter, presenting Cohen’s d for paired observations (Cohen,
68
Chapter 3
3
1988, p.46) as an effect size. Following Cohen (1988) we distinguished among small (0 to 0.5),
medium (0.5 to 0.8) and large (> 0.8) effect sizes.
In order to test if there was a short term after-effect of the vacation (Question 2), we
compared the on-vacation measure (Inter) with the first post-vacation occasion (Post 1). In
a next step, the comparison of Pre versus Post 1 told us if H&W indicators had returned to
baseline levels.
For H&W indicators that did not attain baseline at Post 1, we examined post-hoc
Fisher’s LSD differences between Inter and Post 2 to test if vacation effects still persisted and a
mid-term after-effect applied. The post-hoc Fisher’s LSD test between Pre and Post 2 informed
us about the strength and duration of this potential mid-term after-effect.
Only in case of a mid-term after-effect, we examined the post-hoc differences between
Inter versus Post 3 and Pre versus Post 3 to determine if there was a long term after-effect.
3.3. Results
3.3.1. Preliminary analysis: descriptive statistics
Pearson product moment correlations were examined to establish the relationship between the
seven different H&W indicators on the five measurement occasions. The full 35 by 35 table (five
occasions multiplied by seven H&W indicators) is available on request from the first author.
Autocorrelations that can be interpreted as test-retest reliability coefficients ranged
from .06, ns, for the Pre and Inter measures of sleep, to .67, p < .001, for the Post 2 and Post
Table 3.1:
Description of the seven single-item measures used in this study
H&W
indicators
Single item measure Score of 1
means…
Score of 10
means…
Sleep quality How did you sleep last night? Very badly Very good
Health status How healthy did you feel today? Very unhealthy Very healthy
Mood How was your mood today? Very bad Very good
Fatigue How fatigued did you feel today? Not fatigued at all Very fatigued
Tension How tense did you feel today? Very calm Very tense
Energy level How energetic do you currently feel? Absolutely not
energetic
Very
energetic
Satisfaction How satisfied do you feel about this day? Very dissatisfied Very satisfied
Effects of Vacation on Health and Well-being
69
3
3 measures of energy level. The correlations among the seven H&W indicators on the same
measurement occasions ranged, for Pre, between -.28 (p < .01, fatigue and sleep quality) and
.78 (p < .001, mood and satisfaction), for Inter between .08 (ns, satisfaction and energy level)
and .68 (p < .001, mood and health status), for Post 1 between .04 (ns, energy level and sleep
quality) and .76 (p < .001, satisfaction and mood), for Post 2 between .09 (ns, energy level
and health status) and .82 (p < .001, satisfaction and mood), and for Post 3 between -.16 (ns,
tension and sleep quality) and .71 (p < .001, satisfaction and mood). So, the H&W indicators
were interrelated, but not identical. Mean scores for the seven H&W indicators across the five
measurement occasions are presented in Table 3.2 and Figure 3.2.
2
3
4
5
6
7
8
Pre InterPost 1Post
2P
ost 3
Sleep quality
Health status
Mood
Fatigue
Tension
En ergy level
Satisfaction
Figure 3.2:
Line diagram of means for H&W indicators across the five measurement occasions.
70
Chapter 3
3
Table 3.2:
Means and standard deviations on all five occasions and occasion effects, vacation effects (research question 1) and vacation after-effects (research question 2) for
all H&W indicators
H&W
indicators
Effect sizes (Cohen’s d) for various comparisons
Means and (standard deviations) for 5
occasions
Occasion
effect
Vacation
effect
Short term after-effect Mid-term after-effect
Pre Inter Post 1 Post 2 Post 3
F
partial eta
Pre vs.
Inter
Inter vs.
Post 1
Pre vs.
Post 1
Inter vs.
Post 2
Pre vs.
Post 2
Sleep quality
7.42
(1.05)
7.46
(1.31)
7.62
(1.00)
7.42
(1.24)
7.18
(1.26)
1.93
0.09
- - - - -
Health status
7.53
(1.24)
7.98
(1.26)
7.63
(1.44)
7.53
(1.12)
7.45
(1.04)
3.88**
0.16
0.28** - 0.23* 0.06 Back to baseline at Post 1
Mood
7.28
(1.17)
8.27
(0.99)
7.41
(1.10)
7.28
(1.13)
7.31
(1.12)
15.91**
0.45
0.71** - 0.66** 0.11 Back to baseline at Post 1
Fatigue
4.42
(1.99)
4.64
(1.99)
3.81
(1.72)
4.28
(1.62)
4.60
(1.71)
3.78**
0.16
- 0.09 0.40** 0.29** 0.16 0.07
Tension
3.43
(1.80)
2.31
(1.16)
3.25
(1.53)
3.55
(1.78)
3.71
(1.71)
21.68**
0.52
0.71** - 0.64** 0.11 Back to baseline at Post 1
Energy level
5.90
(1.90)
6.84
(1.37)
5.96
(1.89)
5.90
(1.74)
5.51
(1.74)
9.13**
0.32
0.46** - 0.46** 0.03 Back to baseline at Post 1
Satisfaction
7.32
(1.05)
8.14
(1.10)
7.34
(1.15)
7.25
(1.09)
7.28
(0.95)
11.43**
0.37
0.59** - 0.59** 0.02 Back to baseline at Post 1
Note: * p < .05, ** p < .01. Pre = 2 weeks before vacation, Inter = during vacation, Post 1 = 1
st
week of work resumption, Post 2 = 2
nd
week of work resumption, Post
3 = 4
th
week of work resumption. Baseline = levels of H&W indicators 2 weeks before vacation (Pre).
Effects of Vacation on Health and Well-being
71
3
With regard to the on-vacation measurements of H&W, there were no systematic
differences between reports collected by telephone interviews and the nine reports collected
by paper-and-pencil questionnaires (t (85) < 1.30, p > .05).
3.3.2. Multivariate analysis
Multivariate analysis of variance revealed main effects of Occasion, F (4,79) = 7.29, p < .001,
and of H&W, F (6,77) = 140.35, p < .001, as well as a significant Occasion × H&W interaction
effect, F (24,59) = 7.20, p < .001. Hence, H&W varied significantly across the five occasions,
and this across-time change was different for the various H&W indicators.
3.3.3. Univariate analysis
Follow-up univariate ANOVAs for the H&W indicators across the five measurement occasions
revealed that the levels of six indicators varied significantly across the five occasions (Table
3.2). Sleep quality was the only indicator that did not show an overall occasion effect, F (4, 79)
= 1.93, ns, meaning that sleep quality did not differ significantly before, during and after the
vacation period.
3.3.4. Research question 1: Do H&W of working individuals improve during a winter
sports vacation (i.e., vacation effect)?
To answer the first research question, we compared the pre-vacation measures of the six
H&W indicators with the measures taken during vacation (Inter). Five out of seven indicators
showed an overall occasion effect with Pre levels of H&W being significantly different from Inter
levels (p < .01). During the vacation participants felt healthier, were in a better mood, felt more
energized, were more satisfied and reported lower tension than during the regular working
week before they went on vacation. Effect sizes were medium for satisfaction (d = 0.59), mood
(d = 0.71) and tension (d = 0.71), and small for energy level (d = 0.46) and health status (d =
0.28). The level of fatigue was not significantly different during the vacation period compared
to the pre-vacation baseline (p = .74).
Overall, self-reported H&W significantly improved during vacation. The mean absolute
effect size d for the difference between Pre and Inter in all seven H&W indicators was 0.38,
indicating a small positive vacation effect (ds were 0.02 for sleep, 0.28 for health status, 0.71
for mood, -0.09 for fatigue, 0.71 for tension, 0.46 for energy level, and 0.59 for satisfaction).
72
Chapter 3
3
3.3.5. Research question 2: Once a vacation effect has occurred, how long does it
last after work resumption (i.e., vacation after-effects)?
To test if there was a short term after-effect, we conducted post-hoc Fisher’s Least Significant
Difference (LSD) tests for the difference between the on-vacation measure (Inter) and the first
post-vacation occasion (Post 1). For all six H&W indicators, there was a significant difference
between Inter and Post 1. For five of the six indicators, self-reported H&W had declined
significantly immediately after participants had returned home and resumed work. Effect
sizes were small for health status (d = -0.23) and energy level (d = -0.46), and medium for
satisfaction (d = -0.59), tension (d = -0.64) and mood (d = -0.66). For fatigue, findings were
different: levels of fatigue had decreased rather than increased directly after vacation (d =
0.40), indicating a positive short term after-effect.
An inspection of the means of the H&W indicators (Table 3.2) already provided
interesting insights: an increase from Pre to Inter was followed by an immediate decrease in
H&W of nearly the same amount from Inter to Post 1, resulting in almost baseline levels again.
The mean score for health status increased by 0.45 points during vacation and decreased by
0.35 points from Inter to Post 1. The same pattern could be observed for mood (0.99 increase
during vacation, 0.86 decrease at Post 1), energy level (0.94 increase, 0.88 decrease), and
satisfaction (0.82 increase, 0.80 decrease). Tension showed a similar pattern in the reversed
direction (1.12 decrease during vacation, 0.94 increase at Post 1). Standardized effect sizes d,
which enabled us to compare the rise and fall within the seven H&W indicators relative to each
other, mirrored this development across time.
Post-hoc tests of the difference between Pre versus Post 1 were non-significant in
five of the six H&W indicators, indicating that during the first week after vacation, there was
a return to baseline levels for health status, mood, tension, energy level and satisfaction. The
lowest levels of fatigue were found at Post 1 and accordingly there was a significant decrease
in fatigue from Pre to Post 1, resulting in a positive effect size d of 0.29.
Because every single H&W indicator except fatigue had reached baseline levels
again at Post 1, we only conducted post-hoc tests for a mid-term after-effect in fatigue. As
fatigue was lowest on Post 1 and had similar levels at Pre, Post 2 and Post 3, the differences
between Inter versus Post 2 and Pre versus Post 2 were indeed non-significant (ps were .30 and
.44, respectively). So, fatigue had returned to baseline levels at Post 2.
In conclusion, self-reported H&W had declined rapidly after resumption of work: five
of the six H&W indicators (health status, mood, tension, energy level, satisfaction) had returned
to baseline levels within the first week of work resumption (Post 1), meaning that vacation had
no short term, mid-term or long term after-effect. Fatigue showed a different pattern of rise and
Effects of Vacation on Health and Well-being
73
3
fall, with the lowest level at Post 1 and levels comparable to baseline at Post 2, suggesting a
short term after-effect.
3.3.6. Process evaluation
In an evaluation of the research procedure, 63% of the respondents reported to have
enjoyed participating in our study and only 17% found the research procedure a little boring
or time consuming. The great majority appreciated the digital diaries (94%) and 66% found
the reminder SMS very useful. Only a small percentage (9%) indicated that the phone call
interfered somewhat with their vacation. The great majority (65%) indicated that being called
during vacation was “no problem”. The majority (93%) even judged the vacation phones as a
very good and creative idea.
3.4. Discussion
3.4.1. Vacation effect
Our study provided evidence for improvements in self-reported H&W during a winter sports
vacation. The average effect size for the vacation effect computed across the seven health and
wellbeing indicators was d = 0.38 (small). This effect was present for five of the seven H&W
indicators employed in this study. In particular, workers felt more satisfied and experienced more
positive mood and less tension during vacation compared to a regular pre-vacation working
week. In addition, although to a lesser extent, workers felt more energized and healthier during
vacation than before vacation.
These findings strongly support the idea of a vacation as a powerful opportunity to
recover from work demands and to benefit from positive free-time experiences. Regarding
fatigue and sleep quality, participants’ reports did not differ between the on-vacation and the
pre-vacation occasions. The finding that mood, tension and satisfaction were more strongly
affected by vacation than, for instance, health status may reflect the fact that the former
aspects of H&W are more sensitive to changes in stressors and work demands and fluctuate
more easily from day to day, than the latter.
We believe that current study has several strengths, specifically, a research design
with multiple repeated measures before, during and after vacation. We succeeded in carrying
out 10 repeated measurements per individual (two measurements for each of the five
occasions) during a seven-week period in a substantial group of 96 vacationers. Hereby, we
applied a proper pre-vacation baseline measurement during a regular working week two weeks
prior to vacation and we were able to assess the after-effects of vacation by monitoring H&W on
74
Chapter 3
3
three measurement occasions after vacation. Additionally, our study is one of the few studies
that measured H&W during vacation itself. The importance of the inclusion of on-vacation
measurements to determine the “genuine” vacation effect can easily be illustrated: if we had
left out the on-vacation occasion, we would falsely have concluded that vacation generally had
no positive effect on H&W.
The combination of traditional and new media gave us the opportunity to generate
rich datasets in a reliable, user-friendly way and to reduce missing data and attrition drastically
by acting upon the principle ‘the more you measure, the less the pleasure’. This means, we
measured frequently but in a comfortable manner by restricting the number of questions
to a minimum and by designing easy to use instruments and resources like digital diaries,
telephone surveys and SMS reminders. The process evaluation of the participants confirms
that our approach was generally experienced positively.
Our findings showed that sleep quality and fatigue had not improved on vacation
compared to the pre-vacation baseline. Previous research has suggested, however, that sleep
quality and stress are closely related (e.g. Akerstedt, 2006) and that sleep quality improves
in times of low stress (Dahlgren, Kecklund & Akerstedt, 2005). It is possible that the potential
beneficial effects of low stress and rest on sleep quality may have been outweighed by specific
vacation circumstances, such as a reduced number of hours sleep, an unfamiliar sleeping
environment (e.g. a different bed, different sounds, and light and temperature conditions) and
changes in sleep-relevant behaviour. Regarding the latter, it is not uncommon during a winter
sports vacation to drink substantial amounts of alcohol during the après ski (Meyers, Perrine
& Caetano, 1997), which might in turn lead to sleep disruption (Roehrs & Roth, 2001). It is
conceivable that the beneficial effect of low stress and rest on sleep quality only occurs for
those who sleep enough or consumed low amounts of alcohol before going to bed.
Hence, we tested in a number of post-hoc analyses whether the relationship between
pre-vacation and during-vacation sleep quality varied as a function of the number of hours sleep
and of alcohol consumption before going to sleep during vacation (i.e. the number of glasses
of alcoholic beverages). These analyses revealed no main or moderator effects of sleep hours
(F’s (1, 92) < 1.26, ns) on sleep quality. The same was true for alcohol consumption (Fs (1, 92)
< 1.20, ns). So we concluded that neither the number of hours participants slept, nor alcohol
consumption during vacation explained why sleep quality did not improve during vacation. We
cannot rule out that physical sleeping circumstances may have accounted for the absence of
a vacation effect on sleep quality.
Contrary to our expectations, we found the lowest levels of fatigue immediately
after vacation instead of during vacation. Strictly speaking, this effect cannot be labelled an
after-effect of vacation, since levels of fatigue on vacation did not differ significantly from pre-
Effects of Vacation on Health and Well-being
75
3
vacation levels, indicating the absence of a vacation effect. Still, we assume that decreased
levels of fatigue on post-vacation may represent a vacation after-effect: during winter sports
vacation, people engage in physically demanding, uncommon activities which are presumably
accompanied by feeling physically fatigued, while after work, people may feel primarily mentally
fatigued.
3.4.2. Vacation after-effects
The results regarding vacation after-effects were less favourable for H&W: the five positive
vacation effects had vanished within the first week of work-resumption. Fatigue constituted the
only exception to this rule and was lowest immediately after vacation. Despite the absence of
a vacation effect in fatigue, this finding is in line with the slower fade-out process in burnout
that Westman and Eden (1997) reported and may point to positive mid-term effects regarding
fatigue.
Due to the absence of on-vacation measurement occasions, most previous vacation
studies defined a vacation effect as the difference between the pre-vacation and post-vacation
levels in H&W that ‘sandwiched’ the vacation period. Whereas the meta-analysis of De Bloom
et al. (2009) revealed a small short term after-effect, we found none in the current study. There
are several possible explanations for the immediate fade-out of vacation effects that need to
be discussed.
Firstly, could it be that the type of vacation is important for the duration of the vacation
effects? One might argue that a winter sports vacation as a very active type of vacation may
have less enduring beneficial health effects than for example a predominantly relaxing vacation.
However, research has demonstrated that active leisure activities, in particular physical
activities, improve well-being and may be even more recovering than low-effort activities
like watching television (Sonnentag, 2001; Sonnentag & Natter, 2004; Rook & Zijlstra, 2006).
Accordingly, it is not very likely that the active character of a winter sports vacation explains the
lack of after-effects.
Secondly, an explanation may be that a winter sports vacation normally forms an
interruption of a busy period of the year. Vacationers return home and are immediately trapped
in demanding daily routines and hassles like unpacking and washing clothes, work, and non-
work-obligations. Research on spa therapy suggests that returning home in the second half of
a workweek with the weekend in prospect is more favourable for the conservation of positive
effects than returning on Sunday with a full working week ahead (Strauss-Blasche, Muhry,
Lehofer, Moser & Marktl, 2004). Therefore, it would be interesting to examine in future studies
whether short vacations (active or passive) scheduled at a more relaxed time of the year (e.g.
76
Chapter 3
3
during a long summer vacation) or in a different manner (e.g. one or two more days off after
returning home to prevent “post-vacation stress”) may have more enduring after-effects.
Thirdly, the duration of the vacation period may constitute a major component of its
effectiveness in improving H&W during and after vacation. Just as a lower dose of medicine
may be less effective in curing a disease, a short vacation may have fewer and less profound
effects on H&W than a long vacation period. A winter sports vacation is typically a short
vacation: most of our participants spent only nine days away from home (including two travel
days) and one week away from work. As a consequence of the brief ‘treatment’, the effects
may have been weaker and more short-lived.
Fourthly, it may also be that in previous studies the after-effects of vacation have
been overestimated. If the pre-vacation occasion is programmed immediately before vacation,
it may be confounded by preparation stress for the vacation which is likely to be associated
with decreased levels of H&W. When this pre-vacation occasion is subsequently treated as
baseline, vacation after-effects would artificially increase.
Regarding the rapid fade-out process of positive vacation effects, an intriguing
question may be: Why should we go on vacation at all when effects wash out so fast? However,
like any other freely chosen and pleasant activity, a vacation is a period that people enjoy for
its own sake; vacation makes people happy and healthy as our study unmistakably showed. A
vacation is, therefore, an effective, strong and natural way to boost the well-being of employees.
Furthermore, H&W could deteriorate over time if people would not go on vacation,
as vacation is important for long term health and vitality, and for building up enduring personal
resources and coping capacities. A study of Gump and Matthews (2000), for example,
showed that not taking annual vacations was associated with a higher risk of morbidity and
mortality during a nine-year period. Correspondingly, a more appropriate question regarding
the temporal nature of vacation effects would probably be: Is it possible to conserve positive
vacation effects, and if so, which strategies can be used to slow down fade-out processes and
prolong vacation relief (see also Eden, 2001)?
3.4.3. Limitations
The limited variation in vacation type and duration was a deliberate choice in the current study.
The uniformity with respect to activities, duration and time off the job before and after vacation
(maximally 1 or 2 days) enabled us to generate reliable results for short winter sports vacations.
However, the question remains whether we would have found the same pattern of results for
other vacation types, for other vacations durations, and for other periods (seasons) of the year.
In addition, our sample of skiing enthusiasts may limit the external validity of our
study. Although our sample was heterogeneous in many regards (gender, age, type of work,
Effects of Vacation on Health and Well-being
77
3
family background), winter sports vacationers may be above-average healthy, active and
sporty. Even though we do not have theoretical reasons to assume that vacations will have less
positive effects among less healthy and sporty individuals, we should be careful in generalizing
our findings.
Another limitation is the use of self-reports only. However, H&W are by definition
subjective constructs and self-reports are probably the best way to measure them (Kompier,
2005). But one may also argue that retrospective evening scores may be biased by cognitive
distortions like the “rosy view bias”. Mitchell, Thompson, Peterson and Cronk (1997) found
that people’s post-event recollections are more positive than their evaluations of the actual
experiences. Yet, we reduced such potential biases by measuring several times a week and by
asking respondents to indicate their level of H&W on the same day.
We measured fatigue with a single-item measure because it reduced the burden put
on the participants, prevented non-response and attrition and because it is a valid substitute
for multiple item measures of fatigue (Van Hooff et al., 2007). In spite of that, the use of two
additional single-item measures on mental and physical fatigue could have provided more in-
depth information and understanding of the vacation (after-) effects regarding fatigue.
Finally, there may be an effect of the time of the day at which the pre- and post-
vacation measures (just before going to bed) and the on-vacation measures (between 5 and
7 pm) were taken. It may be that people feel better in the early evening than just before going
to bed because of feeling more tired at bedtime. Nevertheless, fatigue was highest during
vacation, in the early evening, which does not point into the direction of a “before bedtime
effect”.
3.4.4. Suggestions for future vacation research
First and foremost, future vacation research could be optimized by applying research designs
like the one we used with repeated measures before, during and after vacation. Furthermore,
the combination of different technically innovative instruments for data collection (digital diaries,
telephone surveys) and an extensive protocol to guarantee compliance (careful recruitment,
SMS reminders) may help future researchers to start measuring on vacation and to prevent
attrition.
Data triangulation, for example the combination of self-reports, ratings from the
partner or fellow vacationers and performance ratings, would be a means to further improve
vacation research and to generate valid and reliable results.
Some other suggestions for future vacation research regarding sleep quality (i.e. take
physical sleep circumstances into account) and fatigue (i.e. distinguish mental and physical
fatigue) are important and were already briefly mentioned above.
78
Chapter 3
3
Because different types of vacation (active and passive) may have different effects
on H&W, the impact of various vacation types on the strengths and duration of vacation effects
should be investigated (see also Eden, 2001). For instance, would a relatively short relaxing
vacation during the winter period have the same vacation effects and (lack of) after-effects as
an active winter sports vacation? The impact of similar types of vacation (e.g. physically active
vacations) scheduled in different seasons of the year could be examined as well. Would, for
instance, an active vacation in the summer (e.g. sailing or biking) have the same vacation and
after-effects as an active vacation in the winter?
The role of vacation duration is difficult to study, because if duration varies, a lot
of other variables such as vacation type and activities co-vary. As a consequence, it will be
impossible to attribute vacation effects and after-effects mainly to its duration. It does for
example not make sense to compare vacation effects of a 4-week backpacker-trip through
Scandinavia with a 2-week all-inclusive resort stay at Costa del Sol. Also experimentally
assigning participants to different vacation durations is practically impossible (for creative
ideas like give-away paid vacations see Eden, 1990). So, the best way to study the effects of
vacation duration is probably to vary vacation duration while holding vacation type as constant
as possible.
Another interesting research topic is the investigation of the role of work accumulation
as moderator of vacation (after-) effects. For some employees, work may pile up before
vacation (see also Westman, 2004, 2005; DeFrank, Konopaske & Ivancevich, 2000); they have
to work harder in order to go on vacation and experience “working-ahead stress”. On vacation,
their work may accumulate even further and they may be confronted with high workload
after returning home (Fritz & Sonnentag, 2006). We may call this “catch-up stress”. For other
employees, work may be structured in a different way and may not pile up because a colleague
takes over. Accordingly, it would be interesting to include measures of “working ahead-stress”
before and “catch-up stress” after vacation and to study their impact on H&W.
A target for vacation researchers could also be the investigation of the role of vacation
activities and experiences in changing H&W. Up till now, vacation remains an intervention
with more or less unknown content and we do not know if vacation activities like physical
activities, relaxing, household or work-related tasks have a different impact on the strength of
the vacation effect or the fade-out rate (for an exception see Fritz & Sonnentag, 2006). Vacation
expectations and their fulfilment, uplifts and hassles and relations with travel companions and
the life partner during vacation are additional examples for possible moderators of the vacation
effect which should be studied (see also Eden, 2001).
Last but not least, strategies to slow down fade-out processes and to prolong vacation
relief are an important avenue for future research. Positive, frequent vacation reflection may
Effects of Vacation on Health and Well-being
79
3
be a prime candidate for fade-out deceleration because reflecting repeatedly and favourably
on pleasant vacation experiences may reactivate positive vacation cognitions and feelings,
and enhance H&W. In an experiment on cardiovascular reactivity (Fredrickson, Mancuso,
Branigan & Tugade, 2000), positive emotions speeded up cardiovascular recovery from stress,
indicating that positive emotions regulate or even undo negative emotional arousal. These
findings support the assumption from Broaden- and -Build Theory (Fredrickson, 2001) that
positive emotions may improve individual’s coping capacity to deal with stressors. So, positive
emotions experienced during vacation and positive vacation reflection may protect and build
resources that improve H&W by buffering future threats.
In conclusion, it seems that a winter sports vacation certainly improves H&W, but positive
effects are short-lived. Future vacation studies should therefore focus on means to decelerate
the fade-out process in order to prolong vacation relief. Moreover, we propose a longitudinal
framework for vacation research with proper baseline-, on-vacation- and multiple post-vacation
measurements (such as in the framework that we employed) to investigate the effects of
different vacation types, durations, activities and experiences on H&W in future vacation
studies.
80
Chapter 3
3
3.5. References
Akerstedt, T. (2006). Psychosocial stress and impaired sleep. Scandinavian Journal of Work, Environment
& Health, 32, 493-501.
Belkic, K., Landbergis, P.A, Schnall, P.L., & Baker, D. (2004). Is job strain a major source of cardiovascular
disease risk? Scandinavian Journal of Work, Environment & Health, 30, 85-128.
Brosschot, J.F., Van Dijk, E., & Thayer, J.F. (2007). Daily worry is related to low heart rate variability during
waking and the subsequent nocturnal sleep period. International Journal of Psychophysiology,
63, 39-47.
Cohen, J. (1988). Statistical power analysis for the behavioral sciences. Hillsdale (NJ): Lawrence Erlbaum
Associates.
Dahlgren, A., Kecklund, G., & Akerstedt, T. (2005). Different levels of work-related stress and the effects
on sleep, fatigue and cortisol. Scandinavian Journal of Work, Environment & Health, 31, 277-
85.
De Bloom, J., Kompier, M., Geurts, S., De Weerth, C., Taris, T., & Sonnentag, S. (2009). Do we recover
from vacation? Meta-analysis of vacation effects on health and well-being. Journal of
Occupational Health, 51, 13-25.
DeFrank, R.S., Konopaske, R., & Ivancevich, J.M. (2000). Executive travel stress: Perils of the road
warrior. The Academy of Management Executive, 14, 58-71.
De Lange, A.H., Taris, T.W., Kompier, M.A.J., Houtman, I.L.D., & Bongers, P.M. (2003). The very best of
the millennium: Longitudinal research and the Demand-Control-(Support) model. Journal of
Occupational Health Psychology, 8, 282-305.
DeShon, R.P., & Morris, S.B. (2003). Modeling complex data structures: The General Linear Model
and beyond. In S.G. Rogelberg (Ed.), Handbook of Research Methods in Industrial and
Organizational Psychology (pp. 390-411). Malden: Blackwell Publishers.
Eden, D. (1990). Acute and chronic job stress, strain, and vacation relief. Organizational Behavior and
Human Decision Processes, 45, 175-199.
Eden, D. (2001). Vacations and other respites: studying stress on and off the job. In C. Cooper and I.T.
Robertson (Eds), Well-being in organizations (pp.305-330). West Sussex (UK): John Wiley &
Sons.
Elo, A., Leppänen, A., & Jahkola, A. (2003). Validity of a single-item measure of stress symptoms.
Scandinavian Journal of Work, Environment & Health, 29, 444-451.
Etzion, D. (2003). Annual vacation: duration of relief from job stressors and burnout. Anxiety, Stress and
Coping, 16, 213-226.
Ferrie, J.E., Westerlund, H., Virtanen, M., Vahtera, J., & Kivimaki, M. (2008). Flexible labour markets and
employee’ health. Scandinavian Journal of Work, Environment & Health, 6, 98-110.
Fredrickson, B.L. (2001). The role of positive emotions in positive psychology: The Broaden- and -Build
Theory of positive emotions. American Psychologist, 56, 218-226.
Effects of Vacation on Health and Well-being
81
3
Fredrickson, B. L., Mancuso, R. A., Branigan, C., & Tugade, M. M. (2000). The undoing effect of positive
emotions. Motivation and Emotion, 24, 237 - 258.
Fritz C, & Sonnentag, S. (2005). Recovery, health, and job performance: Effects of weekend experiences.
Journal of Occupational Health Psychology, 10, 187-99.
Fritz C, & Sonnentag S. (2006). Recovery, well-being, and performance-related outcomes: the role of
workload and vacation experiences. Journal of Applied Psychology, 91, 936-945.
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 & Health, 32, 482-492.
Gump, B., B., & Matthews, K.A. (2000). Are vacations good for your health? The 9-year mortality
experience after the multiple risk factor intervention trial. Psychosomatic Medicine, 62, 608-
612.
Hjortskov, N., Rissen, D., Blangsted, A.K., Fallentin, N., Lundberg, U., & Sogaard, K. (2004). The effect
of mental stress on heart rate variability and blood pressure during computer work. European
Journal of Applied Physiology, 92, 84-89.
Kompier, M. (2005). Assessing the psychosocial work environment- ‘subjective’ versus ‘objective’
measurement. Scandinavian Journal of Work, Environment & Health, 31, 405-408.
McEwen, B.S. (1998). Stress, adaptation, and disease: allostatis and allostatic load. Annals of the New
York Academy of Science, 840, 33-44.
Meijman, T.F., & Mulder, G. (1998). Psychological aspects of workload. In P.J.D. Drenth, H. Thierry and
C.J. de Wolff (Eds), Handbook of work and organizational psychology (2nd edition). Work
psychology (Vol. 2, pp. 5-33). Hove: Psychology Press.
Meyers, A.R., Perrine, M.W., & Caetano, R. (1997). Alcohol use and downhill ski injuries: A pilot study. In
R.J. Johnson, C.D. Mote, Jr. and A. Ekeland (Eds), Skiing trauma and safety (Vol.11, pp. 14-
22). West Conshohocken (PA): American Society for Testing and Materials.
Mitchell, T. R., Thompson, L., Peterson, E., & Cronk, R. (1997). Temporal adjustment in the evaluation of
events: The “Rosy View”. Journal of Experimental Social Psychology, 33, 421-448.
Roehrs, T., & Roth, T. (2001). Sleep, sleepiness, sleep disorders and alcohol use and abuse. Sleep
Medicine Reviews, 5, 287-297.
Rook, J. W., & Zijlstra, F. R. H. (2006). The contribution of various types of activities to recovery. European
Journal of Work and Organizational Psychology, 15, 218 - 240.
Sluiter, J.K., Frings-Dresen, M.H., Meijman, T.F., & Van der Beek, A.J. (2000). Reactivity and recovery
from different types of work measured by catecholamines and cortisol: a systematic literature
review. Occupational Environmental Medicine, 57, 289-315.
Sonnentag, S. (2001). Work, recovery activities, and individual well-being: A diary study. Journal of
Occupational Health Psychology, 6, 196-210.
Sonnentag, S., & Natter, E. (2004). Flight attendants’ daily recovery from work: Is there no place like
home? International Journal of Stress Management, 11, 366-391.
82
Chapter 3
3
Strauss-Blasche, G., Muhry, F., Lehofer, M., Moser, M., & Marktl, W. (2004). Time course of well-being
after a three-week resort-based respite from occupational and domestic demands: Carry-
over, contrast and situation effects. Journal of Leisure Research, 36, 293-309.
Taris, T.W. (2000). A primer in longitudinal data analysis. London: Sage.
Van Hooff, M.L.M., Geurts, S.A.E., Kompier, M.A.J., & Taris, T.W. (2007). Workdays, in-between workdays,
and the weekend: A diary study on effort and recovery. International Archives of Occupational
and Environmental Health, 80, 599-613.
Van Hooff, M.L.M., Geurts, S.A.E., Kompier, M.A.J., Taris, T.W., Houtman, I.L.D., & Van den Heuvel, F.
(2005). Disentangling the causal relationships between work-home interference and employee
health. Scandinavian Journal of Work, Environment & Health, 31, 15-29.
Van Hooff, M.L.M., Geurts, S.A.E., Taris, T.W., & Kompier, M.A.J. (2007).”How fatigued do you currently
feel?” Convergent and discriminant validity of a single-item fatigue measure. Journal of
Occupational Health, 49, 224-34.
Westman, M. (2004). Strategies for coping with business trips: A qualitative exploratory study. International
Journal of Stress Management, 11, 167-176.
Westman, M. (2005). The impact of short business travels on the individual, the family and the
organization. In A. Antonio, & C. Cooper (Eds.), Research companion to organizational health
psychology (pp. 478-491). Cheltenham, England: Edward Elgar Publishing.
Westman, M., & Eden, D. (1997). Effects of a respite from work on burnout: vacation relief and fade-out.
Journal of Applied Psychology, 82, 516-527.
Westman, M., & Etzion, D. (2001) The impact of vacation and job stress on burnout and absenteeism.
Psychology & Health, 16, 595-606.
85
Chapter 4
How Does a Vacation from Work Affect
Employee’ Health and Well-being?
H&W improve during vacation. However, it is unclear whether this general development
applies to all employees, while also little is known about the underlying processes causing
such an improvement. Our research questions were: 1) Does every worker experience
a positive effect of vacation on H&W?; and 2) Can vacation activities and experiences
explain changes in H&W during vacation?
In a 7-week longitudinal field study, 96 workers reported their H&W 2 weeks before, during,
1 week, 2 and 4 weeks after a winter sports vacation on 6 indicators (health status, mood,
fatigue, tension, energy level, satisfaction).
Sixty percent of the sample experienced substantial improvement of H&W during and after
vacation. Yet, a small group experienced no (23%) or a negative effect of vacation (17%).
Spending limited time on passive activities, pleasure derived from vacation activities, and
the absence of negative incidents during vacation explained 38% of the variance in the
vacation effect.
Although vacation has a positive, longer lasting effect for many, it is not invariably positive
for all employees. Choosing especially pleasant vacation activities and avoiding negative
incidents as well as passive activities during active vacations apparently contribute to the
positive effect of vacation on H&W.
This chapter is based on:
De Bloom, J., Geurts, S.A.E., Sonnentag, S., Taris, T., De Weerth, C. & Kompier, M.A.J.
(2011). How does a vacation from work affect employee’ health and well-being? Psychology
& Health, 26, 1606-1622.
How Does Vacation Affect Health and Well-being?
87
4
4.1. Introduction
Research in occupational health psychology has consistently demonstrated the adverse impact
of stress in the workplace on individuals’ H&W (e.g. Belkic, Landsbergis, Schnall & Baker,
2004; Hjortskov et al., 2004). Meanwhile, the great importance of recovery during nonwork time
to protect workers against the negative effects of job stressors is increasingly acknowledged
(Geurts & Sonnentag, 2006; Sonnentag, Perrewe & Ganster, 2009; Van Hooff, Geurts, Taris,
Kompier, Houtman & Van den Heuvel, 2005).
Vacation as a long and relatively uninterrupted period of absence from work is a
prime candidate for helping employees to recover more completely from work than during
shorter respite intervals like evening hours or weekends (e.g. Eden, 2001; Etzion, 2003). Earlier
vacation studies demonstrated a positive effect of a vacation from work, i.e., workers’ H&W
substantially improved during a vacation compared to work periods before vacation (e.g. De
Bloom, Kompier, Geurts, De Weerth, Taris & Sonnentag, 2009; Kühnel & Sonnentag, 2011;
Westman & Etzion, 2001). Indeed, a longitudinal study by Gump and Matthews (2000) even
showed that not taking annual vacations was associated with a higher risk of mortality during
a nine-year period.
4.1.1. Differential vacation effects
In spite of the general belief that vacation is beneficial for recovery from work in general (e.g. De
Bloom et al., 2009; Hoopes & Lounsbury, 1989), as yet it is unclear whether favorable effects
of vacation apply to all vacationers. It may well be that relatively large differences between
vacationers in terms of the direction and strength of the vacation effect underlie an on average
positive effect of vacation on H&W. Accordingly, it makes sense to investigate whether or not
subgroups exist that differ in the direction of the vacation effect (positive, neutral, or even
negative) and in the strength of the vacation (after-) effect. Therefore, our first research question
is:
RQ1: Does every worker experience a positive effect of vacation on H&W?
We defined the vacation effect as the difference between baseline and the on-vacation level
of H&W, because this comparison represents the most direct and “pure” effect of a vacation.
This effect is comparable to what Westman and Eden (1997) labeled “Immediate Respite”.
Measuring H&W ‘only’ before and after vacation and attributing post-vacation changes in
H&W to the vacation period would be insufficient, because sequential occurrence does not
necessarily mean that there is also a causal relationship between variables (Eden, 2001).
Moreover, post-vacation measurements may be biased by work resumption. However, it is also
88
Chapter 4
4
very important to investigate the after-effects of vacation in order to study the duration of the
vacation effect. Therefore, we also compared post-vacation to baseline levels of H&W.
4.1.2. Processes underlying the vacation effect
De Bloom et al.‘s (2009) meta-analysis revealed that- due to the difficulty of obtaining data while
respondents are on vacation- vacation activities and experiences as possible determinants of
the vacation effect received little attention in earlier research. Regarding vacation experiences,
only five studies ever collected information on issues like vacation satisfaction (Etzion, 2003;
Lounsbury & Hoopes, 1986; Westman & Eden, 1997), work reflection (Etzion, 2003; Fritz &
Sonnentag, 2006), vacation hassles and recovery experiences (Fritz & Sonnentag, 2006).
However, the findings were inconclusive (see also De Bloom et al., 2009), many experiences
have not been studied yet (e.g. pleasure derived from activities) and, even more importantly,
the information on vacation experiences from most earlier studies was potentially biased,
because data were collected after returning home (i.e., after work resumption).
Information on vacation activities is even scarcer: only in two studies (Lounsbury
& Hoopes, 1986; Strauss-Blasche, Ekmekcioglu & Marktl, 2000) vacationers were asked,
again retrospectively, what they did during vacation. This information was not linked to H&W
outcomes. Accordingly, the role of vacation activities and experiences as possible determinants
of vacation effects remains unclear (De Bloom et al., 2009). Therefore, our second research
question is:
RQ2: Can vacation activities and experiences explain changes in H&W during vacation?
4.1.3. Vacation activities
Research on leisure activities suggests that the activities people engage in during non-working
time influence their level of H&W (e.g. Rook & Zijlstra, 2006; Sonnentag, 2001). Therefore, we
will discuss different types of vacation activities as potential determinants of the effects of a
vacation.
It is often assumed that the temporary absence from work demands in itself already
leads to an improvement of H&W during a vacation period. Consequently, we hypothesize:
H1: Workers’ H&W will be higher during vacation than during working periods.
This reasoning also implies that work-related activities during free-time should have a negative
effect on H&W. Indeed, there is ample evidence that prolonged exposure to work demands has
adverse effects on H&W (Demerouti, Bakker, Geurts & Taris, 2009; Van Hooff, Geurts, Kompier
& Taris, 2007). Working during vacation does not only limit directly one’s potential recovery time,
but also puts a demand on the same psychophysiological systems that were activated during
How Does Vacation Affect Health and Well-being?
89
4
work periods, potentially obstructing the crucial process of ‘psychophysiological unwinding’
(Geurts & Sonnentag, 2006). Therefore, we hypothesize:
H2: The increase in H&W during vacation will be smaller for employees who spend more
time on work-related activities during vacation.
Furthermore, it is well-established that physical exercise has beneficial effects on physical
health, mood, and psychological recovery indicators (Demerouti et al., 2009; Reed & Ones,
2006; Rook & Zijlstra, 2006). This effect may be explained both by psychological and
physiological mechanisms (Mead et al., 2009; Nabkasorn et al., 2005). Exercise may distract
from unpleasant stimuli, daily hassles and job-related duties, may encourage positive feelings
about oneself, and enhance the secretion of neurotransmitters with an antidepressant effect
(Hansen, Stevens & Coast, 2001; Moran, 2004; Sonnentag & Jelden, 2009). Accordingly, we
hypothesize:
H3: The increase in H&W during vacation will be larger for employees who spend more
time on physical activities during vacation.
The need to be with others and the desire to engage in social activities is considered to be
an inborn, evolutionary adaptive, fundamental human need (e.g. Baumeister & Leary, 1995).
Research showed that social activities and social support improve H&W in humans ranging
from childhood to older adults (e.g. Dormann & Zapf, 1999; Hale, Hannum & Espelage,
2005). Social support may function as a stress buffer, for instance, by lowering cardiovascular
reactivity to psychosocial stress (e.g. Gerin, Pieper, Levy & Pickering, 1992). We hypothesize:
H4: The increase in H&W during vacation will be larger for employees who spend more
time on social activities during vacation.
Sonnentag (2001) argued that low-effort (or passive) activities put no demands on the individual.
Therefore, she attributed recovery-potential to these activities and her results supported this
assumption. However, Rook and Zijlstra (2006) found in their study on recovery during a normal
work week that low-effort activities were non-beneficial for recovery after work. So, the findings
regarding low-effort activities are inconclusive. Therefore, we will not generate a hypothesis
regarding the role of passive activities in our study.
4.1.4. Vacation experiences
A vacation period may well add pleasure to peoples’ lives and may help to build up enduring
personal resources, because people are free to engage in self-chosen activities which will
lead to positive emotions like joy and freedom. Correspondingly, the engagement in activities
experienced as pleasant may boost the vacation effect on H&W and we hypothesize:
H5: The increase in H&W during vacation will be larger for employees who report higher
levels of pleasure derived from their vacation activities.
90
Chapter 4
4
Another important type of vacation experiences are negative incidents. During vacation sad,
bothersome or irritating things may happen (e.g. illnesses, accidents, conflicts), like at any
moment in time. However, when expectations for pleasure and fun are high, such as during
vacation, these incidents will presumably have a particularly strong negative effect on H&W.
Negative incidents can be considered stressors that undermine the recovery process and that
may lead to lower levels of H&W (Fritz & Sonnentag, 2006). According to Ryan and Deci’s
(2000) Self-Determination Theory, negative incidents like getting ill or injured can be considered
a threat to self-determination and may limit the vacationer’s ability to engage in an activity of
his/her own choice. Therefore, we hypothesize:
H6: The increase in H&W during vacation will be smaller for employees who experience
negative incidents during vacation.
4.2. Method
4.2.1. Data collection procedure and design
Our two research questions and our six hypotheses (question 1: hypothesis 1; question 2:
hypotheses 2-6) were addressed in a longitudinal field study on winter sports vacations. This
study covered a time span of seven weeks, including the winter sports vacation itself that took
place between February 15
th
and April 15
th
2008. Two weeks before vacation, a paper-and-
pencil questionnaire assessing demographic information and basic job information was sent to
the participants, to be returned in a postage-paid pre-addressed envelope that was attached
to the questionnaire. Participants also received a card with an overview of their personal
measurement occasions during the seven-week period. In order to encourage participation
and adherence to the research protocol, we announced a lottery prize among all participants
(a winter sports holiday for the next winter sports season). Chances of winning were higher for
participants who returned all questionnaires than for participants who missed measurement
occasions.
Subsequently, the participants received an e-mail with a link to a digital diary twice a
week on all measurement occasions during working periods i.e., two weeks before vacation
(Pre), and the first (Post 1), second (Post 2) and fourth week (Post 3) after returning home.
These digital diaries had to be completed just before bedtime on a fulltime working day. To
ensure that participants completed the digital diaries in the evening, they additionally received
a reminder-SMS on their cell phone earlier that day.
The participants were also provided with cell phones from the university with
international pre-paid SIM cards to take with them on holiday in order to collect on-vacation
How Does Vacation Affect Health and Well-being?
91
4
measures of H&W. Every participant received 10 Euro as pre-paid talk credit for their vacation-
phone. Each participant was called on this phone and interviewed by one of the researchers
on the second day after arrival and on the second-last day before departure between five and
seven pm (Inter).
When the full cycle of data collection (i.e., 10 measurements per participant) had
been completed, respondents were thanked for their participation, received information about
when the results would be available and about when the winner for the lottery prize was drawn.
The cell phones were returned to the researchers after the data collection phase had been
completed.
4.2.2. Participants
To recruit participants, we distributed information via travel agencies, winter sports websites,
shops for skiing-equipment, winter sports journals and newspaper ads. Additionally, we visited
a winter sports fair and contacted ski-clubs.
Initially, 176 Dutch persons voiced that they were interested in taking part in this
study. After administering detailed information about the research procedure and promising
confidentiality, these 176 persons received a phone call from one of the researchers. During
this call, the participants were screened on their eligibility for the study: participants (i) had to
work at least 24 hours per week (18 exclusions), (ii) go on winter sports vacation for at least
one week between February 15th and April 15th, 2008 (22 exclusions), and (iii) should enrol
on time (17 exclusions). Persons working extremely irregular schedules were also excluded
(4 exclusions). Occasionally, persons did not appreciate to be called during vacation (4
exclusions), did not use electronic mail (5 exclusions) or found the research procedure too
burdensome (3 exclusions). Another seven persons were excluded because they did not go on
vacation after all due to illness. Of the 176 interested persons, 108 met the inclusion criteria. Of
this group of 108 potential participants, 96 persons actually took part in the study.
The majority of this sample was male (65%), the mean age was 44 years (SD = 10
years) and 55% of the sample held a college or university degree, 40% were medium (senior
general secondary and university preparation education) and 5% lower educated (no, lower
secondary or junior secondary education only).
The largest part of the respondents was employed (82%) whilst 18% were self-
employed. The participants worked in a variety of vocations: 23% worked in the commercial
sector, 20% were high-educated specialists (e.g. engineers, ICT-workers), 14% worked in the
service sector, 12% in health care, 11% were administrative employees, and the remaining 20%
worked in other sectors.
92
Chapter 4
4
The participants worked 38 hours per week on average (SD = 8 hours) with the total
number of weekly work hours (including overtime) varying from 24 to 60 hours. Forty-seven
percent of the participants supervised other persons; the remaining 53% had no supervisory
tasks.
The average vacation duration was 9 days (SD = 2 days, range: 7-19 days) and
vacation destinations were typical winter sports areas, with the top-three destinations being
Austria (70%), France (15%) and Switzerland (6%).
4.2.3. Measures
Health and well-being. In order to present a comprehensive account of H&W, we
incorporated six main indicators of H&W: health status, mood, fatigue, tension, energy level and
satisfaction. Single-item measures were used to tap these concepts. In this way we attempted
to minimize the effort required from the participants and to maximize user-friendliness, which
was supposed to have beneficial effects on response rates. Single-item measures often have
a high face-validity and participants value their directness and lack of redundant and repeated
comparable items (e.g. Elo, Leppänen & Jahkola, 2003). Accordingly, multiple item measures
may be validly replaced by single-item measures and still be psychometrically acceptable if the
underlying constructs are sufficiently one-dimensional and unambiguous to the participants
(e.g. Van Hooff, Geurts, Taris & Kompier, 2007). For simplicity, we adapted response-scales
based on the well-known basic Dutch grade notation system ranging from 1 (extremely low/
negative) to 10 (extremely high/positive) and anchored the first and the last grade.
We measured health status on each measurement occasion by the single-item
measure: “How healthy did you feel today?” (1 = “very unhealthy”, 10 = ”very healthy”). Mood
was measured with the item: “How was your mood today?” (1 = ”very bad”, 10 = ”very good”).
Levels of fatigue were assessed with the measure: “How fatigued did you feel today?” (1 =
“not fatigued at all”, 10 = ”very fatigued”). We measured tension with a single-item worded:
“How tense did you feel today?” (1 = ”very calm”, 10 = “very tense”). In addition, we asked
respondents to indicate how energetic they felt (“How energetic do you currently feel?” (1 =
“absolutely not energetic”, 10 = ”very energetic”). Finally, respondents were asked to indicate
day satisfaction on a single-item measure: “How satisfied do you feel about this day?” by
means of a report mark ranging from 1 (“very dissatisfied”) to 10 (“very satisfied”).
We included the six H&W indicators in an exploratory factor analysis to find out
whether one underlying construct existed. This factor analysis indeed resulted in a one-factor
solution with an Eigenvalue greater than 1 and factor loadings .50. Therefore, we combined
the six H&W indicators into one overall H&W construct. Cronbach’s α of H&W was .84 (Pre),
.78 (Inter), .81 (Post 1), .82 (Post 2) and .78 (Post 3).
How Does Vacation Affect Health and Well-being?
93
4
Vacation activities. For each of the four vacation activities (work-related, physical,
social, passive), participants indicated the amount of time they had devoted to it during the day
they were interviewed. We also gave at least two examples for each activity to help vacationers
categorize their activities: checking work mail or having a phone call with the office (work-
related), skiing or walking (physical), après ski or playing games (social), and reading a novel
or watching television (passive).
Pleasure derived from vacation activities. We also measured the quality of the
engagement in each activity by asking participants to rate the pleasure they experienced
while executing it. An example item is: “Please indicate how pleasant you experienced the
physical activities you carried out today” (1 =”very unpleasant”, 10 = ”very pleasant”). Not
every vacationer carried out every single activity. Therefore, in order to get an overall score of
the pleasure derived from vacation activities for each participant, we averaged the pleasure
scores across the activities the vacationers did participate in
Negative incidents during vacation. Negative incidents were measured with two
questions: 1) Did you experience something very unpleasant today?; and 2) Did you experience
something very unpleasant within the previous vacation days? Participants responded
dichotomously with Yes or No. We divided the vacationers into two groups: one group that
experienced at least one negative incident during vacation and one group that experienced no
negative incidents during vacation. Using an open question, we also assessed the nature of
the negative incident (“Could you give a short indication of the nature of the negative incident
you experienced?”).
4.2.4. Missing data: prevention and treatment
In order to prevent and deal with missing data, we scheduled two measurement occasions
within each week. As a series of 5 Student’s t-tests demonstrated no significant differences
between the first and second measurement of each week, the two within-week measures of a
particular H&W indicator were averaged to obtain a reliable week-level indicator. In case of a
non-answered prompt during a workweek, the other measurement in that week was treated as
the week-average.
With the exception of the general questionnaire, we used electronic mail and SMS
to remind the participants to fill in the questionnaires at the correct moment in time. We could
detect protocol deviations very rapidly (i.e., the next morning) due to these digital diaries. A
detailed non-completion script was applied for the digital diaries as well as the telephone
surveys. This script included a reminder e-mail, a second SMS and finally a phone call on
the participants’ own cell phone to solve possible problems with the protocol, to elucidate
94
Chapter 4
4
ambiguities and to explain the importance of duly reactions and compliance to the research
procedure.
The general questionnaire was returned by the full sample. Based on a maximum
of 960 possible single diary measurements in this study (10 measurements in 96 persons),
the overall completion rate was 87% (834 measurements). The combination of the ten
measurements (two measurements a week) into five occasions resulted in even more reliable
week-indicators and high completion rates: 100% (N = 96) on Pre, 98% (N = 94) on Inter, 90%
(N = 86) on Post 1 and 96% (N = 92) on Post 2 and Post 3. For 83 of the 96 persons, data sets
were complete (no missing data on any of the five occasions).
4.2.5. Statistical analyses
To answer research question 1, and to test Hypothesis 1, we analyzed the data in an analysis
of variance (ANOVA) with repeated measures on the five occasions before, during and after
vacation. Post-hoc Fisher’s Least Significant Difference (LSD) tests were used to detect
variations in H&W across this seven week period: the vacation effect was tested by comparing
the pre-vacation measures of the H&W indicator with the measures taken during vacation (Pre
versus Inter). Vacation after-effects were examined by conducting LSD’s for the difference
between H&W on Pre versus Post 1, Post 2 and Post 3 respectively.
For all significant differences between measurement occasions (Pre versus Inter,
Post 1, Post 1, and Post 3, representing the vacation effect and its after-effects respectively),
we present Cohen’s d for paired observations (Cohen, 1988, p.46). A negative effect size
indicates decreases in H&W compared to the pre-vacation level. Following Cohen (1988) we
distinguished small (0 to 0.5), medium (0.5 to 0.8) and large (> 0.8) effects.
To test our five hypotheses related to research question 2, we conducted a hierarchical
regression analysis with H&W during vacation (Inter) as the dependent variable. Because we
defined a vacation effect as a change in H&W during vacation compared to before vacation
(Pre), we entered baseline levels of H&W (i.e., H&W two weeks before vacation) as the first
variable in the regression equation. In the second step, we entered age and gender as control
variables.
Because we assumed that the passive mechanism of relief from work is the most
basic process underlying the vacation effect, and working would lead to smaller increases
in H&W during vacation, we entered time spent on work-related activities during vacation as
a predictor in the third step. In the fourth step, we entered time spent on physical, social,
and passive activities into the regression analysis. In Step 5 we added pleasure derived
from vacation activities. In the sixth and final step, we added negative incidents as possible
determinants of H&W during vacation.
How Does Vacation Affect Health and Well-being?
95
4
To ensure that the effects were relatively independent from the order of entry of the
independent variables, we repeated the analysis in several different orders. The results were
very similar (and can be obtained upon request from the first author). Therefore, we only report
the results from the analysis described above for which we also had the most solid theoretical
basis.
4.3. Results
4.3.1. Question 1: Does every worker experience a positive effect of vacation on
H&W? (Hypothesis 1)
Repeated measures ANOVA revealed a main effect across time F (4,79) = 14.06, p < .01,
meaning that H&W levels significantly varied across the seven week time period (pre-inter-
post). Post-hoc LSD tests further showed that H&W levels on Inter differed significantly from
baseline (p < .01). The average change in H&W from Pre (M = 7.0) to Inter (M = 7.7) was
moderate and represented a medium-sized positive effect (d = 0.55). Accordingly, employees
felt better during vacation than two weeks before vacation (Hypothesis 1 supported).
However, the difference scores for Pre and Inter (Inter minus Pre) for individual participants
ranged from – 3.58 to + 4.58 (M = 0.7, SD = 1.3, 95% confidence interval ranged from 0.44
to 0.97, reliability of difference score = 0.74 as calculated using Schulte & Borich’s, 1984,
approach, which exceeds the .70 threshold proposed by Nunally, 1978) indicating that for at
least some vacationers H&W decreased during vacation. Accordingly, the answer to research
question 1 is: No, not every worker experienced a positive effect of vacation on H&W.
In order to analyze the development of H&W in vacationers with different vacation
effects in more detail, we divided the vacationers into three groups. We defined the ‘neutral
vacation effect’ group as vacationers with a difference score (Inter minus Pre) around zero (with
a quarter standard deviation, range from - 0.32 to + 0.32). The ‘positive vacation effect’ group
was composed of vacationers with a difference score in H&W larger than zero (difference
Inter-Pre 0.33). The ‘negative vacation effect’ group consisted of vacationers with a difference
score in H&W smaller than zero (difference Inter-Pre - 0.33).
This subgroup analysis showed that 60% of the respondents experienced a positive
vacation effect (the ‘positive vacation effect’ group, M Pre = 6.5, M Inter = 8.0, d = 2.02). It
also showed that 23% experienced no difference in H&W during vacation compared to before
vacation (the ‘neutral vacation effect’ group, M Pre = 7.4, M Inter = 7.5), and that 17% of
the sample reported lower H&W during vacation compared to before vacation (the ‘negative
vacation effect’ group, M Pre = 8.2, M Inter = 7.0, d = -1.55).
96
Chapter 4
4
We also studied the vacation after-effects, first for the full sample and then also for the
three different subgroups. We just reported that for the full sample, H&W increased. However,
directly after returning home and resuming work (on Post 1), H&W had returned to baseline
levels (M = 7.2), meaning that in general positive effects had immediately faded out and there
was no vacation after-effect. H&W remained on this level on Post 2 and Post 3 (Post-hoc LSD
tests for the differences between Pre and Post 1, Pre and Post 2 and between Pre and Post 3
were non-significant).
Inspection of the development of H&W across time for the three specific vacation
effect groups showed significant differences. For the group with a negative vacation effect, the
level of H&W was also significantly lower on Post 1 (M = 7.7, d = -0.86), Post 2 (M = 7.3, d =
-1.10) and Post 3 (M = 7.3, d = -1.31) compared to the pre-vacation level (M = 8.2).
Conversely, the group with a positive vacation effect experienced also a positive after-
effect: On Post 1, the level of H&W of this group (M = 7.2) was still significantly different from
baseline (M = 6.5, d = 0.59) and even on Post 2, 2 weeks after vacation, H&W surpassed
baseline scores significantly (M = 6.9, d = 0.44). Only 4 weeks after vacation, H&W had
returned to the pre-vacation level (M = 6.7).
The baseline of the neutral group was in between the negative and the positive group
(M = 7.4). After vacation, on Post 1 (M = 7.0) and Post 3 (M = 7.0), H&W of this group was
even significantly lower than upon baseline (d’s were -0.57 and -0.48, respectively).
In sum, 60% of the vacationers did benefit from a vacation in terms of increased levels
of H&W during vacation. In this group, there was also a positive vacation after-effect that lasted
at least 2 weeks after vacation. However, there was also a minority of vacationers (40%) that
experienced no or a negative vacation effect on H&W.
4.3.2. Question 2: Can vacation activities and experiences explain the changes in
H&W during vacation? (Hypotheses 2-6)
The answer to our second research question was based on hierarchical regression analysis.
Zero-order correlations can be found in Table 4.1. The six steps of the hierarchical regression
are summarized in Table 4.2.
In the first step of the regression analysis, we regressed H&W on Pre. H&W on Pre
was positively related to H&W during vacation, meaning that employees who felt well before
vacation also felt well during vacation. In the second step we entered age and gender as control
variables, but these variables did not contribute to the prediction of H&W during vacation.
Work-related vacation activities were entered in the third step, but did not explain
variance beyond H&W on Pre. Accordingly, work-related activities during vacation did not
account for changes in H&W during vacation (Hypothesis 2 not supported).
How Does Vacation Affect Health and Well-being?
97
4
Table 4.1:
Means, Standard Deviations, and Zero–Order Correlations Between Study Variables
Variables 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15
Descriptives
M
7.0 7.7 7.2 7.0 6.9 0.1 4.9 2.0 0.6 6.3 8.5 8.3 7.3 8.1 0.4
SD
1.2 0.9 1.1 1.1 1.0 0.2 1.4 1.5 1.3 1.6 1.0 1.0 1.5 1.0 0.5
N
96 94 86 92 92 94 94 94 94 19 94 86 51 91 94
Possible range 1-10 1-10 1-10 1-10 1-10 0-24 0-24 0-24 0-24 0-10 0-10 0-10 0-10 0-10 0-1
1. H&W Pre
2. H&W Inter .28*
3. H&W Post 1 .45* .51*
4. H&W Post 2 .58* .40* .55*
5. H&W Post 3 .59* .31* .44* .68*
Vacation activities & experiences
6. Time work-related activities .09 .10 .14 .03 -.01
7. Time physical activities .05 .32* .03 .07 .06 .03
8. Time social activities .02 .19 .02 -.10 -.14 -.00 -.20*
9. Time passive activities .14 -.36* -.08 .08 .06 -.04 -.35* .04
10. Pleasure work-related activities .09 .17 .00 .07 -.25 .14 .18* -.24 -.31
11. Pleasure physical activities .10 .44* .24* .17 .16 -.14 .35* .03 -.11 .11
12. Pleasure social activities .19 .45* .20 .26* .15 -.04 .37* .13 -.26* -.01 .60*
13. Pleasure passive activities .14 .36* .18 .03 .08 -.06 .15 .38* -.28* .23 .25 .38*
14. Pleasure from activities .11 .49* .21 .18 .14 -.21* .35* .17 -.23* .61* .76* .74* .79*
15. Negative incidents .10 -.32* -.20 -.06 -.08 -.05 -.32* .01 .23* -.23 -.18 -.08 -.19 -.20
Note. * p < .05. Time = number of hours spent on activity. Negative incidents: 0 = no, 1 = yes.
98
Chapter 4
4
To test whether the time devoted to other vacation activities explained variance in
the vacation effect, we entered time devoted to physical, social and passive activities in the
fourth step. Table 4.2 indicates that engagement in physical, social and passive activities all
contributed to changes in H&W during vacation. The strong effect of time devoted to physical
activities on H&W during vacation fitted the high zero-order correlation between these two
variables, indicating that H&W during vacation improved more strongly when vacationers
spent more time on physical activities (Hypothesis 3 supported). However, the non-significant
zero-order correlation between time devoted to social activities and H&W during vacation
suggested that the significant regression weight should be considered an artifact (Hypothesis
4 not supported).
In step 5, pleasure derived from vacation activities significantly contributed to
improvements in H&W during vacation (Hypothesis 5 supported). The significant relationship
of time devoted to physical activities supporting Hypothesis 3 was no longer significant after
pleasure had been added to the analysis. This finding suggests that pleasure derived from
activities plays a crucial role in the relationship between time devoted to physical activities and
H&W during vacation.
Table 4.2:
Hierarchical Regression of H&W during vacation (Inter) on Vacation Activities and Experiences
Variable
H&W Inter
Step 1 Step 2 Step 3 Step 4 Step 5 Step 6
β β β β β β
Step 1: H&W Pre .28* .27* .27* .28* .23* .25*
Step 2: Control variables
Age -.02 -.01 .01 -.01 .01
Gender -.04 -.03 -.09 -09 -.05
Step 3: Time work-related activities .07 .05 .13 .12
Step 4: Time spent on activity (Inter)
Time physical activities .25* .13 .08
Time social activities .25* .15 .14
Time passive activities -.33* -.28* -.26*
Step 5: Pleasure from activities (Inter) .36* .34*
Step 6: Negative incidents -.18*
R² .08* .00 .01 .26* .10* .03*
Total R² .08* .08 .08 .34* .43* .46*
Note. * p < .05. Time = number of hours spent on activity. Gender: 1 = male, 2 = female. Negative incidents:
0 = no, 1 = yes.
How Does Vacation Affect Health and Well-being?
99
4
In the final step, negative incidents during vacation contributed three percent of
explained variance in the changes in H&W during vacation, indicating that vacationers who
experienced at least one negative incident during vacation showed smaller increases in H&W
during vacation compared to vacationers without negative incidents (Hypothesis 6 supported).
We analyzed the answers regarding the nature of the negative incident(s) as well.
A substantial part of the respondents (44%) reported at least one incident on the day that we
called or within the previous vacation days (in total 65 negative incidents were reported). For
those who experienced negative incidents, 17% was confronted with a close other getting
injured, 15% with travel stress, 15% with being ill, 11% with getting injured themselves, 11% with
a close other being ill, and 11% suffered from bad weather and skiing conditions. The remaining
negative incidents reported were quarrels (3 incidents), children being upset (2 incidents), bad
news from families or friends at home (3 incidents) or were not further specified (5 incidents).
Summing up, work-related and social activities during vacation were not related to
H&W changes during vacation whereas physical activities and pleasure derived from vacation
activities were related to positive changes in H&W during vacation. Moreover, passive activities
and negative incidents were related to negative changes in H&W during vacation. The final
model accounted for 46% of the variance in H&W during vacation, or expressed differently,
for 38% of the variance in the vacation effect (i.e., computed as the percentage of explained
variance in H&W Inter (46%) minus the percentage explained variance of H&W Pre (8%)).
4.4. Discussion
The present study sought to examine vacation (after-) effects on H&W and the relation between
vacation activities and experiences, and changes in worker H&W during vacation. The three
most interesting sets of findings are the following.
4.4.1. A positive vacation effect does not apply to every employee
Our results showed that the examination of average scores of H&W during vacation compared
to work periods has a major drawback because it collapses individual well-being scores into
a somewhat insensitive average score that mistakenly suggests that vacation always (or at
least usually) results in a short-lived improvement in H&W. By instead inspecting the individual
difference scores of H&W between (pre-vacation) baseline and the vacation period, we
discovered an interesting trend: 60% of the vacationers experienced a strong positive change
in H&W during vacation which was maintained at least 2 weeks after work resumption. Put
differently, there are clear indications that a good vacation (i.e., with a strong positive vacation
100
Chapter 4
4
effect) charges the batteries of working people during vacation and for about two weeks
after work resumption before they return to baseline levels of H&W. However, a minority also
experienced none (23%) or even a negative (17%) effect of vacation. By averaging the H&W
scores, this pattern would not have been detected.
4.4.2. The jury is still out on the role of work activities while on vacation
Our results showed that H&W improved during vacation compared to working periods, which
supports our assumption that the temporary absence of work leads to increases in employee’
well-being. But how can we explain this study’s finding that time spent on work-related activities
during vacation was unrelated to the vacation effect? The prevalence of being engaged in
work-related activities during vacation was extremely low: all vacationers worked on average
less than 10 minutes a day, less than 20% (N = 19) of the vacationers spent time on work-
related activities during vacation at all, and the average working time for this ‘working while on
vacation’ group was less than half an hour per day (24 minutes). So, due to the low prevalence
of working during vacation, we cannot draw firm conclusions about the impact of working
during vacation on changes in H&W during vacation.
4.4.3. Activities and experiences during vacation are important
Time devoted to physical activities emerged as an influential vacation activity contributing to
positive changes in H&W during vacation. Pleasure derived from vacation activities appeared
to play a role in this relationship: the more time vacationers had spent on physical activities
and the more pleasure they experienced, the higher their H&W improvement during vacation.
Our study also showed that negative incidents happened frequently during a winter
sports vacation (in 44% of the sample) and that these incidents were related to decreases in
H&W during vacation. Accordingly, it would be interesting to examine whether vacations with
less risks of negative incidents (e.g. a relaxing summer vacation) would render even more
positive results for H&W during vacation.
Passive activities also emerged as influential because decreases in H&W during
vacation were associated with a higher engagement in passive activities. The strong positive
correlation between passive activities and negative incidents (see Table 4.1) suggests that
vacationers are probably ‘convicted to’ passive activities during their winter sports vacation
due to the occurrence of negative incidents. It would be interesting to find out how passive
vacation activities are related to H&W during vacations that are supposed to be passive and
relaxing. It may well be that the negative relationship between passive activities and H&W
during vacation will disappear or will turn into a positive effect when the vacation type becomes
less active, for instance during a relaxing vacation on the beach.
How Does Vacation Affect Health and Well-being?
101
4
4.4.4. Strengths and limitations
To our knowledge, the present study is the first that systematically investigated the role of
vacation activities and experiences during vacation itself. Vacation researchers correctly
stated that the logistics of locating people during vacation can be “nightmarish” (Eden, 1990,
p.182). However, we found a way to overcome many of the technical and practical problems
and collected information on vacation activities and experiences in a reliable, user-friendly
way during vacation. We also reduced attrition and missing data drastically by applying a
combination of innovative instruments for data collection (digital diaries, telephone surveys)
and an extensive protocol to guarantee compliance (careful recruitment, SMS-reminders, non-
completion script).
Yet, this study also has some limitations. Firstly, one might argue that the strength of
the vacation effect highly depends on the timing and the validity of the baseline measurement.
Nevertheless, this measurement occasion, i.e. two weeks before vacation, is presumably the
best estimate of baseline H&W because it is not biased by high pre-vacation workload and/
or pre-vacation pleasure which may colour measurement occasions scheduled immediately
before vacation (e.g. De Bloom, Geurts, Taris, Sonnentag, De Weerth & Kompier, 2010).
Moreover, we also measured H&W twice in that week to get a more reliable estimate of the
baseline.
A related problem is the use of difference scores for our subgroup analyses of the
difference between Pre and Inter. Some researchers (e.g. Cronbach & Furby, 1970) point out
that difference scores have a lower reliability than the single scores they were based upon
and should therefore not be used. Other scholars do not share this opinion (at least for the
application for research purposes), but recommend the careful use of these scores implying
reporting of standard errors as well as confidence intervals (e.g. Schulte & Borich, 1984;
Allison, 1990). This is what we did in this paper. The reliability of the difference score was high
in our study. Analysis of the difference scores added valuable information beyond our ANOVA’s
and regression analyses: reporting only average scores of H&W would obscure the fact that
there are different, even opposing trends across time in meaningful subgroups of employees.
Secondly, the limited variation in vacation type and duration leaves the question
unanswered if we would have found the same pattern of results for other vacation types, for
other vacations durations, and for other periods (seasons) of the year. Thirdly, winter sports
vacationers may be above-average healthy, active and sporty, which might limit the external
validity of our study.
Fourthly, we need to examine our study sample in relation to the external validity
of this study. This study’s sample includes both full-time and part-time employees, and one
might argue that the effects of vacation may differ for both categories. We have repeated our
102
Chapter 4
4
analyses for subsamples composed of full-time and part-time workers, but the results were
highly similar for both groups of workers and are therefore not presented in this paper.
Last but not least, one might argue that vacation activities were measured on only two
instead of all vacation days. However, because it is such a difficult task to study vacationers
in the field, we had to find a compromise between investigating the role of vacation activities
and experiences and interfering too much during people’s vacations with the risk of negatively
influencing our key study variables (we did not want our phone call to be a daily hassle and
negative incident in itself). The high response rates and positive process evaluations by the
participants suggest that our approach was bearable for the participants and scientifically
valuable.
4.4.5. Suggestions for future research
Future vacation research should, apart from an analysis of general trends across time and
means, include subgroup analyses to discover trends in H&W in different groups across a
vacation period. Moreover, our findings suggest that vacationers chose to engage in vacation
activities they preferred, i.e. experienced as pleasant. Regarding the positive impact of pleasure
from vacation activities on H&W during vacation, future research could focus on the role of
the active choice for pleasant activities by asking vacationers to indicate to what degree they
were able to choose for the activities they engaged in. Self-Determination Theory suggests
that autonomy to initiate behavior of one’s own choice fulfills a fundamental human need and
therefore leads to positive emotions (Ryan & Deci, 2000). However, it may well be possible that
not every vacationer is able to engage in the vacation activity of his/her own choice. Therefore,
vacation researchers should incorporate questions about the extent to which vacationers
are able to determine their own vacation activities, their day schedule during vacation, their
vacation destination, and the type of vacation.
Apart from choice of vacation activities and experiences of pleasure derived from
these activities, negative incidents during vacation appeared to be an important determinant
for the strengths of the vacation effect as well and should therefore be included in upcoming
studies. Regarding vacation activities, it would be interesting to examine the effect of passive
activities during more passive vacations (e.g. a relaxing vacation on the beach). Furthermore,
a study with a greater number of work-related activities during vacation would be useful to test
if not-working is indeed one of the keys to the vacation effect.
One of the most important findings of our study is that people did benefit mostly
from their winter sports vacation (i) by doing what they were supposed to do during this type
of vacation, i.e., being physically active, (ii) and by experiencing pleasure from their vacation
activities. Therefore, research on vacation should focus on vacation activities and the pleasure
How Does Vacation Affect Health and Well-being?
103
4
people derive from these vacation activities. Savoring positive vacation experiences, or “the
capacity to attend to, appreciate, and enhance the positive experiences of one’s life” (Bryant &
Veroff, 2007, p. 2) may be the key to maximize the positive vacation effect and prolong vacation
pleasure and relief after work resumption.
104
Chapter 4
4
4.5. References
Allison, P D. (1990). Change scores as dependent variables in regression analysis. In C.C. Clogg (Ed.),
Sociological Methodology (pp. 93-114). Oxford: Blackwell.
Baumeister, R.F., & Leary, M.R. (1995). The need to belong: desire for interpersonal attachment as a
fundamental human motivation. Psychological Bulletin, 117, 497-529.
Belkic, K., Landsbergis, P.A, Schnall, P.L., & Baker, D. (2004). Is job strain a major source of cardiovascular
disease risk? Scandinavian Journal of Work, Environment, & Health, 30, 85-128.
Bryant, F.B., & Veroff, J. (2007). Savoring. A new model of positive experience. Mahwah (NJ): Lawrence
Erlbaum Associates.
Cohen, J. (1988). Statistical power analysis for the behavioral sciences. Hillsdale (NJ): Lawrence Erlbaum
Associates.
Cronbach, L. J & Furby, L. (1970). How should we measure change- or should we? Psychological
Bulletin, 74, 68-80.
De Bloom, J, Geurts, S.A.E., Taris, T.W., Sonnentag, S, De Weerth, C., & Kompier, M.A.J. (2010). Effects
of vacation from work on health and well-being: Lots of fun, quickly gone. Work & Stress, 24,
196-216.
De Bloom, J., Kompier, M., Geurts, S., De Weerth, C., Taris, T., & Sonnentag, S. (2009). Do we recover
from vacation? Meta-analysis of vacation effects on health and well-being. Journal of
Occupational Health, 51, 13-25.
Demerouti, E., Bakker, A.B., Geurts, S.A.E. & Taris, T.W. (2009). Daily recovery from work-related
effort during non-work time. In S. Sonnentag, P. Perrewe, & D. Ganster (Eds), Research in
occupational stress and well being: Current perspectives on job-stress recovery (Volume 7,
pp. 85-123). Bingley (UK): JAI Press.
Dormann, C., & Zapf, D. (1999). Social support, social stressors at work, and depressive symptoms:
Testing for main and moderating effects with structural equations in a three-wave longitudinal
study. Journal of Applied Psychology, 84, 874-884.
Eden, D. (1990). Acute and chronic job stress, strain, and vacation relief. Organizational Behavior and
Human Decision Processes, 45, 175-199.
Eden, D. (2001). Vacations and other respites: studying stress on and off the job. In C. Cooper and I.T.
Robertson (Eds), Well-being in organizations (pp.305-330). West Sussex (UK): John Wiley &
Sons.
Elo, A., Leppänen, A., & Jahkola, A. (2003). Validity of a single-item measure of stress symptoms.
Scandinavian Journal of Work, Environment, & Health, 29, 444-451.
Etzion, D. (2003). Annual vacation: duration of relief from job stressors and burnout. Anxiety, Stress, and
Coping, 16, 213-226.
Fritz, C., & Sonnentag, S. (2006). Recovery, well-being, and performance-related outcomes: the role of
workload and vacation experiences. Journal of Applied Psychology, 91, 936-945.
How Does Vacation Affect Health and Well-being?
105
4
Gerin, W., Pieper, C., Levy, R., & Pickering, T.G. (1992). Social support in social interaction: a moderator
of cardiovascular reactivity. Psychosomatic Medicine, 54, 324-336.
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, & Health, 32, 482-492.
Gump, B., B., & Matthews, K.A. (2000). Are vacations good for your health? The 9-year mortality
experience after the multiple risk factor intervention trial. Psychosomatic Medicine, 62, 608-
612.
Hale, C. .J., Hannum, J. W., & Espelage, D. L. (2005). Social support and physical health: The importance
of belonging. Journal of American College Health, 53, 276-284.
Hansen, C.J., Stevens, L.C. & Coast, R.J. (2001). Exercise duration and mood state: How much is
enough to feel better. Health Psychology, 20, 267-275.
Hjortskov, N., Rissen, D., Blangsted, A.K., Fallentin, N., Lundberg, U., & Sogaard, K. (2004). The effect
of mental stress on heart rate variability and blood pressure during computer work. European
Journal of Applied Physiology, 92, 84-89.
Hoopes, L.L. & Lounsbury, J.W. (1989). An investigation of life satisfaction following a vacation: a
domain-specific approach. Journal of Community Psychology, 17, 129-140.
Kühnel, J., & Sonnentag, S. (2011). How long do you benefit from vacation? A closer look at the fade-out
of vacation effects. Journal of Organizational Behavior, 32, 125-143.
Lounsbury, J. W., & Hoopes, L.L. (1986). A vacation from work: changes in work and nonwork outcomes.
Journal of Applied Psychology, 71, 392-401.
Moran, A.P. (2004). Does a healthy body always lead to a healthy mind? Exploring exercise psychology.
In A.P. Moran (Ed.), Sport and exercise Psychology: A critical introduction (pp. 225-251). New
York: Routledge Taylor & Francis Group.
Mead, G.E., Morley, W., Campbell, P., Greig, C.A., McMurdo, M., & Lawlor, D.A. (2009). Exercise for
depression. Cochrane Database of Systematic Reviews, Issue 3. Art.No.: CD004366. DOI:
10.1002/14651858.CD.004366.pub4.
Nabkasorn, C., Miyai, N., Sootmongkol, A., Junprasert, S., Yamamoto, H., Arita, M., & Miyashita, K.
(2005). Effects of physical exercise on depression, neuroendocrine stress hormones and
physiological fitness in adolescent females with depressive symptoms. European Journal of
Public Health, 16, 179-184.
Nunnally, J. C. (1978). Psychometric theory (2nd ed.). New York: McGraw-Hill.
Reed, J., & Ones, D. S. (2006). The effect of acute aerobic exercise on positive activated affect: A meta-
analysis. Psychology of Sport and Exercise, 7, 477-514.
Rook, J. W., & Zijlstra, F. R. H. (2006). The contribution of various types of activities to recovery. European
Journal of Work and Organizational Psychology, 15, 218-240.
Ryan, R.M., & Deci, E.L. (2000). Self-Determination Theory and the facilitation of intrinsic motivation,
social development, and well-being. American Psychologist, 55, 68-78.
106
Chapter 4
4
Schulte, A., & Borich, G.D. (1984). Considerations in the use of difference scores to identify learning-
disabled children. Journal of School Psychology, 22, 381-390
Sonnentag, S. (2001). Work, recovery activities, and individual well-being: A diary study. Journal of
Occupational Health Psychology, 6, 196-210.
Sonnentag, S. & Jelden, S. (2009). Job stressors and the pursuit of sport activities: A day-level
perspective. Journal of Occupational Health Psychology, 14, 165-181.
Sonnentag, S., Perrewe, P.L., & Ganster, D.C. (Eds, 2009). Current perspectives on job-stress recovery.
Bingley: Esmeralda Group Publishing Limited.
Strauss-Blasche, G., Ekmekcioglu, C. & Marktl, W. (2000). Does vacation enable recuperation? Changes
in well-being associated with time away from work. Occupational Medicine, 50, 167-172.
Van Hooff, M., Geurts, S., Taris, T., Kompier, M., Houtman, I. & Van den Heuvel, F. (2005). Disentangling
the causal relationships between work-home interference and employee’ health. Scandinavian
Journal of Work, Environment, & Health, 31, 15-29.
Van Hooff, M.L.M., Geurts, S.A.E., Taris, T.W., & Kompier, M.A.J. (2007).”How fatigued do you currently
feel?” Convergent and discriminant validity of a single-item fatigue measure. Journal of
Occupational Health, 49, 224-234.
Westman, M., & Eden, D. (1997). Effects of a respite from work on burnout: vacation relief and fade-out.
Journal of Applied Psychology, 82, 516-527.
Westman, M., & Etzion, D. (2001) The impact of vacation and job stress on burnout and absenteeism.
Psychology & Health, 16, 595-606.
109
Chapter 5
Effects of Short Vacations, Vacation Activities
and Experiences on Employee’ Health
and Well-Being
It was investigated 1) whether employee’ H&W improve during short vacations (4-5 days),
(2) how long this improvement lasts after returning home and resuming work, and (3) to
what extent vacation activities and experiences explain health improvements during and
after short vacations.
Eighty workers reported their H&W two weeks before vacation (Pre), during vacation
(Inter), on the day of return (Post 1), and on the 3
rd
and 10
th
day after returning home (Post
2 and Post 3, respectively).
The results showed improvements in H&W during short vacations (d = 0.62), although this
effect faded out rather quickly. Partial correlations and regression analyses showed that
employees reported higher H&W during vacation, the more relaxed and psychologically
detached they felt, the more time they spent on conversations with the partner, the more
pleasure they derived from their vacation activities, and the lower the number of negative
incidents during vacation. Experiences of relaxation and detachment from work positively
influenced H&W even after returning home. Working during vacation negatively influenced
H&W after vacation.
In conclusion, short vacations are an effective, though not very long lasting, ‘cure’ to
improve employees’ H&W.
This chapter is based on:
De Bloom, J., Geurts, S.A.E. Kompier, M.A.J. (in press). Effects of short vacations, vacation
activities and experiences on employee’ health and well-being. Stress & Health
Effects of Short Vacations, Activities and Experiences
111
5
5.1. Introduction
Exposure to job stressors has negative effects on H&W (e.g. Akerstedt, 2006; Vrijkotte, Van
Doornen & De Geus, 2000). Consequently, recovery from work stress is essential to preserve
employee’ well-being. Recovery, defined as a period of absence from work and “[…] a situation
in which no special demands are made on the individual” (Geurts & Sonnentag, 2006), enables
the psychophysiological systems that were activated while expending effort at work to return to
and stabilize at baseline levels. In other words, recovery implies a reduction in stress.
According to Effort-Recovery Theory (Meijman & Mulder, 1998) and Allostatic Load
Theory (McEwen, 1998), initial normal load reactions associated with effort expenditure during
work (e.g. fatigue), can develop into more chronic load reactions if recovery is incomplete
during off-job time. Recovery occurs regularly in-between work periods, e.g. during evening
hours and during weekends. However, diary studies demonstrated that employees often
recover insufficiently during these short periods of respite (Fritz & Sonnentag, 2005; Van Hooff,
Geurts, Kompier & Taris, 2007). Vacation as a relatively long and less interrupted period of off-
job time could, therefore, be a more effective opportunity to recover from work.
Indeed, a meta-analysis on vacation effects showed that vacation has a small
positive effect on H&W when baseline levels before and the first measurement occasion after