Amsterdam Business School
Bachelor Future Planet Studies
Major Business Administration
Daily reconstruction method (DRM):
Life to work or work to life?
To what extent does leisure satisfaction influence happiness derived from work satisfaction of paid workers?
BSc Thesis by
Rutger Tiesma
10772995
Supervisor: Wendelien van Eerde
Amsterdam, 27
thJune 2017
2
Statement of Originality
This document is written by Rutger Hans Tiesma who declares to take full responsibility for the contents of this document. I declare that the text and the work presented in this document is original and that no sources other than those mentioned in the text and its references have been used in creating it. The Faculty of Economics and Business is responsible solely for the supervision of completion of the work, not for the contents.
3
Abstract
Not your genes or life circumstances determine your happiness, but the activities you undertake. This study uses two conceptual models to investigate the influence of work and leisure activities on happiness. First, it is hypothesized that leisure engagement is positively associated with one’s happiness, and that this relation is mediated by leisure satisfaction (1). Besides that, leisure can forestall work-related problems, and satisfaction derived from work may thus be translated into one’s happiness. It was therefore examined whether leisure satisfaction could positively influence the relation between work satisfaction and happiness (2). To test the hypotheses of both conceptual models data of ‘de Gelukswijzer’ has been examined. This website recorded what 8.188 participants had done on the previous day and their satisfaction levels during these activities. First, no evidence in support of conceptual model 1 was found. However, the variable leisure satisfaction did significantly affect the relation between work satisfaction and happiness in conceptual model 2. This suggests that higher levels of leisure satisfaction influence the work satisfaction on happiness positively.
Keywords: happiness, leisure satisfaction, work satisfaction, leisure engagement, work-detachment, daily reconstruction method, cross-sectional research
4
Acknowledgement
Besides answering the main question I endeavour to give the reader the message that other factors than monetary resources can cause happiness:
Not economic welfare, but other factors, such as paying attention to family life and health contribute to happiness (Easterlin, 2004). Humankind has to re-examine the view that monetary resources, mentioned by Barrington-Leigh (2013) as the strong material consumption-oriented view of welfare, directly leads to well-being. Especially, since the world is facing enormous global problems (Barrington-Leigh, 2013). Challenge yourself by thinking critically about the following question proposed by the World Happiness Report 2012 (as cited by Helliwel et al., 2012): ‘‘Should the world pursue Gross National Product the point of environmental ruin, even when incremental gains in GNP are not increasing much (or at all) the happiness of affluent societies?’’
5
Content
1. Introduction p. 7
2. Theoretical framework p. 9
2.1. The context of happiness p. 9
2.2. Positive psychology p. 9
2.3. The pursuit of happiness p. 11
2.4. Leisure engagement p. 12
2.5. Leisure satisfaction p. 13
2.6. Work and Leisure p. 14
3. Method p. 15 3.1. Design p. 15 3.2. Procedure p. 15 3.3. Measurements p. 16 3.3.1 Happiness p. 16 3.3.2 Leisure engagement p. 16
3.3.3 Leisure and work satisfaction p. 17
3.4. Descriptive statistics p. 17
3.4.1 Sample p. 17
3.4.2 Leisure time and mode p. 19
3.4.3 Work time and mode p. 20
3.4.4 Activity satisfaction and overall happiness p. 20
3.5. Analysis and prediction p. 21
4.5.1 Conceptual model 1 p. 21
6
4. Results p. 22
4.1. Reliabilities and correlation p. 22
4.2. Main analysis p. 23 4.2.1 Conceptual model 1 p. 24 4.2.2 Conceptual model 2 p. 25 4.3. Additional analysis p. 26 4.3.1 Motivation p. 26 4.3.2 Results p. 27 5. Discussion p. 28 5.1. Summary p. 28 5.2. Unexpected results p. 29
5.2.1 Leisure engagement and happiness p. 29
5.2.2 Leisure satisfaction and happiness p. 29
5.2.3 Work satisfaction and happiness p. 30
5.2.4 Other factors p. 30
5.3. Interpretation of the findings p. 30
5.4. Limitations and future research p. 31
5.5. Concluding thoughts p. 32
6. References p. 33
7
1. Introduction
More and more studies are revealing that characteristics and resources considered as valuable by a society are positively related to happiness. Research on happiness includes multiple life domains, such as marriage, companionship and work performance (Lyubomirsky et al., 2005a; Easterlin, 2003).
Happiness can in turn positively affect health, both physical and mental (Davidson et al. 2010; Koivumaa‐Honkanen et al., 2001; Koivumaa-Honkanen, 2004). To illustrate, a research conducted by Davidson et al. (2010) found that higher levels of happiness were associated with a lower risk of heart disease. Aristotle, one of the most influential philosophers in history, considered happiness as (cited by Prinsloo, 2013): ‘‘the meaning and the purpose of life, the whole aim and end of human existence.’’
Another resource that is considered to enhance happiness is income (Lyubomorsky et al., 2005a). This verdict is ratified by Hagerty and Veenhoven (2013), who noted that an increase of national income goes hand in hand with increasing national happiness, especially on the short-term. However, others challenge this assumption and found that not income (Easterlin, 1995), but other characteristics and resources of a society co-vary with wealth, such as social trust (Easterlin, 2004; Helliwell et al., 2012). The relation between earning more money and an individual’s level of happiness seems also ambiguous. Kahneman et al. (2006) even assessed that this relation seems to be mostly illusory. One explanation proposed by the researchers is that a rise in income leads to a decreased participating in leisure activities that can cause an improving effect (Kahneman et al., 2006). Additionally, jeopardizing the enjoyment of someone’s leisure activities is extremely unwise when striving happiness (Aaker et al., 2010).
Instead, participating in leisure activities, such as shopping, exercising and watching TV, will be more effecting in pursuing happiness. According to Diener and Newman (2013), leisure is able to promote happiness, since it triggers the following psychological mechanisms: detachment-recovery, autonomy, meaning, mastery and affiliation.
However, the lack of recovering from work has been observed as a problem nowadays, since the fast-paced 24/7 economy causes that employees are constantly occupied with work-related issues (Sonnentag et al., 2012). Employees are not able to unwind and relax and instead face much stress and other work-related constraints (Sonnentag et al. 2012; Moreno-Jiménez et al., 2009).
Scholars have extensively investigated the relations between work-related factors and happiness, and leisure activities and happiness. Although, less research has focused on the combination. This is remarkable, since issues such as the work-life balance are much more a problem in these times where more and more people live to work, instead of working to live
8 (Walker, 2016). In response to this observation, the study wants to contribute to the literature that leisure can positively influence the relation between work satisfaction and happiness. Even satisfied hard workers can constrain problems related to work in the long run. To illustrate, it turned out that not detaching from work was highly associated with emotional exhaustion (Sonnentag et al., 2010). Do leisure activities compensate the effects of work on happiness, as being wondered by Walker (2016)?
In order to investigate the question Walker (2016) proposes, this study examines the following question: To what extent does leisure satisfaction influence happiness derived from work satisfaction of paid workers?
For answering this question a quantitative approach has been chosen. Gathering data is executed by ‘de Gelukswijzer’; a website consisting of around 9000 unique users who rate the happiness they experience during an activity, during the day and even a complete month (11-point Smiley Scale; Ruut Veenhoven, 2014). The data that has been collected can be characterised as diary data. Diary studies, well known as the ‘‘daily diary methodologies’’, refer to respondents that self-report on daily scale (Lischetzke, 2014). Specifically, to collect the quantitative data, the Day Reconstruction Method (hereafter DRM) grounded by Kahneman et al. (2004) has been used. DRM helps minimizing distortions and is superior above other daily self-measurements, since it asks people to reconstruct their day chronologically. By doing this, people can more precisely remember what their mood was during a certain activity (Kahneman et al., 2004).
Although the sufficient DRM is used, much inconvenience exists about the direction of the relation between variables in the research area of happiness (Tait et al., 1989). For in-stance: are people who are married happier, or are people more is it that people who are happy are likely to get married (Lyubomorsky, 2005b)? More related, does leisure enhance happiness or do happy people do more leisure?
Therefore, I want to examine if the domain of leisure enhances happiness. Kuykendall et al. (2015) found that leisure engagement could influence someone’s happiness positively by using leisure satisfaction as a mediator. In other words, the relation could be explained if someone is happy about the leisure activity undertaken. Still, since many studies are firstly investigating if the direction is as predicted, this study endeavours to do this too. This is espe-cially important since leisure satisfaction is the moderating variable in the second conceptual model. In the next section a literature review will be provided.
9
2. Theoretical Framework
2.1 The context of happiness
At the beginning of the civilization, people already tried to fully understand get rid of the concept of happiness. The ancient Greek spoke of eudaimonia, coined by Aristotle, referred to as ‘‘the state of having an objectively valuable life’’ (Aristotle, 1094; Griffin, 2007). The Islamic sacred book, the Koran, describes happiness as the following: "The hearts of those who believe find contentment in the remembrance of Allah; for without doubt in the remembrance of Allah do hearts find contentment" (Sûrah al-Ra`d: 28).
However, happiness does not directly emanate from leading a good life, according to several measurable objectives, which was found to be important by Aristotle and the Koran. Happiness is still reachable for those whose lives are not considered to be ‘‘good’’ (Robertson, 2016). To illustrate, imagine that someone can gain happiness, for instance, by playing the violin well in an orchestra. Still, one can extract happiness in attending the symphony just by listening (Robertson, 2016). This refers to the subjective aspect of happiness; as with the origin of the English word happiness, which comes from the noun ‘‘hap’’ (Robertson, 2016; Griffin, 2007).
The common present use of this word contains the residue of both historical thoughts on happiness. ‘‘Happy’’, in this sense, encounters both aspects of objectivity and subjectivity. Griffin (2007) explains happiness as the following: ‘‘happy’’ is to be glad or satisfied or content (a suggestion there of subjectivity) with having a good measure of what one regards as important in life (a suggestion there of objectivity).’’. An extensive list of literature has proposed that a combination of the subjective approach and the objective approach should be the leading definition. (Alexandra, 2005; Griffin, 2007).
Another label used for happiness is subjective well-being (hereafter SWB). However, these concepts are not strictly seen as equal, since SWB encompasses only the subjective approach (Easterlin, 2004; Alexandrova, 2005). Researchers are using these definitions interchangeably. To prevent the reader from confusion the research only wields happiness as the comprehensive definition. Easterlin (2004) concludes it is suitable to use happiness for SWB, or vice versa. The next section will elaborate further on happiness and its antecedents.
2.2 Positive psychology
The humanistic movement of positive psychology, which was grounded around the mid-twentieth century, has gained recognition lately (Seligman & Csikszentmihalyi, 2014). The aim of positive psychology is to minimize pathological thoughts by attempting psychological interventions to foster positive attitude toward one’s happiness (Seligman & Csikszentmihalyi, 2014).
10 Until 1998, the research topic within psychology on the relation between factors influencing happiness were only focussing on mental illness, instead of mental health. Happiness could hardly be altered, since the common view was that 80 percent of one’s happiness was predetermined by its genes (Sheldon & Lyubomirsky, 2007). Ultimately, Martin Seligman chose positive psychology as subject for his term when elected as president of the American Psychological Association (Time Magazine, 2005). This encouraged researchers to focus on psychological interventions that promote positivity and happiness (McDonald & O’callaghan, 2008).
Besides that, the majority of studies has examined the effects of negative daily life events on happiness, such as the effect of daily negative financial hassle (Gable & Reis, 2010; Sturgeon, 2014). The lack of recognition of positive psychology has probably ensured that surprisingly little is known about the contribution of positive daily life events. Gable and Reis (2010) propose two fundamental reasons why it is worth examining how individuals react to positive events. Initially, events related to positive feelings emerge more often than the negative ones, and secondly, positive events could have meaningful implications for one’s well-being (Gable and Reis, 2010). For example, it turned out that fewer depressive symptoms were identified when one experienced higher levels of daily positive events (Zautra et al, 2010). A wide range of researches proposed techniques to increase one’s happiness. One early technique that tried to assess happiness was Fordyce’s Happiness Training Program, developed in 1993. This technique presented the ‘‘fourteen fundamentals of happiness’’, falling in the following five categories: change your thinking, value personal growth, invest in friendship and companionship, eliminate negative emotions, and change your activities (Fordyce, 1993).
Furthermore, Luyubomirsky et al. (2005b) proposed that happiness can actually be altered. This view on happiness was proposed in a model called the Sustainable Happiness Model (hereafter SHM; 2005b). This technique proclaims that long-term sustainable happiness relies on the life circumstances, someone’s inherently determined set-point and intentional activity (Lyubomirsky et al., 2005b). Intentional activities are particular actions or practices in which people have the choice to involve (Lyubomirsky et al., 2005b). Lyubomirsky et al. (2005b) concluded that the population variance on happiness was due to these three determinants and they were able to attach approximate percentages to each of these factors. It appears that 50 percent of the variance is simply due to a person’s genetics (Lykken & Tellegen, 1996). Circumstances only account for a 10 percent share of someone’s long-term happiness (Sheldon & Lyubomirsky et al., 2007; Diener et al., 1999), and intentional activities explain almost 40 percent of the durable happiness (Sheldon & Lyubomirsky et al., 2006).
Circumstances, which can be defined as one-time changes include, for example, marital status change, increase in income, promotion, or buying a new car. Oerlemans et al.
11 (2014) concluded that getting married and having your first child are the events that have the biggest positive influences on someone’s happiness (Oerlemans et al., 2014). Respectively, an average gain in happiness of 0,4 and 0,5 occurred after these ‘one-time’ occasions (10-point Likert-scale). Surprisingly, these ‘‘immense’’ changes on happiness just account for a 10 percent share of someone’s long-term happiness.
According to this, one can assess that altering someone’s long-term happiness can be achieved through changing activities, more than through one’s circumstances (Lyubomirsky et al., 2005b). But why is it that that intentional activities can, compared to circumstance changes, foster one’s happiness? The next section will give an overview of research devoted to answering this question.
2.3 The pursuit of happiness
In the Conquest of Happiness, written in 1930, Bertrand Russell states that happiness is not a privilege of the educated (Holowchak, 2006 p. 1). The door towards contentment can be found by all people, as the following paraphrase of the first chapter highlights: ‘‘The world is vast and our own powers are limited. If all our happiness is bound up entirely in our personal circumstances it is difficult not to demand of life more than it has to give (Holowchak, 2006 p. 1, p. 5)’’. So, endeavouring to change your circumstances can alter your happiness, but why? There is a central assumption that circumstance changes, compared to intentional activity changes, are much more quickly subject to hedonic adaption. Hedonic adaption refers to the idea that humans quickly return to a relatively stable set-point of happiness (Diener et al., 2006; Headey, 2008; Sheldon et al., 2012). This can be seen as a critical difference. Strictly speaking, the effects on happiness of these circumstance changes tend to decay more quickly (Sheldon & Lyubomirsky, 2007).
To digress, circumstance changes in one’s life can trigger substantial increases in happiness, such as with getting married and having your first child. Still, these circumstance changes tend to be short-lived and people quickly start to take the new situation for granted (Headey and Wearing 1989). This can also explain the results in the weakly associated relation between income and long-term happiness (Sheldon & Lyubomirsky, 2006; Diener et al., 1999), and the connection between happiness and health status and geographic region (Diener et al., 1984). Brickman et al. (1978) concluded that lottery winners, who obtained a one-time change, were hardly happier than the controls in both short and long term. Winning an immense amount of money lessens the pleasure found in relatively normal events (Brickman et al., 1978).
Intentional activities, on the contrary, contain properties that could forestall the hedonic adaptation. First, these activities place the attention on one’s energy and behaviour in a variety of different ways, causing a wide and diverse array of experiences (Sheldon & Lyubomirsky,
12 2007). Second, new possibilities and opportunities arise from intentional activities, which can possibly lead to positive sustainable effects in one’s life (Fredickson and Joiner, 2002).
Moreover, one fundamental outcome in the prevention of hedonic adaption, assessed by Sheldon et al. (2012), is that intentional activities should not be undertaken at the same time of day, in the same way and pursuing the same purposes and goals. Adaption applies only to activities that are held constant during a time period (Frederick & Loewenstein, 1999). Therefore, striving to encounter new pleasing insights and context of an activity will enhance the forestalling of hedonic adaption (Sheldon et al., 2012).
2.4 Leisure Engagement
Leisure engagement can be seen as an umbrella definition that involves the amount of time spent on leisure, the diversity of leisure activities undertaken and/or frequency of leisure participation (Kuykendall et al., 2015). Activities involved are, for example, playing games, playing sports and social activities (Kleiber et al., 2011). Sonnentag & Fritz (2007) found that relaxing and reading books were associated with higher levels of happiness. Kuykendall et al. (2015) provided evidence that the relation between leisure engagement and enhancing happiness was strong.
To digress, Csikszentmihalvi et al. (1980) found that leisure activities were mostly con-sidered to entail the characteristics of intrinsic motivation and perceived freedom. Similarly, Siegenthaler and Vaughan (1998) concluded that expressing yourself and the idea of having freedom are the crucial points that distinguish leisure from other activities. This is in line with the paradigm, better known as Neulinger’s leisure theory, that argues that intrinsic motivation and perceived freedom can be seen as characteristics that are the basis for why some activities can alter one’s happiness positively (Holt & Ashton-Shaeffer, 2001). Moreover, one could ar-gue that intentional activities are carrying the same characteristics, since intentional activities are considered to be activities where one has the choice to involve in particular actions or practices (Lyubomirsky et al., 2005b; Sheldon & Lyubomirsky, 2007). Besides that, leisure participation provides the opportunity to reduce negative feelings and distract one from stress-ful problems, which normally elevates the individual’s happiness (Haines et al., 2007).
These theories and results are in accordance with the outcomes of researchers from a wide range of different disciplines, such as sociology, psychology, and gerontology, who also found increasing scientific support that leisure is the decisive activity for achieving higher levels of happiness (Pressman et al., 2009; Lloyd & Auld, 2002; Brajša-Žganec et al., 2011). Ulti-mately, it is defensible to state the following hypothesis:
13
2.5 Leisure Satisfaction
First, it is demonstrated that leisure engagement influences happiness (Kuykendall et al., 2015). Therefore, a bottom-up model can be seen as the appropriate one, since this theory posits that happiness is caused by particular life experiences (Kuykendall et al. 2015). Other scholars, such as Headey et al. (1991) and Diener (1984) proclaimed that this model was more substantial compared to the top-down approach, in which happiness influences life experi-ences. Kuykendall et al. (2015) acknowledges that leisure engagement influences happiness through leisure satisfaction.
Amestoy et al. (2008, p. 65) define leisure satisfaction in the following way (based on Nawijn & Veenhoven, 2013): “positive perceptions or feelings that an individual forms, elicits, or gains as a result of engaging in leisure activities and choices. Happiness through leisure is the degree to which one is presently content or pleased with one’s general leisure experiences and situations. This positive feeling of pleasure results from the satisfaction of felt or unfelt needs of the individual.” In other words, leisure satisfaction is the degree to which one can derive enjoyment or satisfaction by participating in leisure activities. Leisure satisfaction is dif-ferent from leisure engagement, since it does not reflect the content of undertaken leisure activities (Newman et al., 2014).
Although the top-down approach has gained the most recognition lately this study wants to give the decisive answer and will, therefore, test this. Subsequently, as in the research of Kuykendall et al. (2015) the relation between leisure engagement and happiness with the mediator leisure engagement will be tested.
H2: Leisure engagement is positively associated with leisure satisfaction.
H3: Leisure satisfaction will positively mediate the relation between leisure engagement
and one’s happiness.
2.6 Work and Leisure
Research has been developed extensively on what makes people happy at work, which is not surprising, since the hours spent on work are substantial in our lives (Fisher, 2010; De Neve & Ward, 2017). It seems that work could directly affect happiness (Geurts et al., 2003; Ryan et al., 2010; Sonnentag et al., 2008). Especially, the strong relationship with happiness can be observed when overall work satisfaction is measured. In contrast, only testing job facets, such
14 as the relation with your colleagues and boss, were considered to have little impact on happi-ness (Tait et al., 1989).
Research indicates that work satisfaction was positively associated with happiness (Tait et al., 1989; Sonnentag, 2012). However, studies had difficulty determining the direction of effect and it was assumable to consider the link between work satisfaction and happiness as a reciprocal relation (Tait et al., 1989). However, some researchers found that work satisfaction causes a positive effect on happiness, instead of the other way around (Judge & Lock, 1993). They argue that this is caused by the part-whole theory (Tait et al. 1989), where particular sub domains can predict an overall outcome. In this case, a specific facet of someone’s life predicts the outcome of one’s overall happiness. In sum, research seems to indicate that it is most likely that work affects happiness and therefore this study hypothesises:
H4: Work satisfaction can positively influence one’s happiness.
Of course, work is not the only activity individuals fill their lives with. Sociologists, psychologists and leisure scholars have all wondered how paid work and leisure relate to each other (Walker, 2016). Especially, scholars have argued that it is useful to find out if leisure activities can compensate for particular levels of satisfaction in other life domains, and especially the domain of paid work (Ryan & Deci, 2000), which could enhance levels of well-being (Walker, 2016).
To illustrate, Allport (1924), a progressive psychologist, proclaimed that the relationship between the domains of paid work and leisure should be considered by the following example (as cited by Walter, 2016 p. 231): ‘‘a worker is fated to remain at a modest economic and vocational level, vicarious compensations should be sought in avocational interests [and the] wise employment of leisure”. In general terms, workers can stay content with their current work activities and lives when they involve in leisure they find satisfying.
This verdict can be confirmed and substantiated by a more recent research conducted by Sonnentag et al. (2008). They argue that work engagement should always go together with periods of time in which people can physically and psychologically detach from work and involve in other activities. Sonnentag et al. (2008 p. 259) refers to work engagement as: ‘‘a positive, fulfilling work-related state of mind that is characterized by vigour, dedication, and absorption’’. Van Hooff (2011) found that lower levels of work detachment were characterized by indicators of poor well-being. Moreover, a lack of detachment from work was found to be related to a negative effect on an individual’s mood and it’s energy resources (Richardson & Thompson, 2012). On the contrary, people who detached from work reported higher levels of happiness (Rodríguez-Muñoz et al. 2014; Sonnentag & Fritz, 2007).
15 Still, psychological detachment from work seems much more complicated than being physically detached from work (Sonnentag & Fritz, 2007). According to the compensation theory proposed by Chick and Hood (1996) a person is likely to engage in leisure activities that energizes him or her in other ways then work does. Demerouti et al. (2012) argue that through this a person forgets the daily hassle of work on in the short-term. Moreover, in the long run, a longitudinal study over 1 year showed that the lack of psychological detachment from work was highly associated with an increase of employees that were emotionally exhausted (Sonnentag et al., 2010). So, research seems to suggests that leisure satisfaction could encourage an employee to detach from work. Therefore, it could be that the more one can enjoy their leisure, the more satisfaction through work is translated into overall happiness, since work constraints, such as emotional exhaustion and poor sleep, will not be translated into one’s happiness in daily life. In sum, the next hypothesis is:
H5: Leisure satisfaction will positively moderate the relationship between work
satisfaction and happiness.
3. Method
3.1 Design
A quantitative study is used in order to investigate the research question and hypotheses. This study is cross-sectional, since respondents specific data is collected at a specific time point and not compared with other data points (Levin, 2006). The data collected is used in a deductive way, which concerns that hypotheses are developed based upon the existing literature (Wilson, 2010 p. 7), as mentioned in the previous section. Moreover, the DRM has been used by ‘de Gelukswijzer’, which can minimize distortions as already mentioned in the introduction.
3.2 Procedure
Participants were recruited by means of asked using a wide range of channels, including Dutch popular magazines, social media and customer communications from the health insurer VGZ (Bakker et al., 2015). People were subscribed to the site of ‘de Gelukswijzer’ and were therefore initially targeted to fill in the questionnaire called the ‘Happiness Diary’, in Dutch ‘Gelukswijzer‘ (Veenhoven, 2015; Bakker et al., 2015; Burger & Veenhoven, 2016). They receive a hyperlink to their own page on ‘de Gelukswijzer’, where they have to fill in a short questionnaire. Subsequently, respondents are able to fill in the ‘Happiness Diary’, if desired.
16 The data this study investigates ranges from entries. In total 8.118 users have filled in the ‘Happiness Diary’. The characteristics of this sample of persons and the aligned statistics are clarified in section 3.4.
3.3 Measurements
3.3.1 Dependent variable: happiness
Happiness is self-rated by the respondents. To measure happiness, the following question is asked: ‘‘How happy are you today? (in Dutch: Hoe gelukkig voel je je vandaag?)’’. The item is
ranked on a 11-point Smiley scale, ranging from (0) to (10) (Davies & Brember, 1994). A high rank indicates that people are happy with their life on that particular day. The Cronbach’s α cannot be measured, since the variable relies on just one single item. This is also the case for the other variables involved in this research.
3.3.2 Independent variable: leisure engagement
To begin, researchers distinguish two leisure measurement approaches from each other, namely the structural approach from the subjective approach (Newman et al., 2014). The struc-tural approach refers to the extent to which someone undertakes leisure activities. Here, the researcher defines which activities can be labelled as leisure (Kleiber et al., 2011), thus what is normatively defined by society as leisure (Kelly & Godbey, 1992). The subjective approach, on the other hand, only attaches leisure to the activity when the individuals itself see it as a leisure activity (Kleiber et al., 2011). Kuykendall et al. (2015) assesses that both measures form the conceptualization of leisure engagement; the extent to which one undertakes leisure activities. Nevertheless, measuring the subjective approach from the dataset provided by ‘de Gelukswijzer’ seems not possible, since participants could only register what activities were undertaken, and not necessarily if the activity felt as leisure. It is therefore a requisite to assess which activities can be seen by our respondents as leisure.
Ultimately, relaxation, exercising, going out and association are labelled as leisure. These activities, and their sub-activities are generally seen as leisure activities (Wei et al., 2015). The leisure activities and their corresponding sub-activities are all listed in Table 1.
As described by Kuykendall et al. (2015) leisure engagement is made up of the following three components: frequency, total time and diversity. Only the total time per activity could be determined from the provided dataset of ‘de Gelukswijzer’. For this, respondents were asked to make an indication when they started and ended the particular activity. Time intervals are based upon time units of half an hour.
17
Leisure activity
Relaxation Exercising Going out Association
Hobby Walking Cinema Politics
Computer Cycling Theatre Church, Mosque
Reading Sporting Café Sport
Watching TV Other Watching sports Other
Listening to music Disco
Conversation Shopping
Other Cultural
Other
3.3.3 Mediation: leisure satisfaction and moderator: work satisfaction
After listing the activities of a particular day, participants are asked to rate how well they had felt during each particular activity (in Dutch: Hoe gelukkig voelde je jezelf bij de activiteit?). This item, considered as satisfaction, was ranked on an 11-point Smiley scale, which has already been mentioned as the scale for happiness (Davies & Brember, 1994). This was also done for all the leisure activities (Table 1).
In the same way work satisfaction is measured on an 11-point Smiley scale (Davies & Brember, 1994). Likewise, not the satisfaction of leisure is measured, but the satisfaction of work activities.
3.4 Descriptive Statistics
3.4.1 SampleFirst, the sample must be clarified; it consists of 219.908 entries from 8.118 unique users of ‘de Gelukswijzer’. Of those, 7.651 filled in the most elemental demographic characteristics on the website, such as gender and their living situation. The research examines the demographic characteristics of only these respondents, since for them comparisons can eventually be made. Therefore, N = 7.651 is used. The demographic characteristics of the respondents that are highlighted are depicted in table 1a. Most of them were female (N = 7.651, 79%), were at least medium-level educated (87%), and had paid jobs (68%).
From this, we can conclude that the sample is not representative for the Dutch society. For example, 52% of the respondents are high educated. However, the Dutch Central Bureau for Statistics (in Dutch: Centraal Bureau voor de Statistiek (CBS), 2015) has identified that only 29% of the Dutch population is high educated (N = 7.651, 29%). It is therefore not possible to generalize the results. Still, conclusions can be drawn for the persons who use ‘de
18 Gelukswijzer’. Lancée et al. (in press.) assesses that significant deviations in the sample do not necessarily generate a big problem for researches similar to this one. Their study is about how much time spent on commuting is optimal for one’s happiness. The generated results are for particular people, namely people that aim to improve one’s happiness (Lancée et al., in press.). This study focuses, therefore, also on this specific group; representativeness for the Dutch society population is not a requisite in this research. Moreover, for work-related analysis N = 5,492 has been used (table 1), which is based on the 7.651 users with only paid jobs. Volunteer jobs and unpaid jobs are excluded, since outcomes on one’s happiness are substantial. For instance, for unpaid jobs psychological stress is less present (Lundberg, 1996). In fact, Mojza et al. (2011) found that unpaid jobs could especially help people detach from work. Of the paid workers 27% was working in the healthcare and 29% in other sectors. Moreover, employees of this sample work for almost 27 hours per week (M = 26.96, SD = 20.45) on average. All demographics of the respondents are depicted in table 1b displayed in the appendices.
Table 1a: Demographic characteristics of respondents
Variable N Mean S.D. Min. Max. %
Gender (1=male, 2= female) 7.651 1,79 0,41 1 2
Education Low-level (ISCED 1997 1-2) 7.651 13% 0 1 Medium-level (ISCED 1997 3-5) 7.651 35% 0 1 High-level (ISCED 1997 5-6) 7.651 52% 0 1 Living together 7.651 89% 0 1 Paid work 7.651 68% 0 1 Sector of employment Government 5.492 9% 0 1
Education and culture 5.492 13% 0 1
Healthcare 5.492 27% 0 1
Business and financial services 5.492 16% 0 1
Retail 5.492 6% 0 1
Other 5.492 29% 0 1
19
3.4.2 Leisure time and mode
The frequencies for all the leisure sub-groups are shown in Table 2. Most participants indicated that they were undertaking particular activities for less than 30 minutes (24%). From the four groups of activities, relaxation was the most undertaken activity (78%), followed by exercising (14%) and going out (6%). The category ‘Association’ accounts for only 2% of all leisure activities undertaken. Going out activities were to a large extent undertaken with someone else (82%), compared to association activities (72%), exercising (52%) and relaxation (47%).
Table 2: Aspects of Leisure Activities
Variable
% of total number of episodes or diary entries (N= 51.594) Leisure activity Relaxation 78% Exercising 14% Going out 6% Association 2% Time of leisure activity
Very short (30 minutes or less) 24%
Short (30-60 minutes) 18%
Bit short (60-90 minutes) 14%
Medium length (90-120 minutes) 12%
Bit long (120-150 minutes) 8%
Long (150-180 minutes) 7%
Very long (more than 180 minutes) 17%
Leisure with someone
Relaxation 47%
Exercising 52%
Going out 82%
Association 72%
3.4.3 Work time and mode
The descriptive statistics for work variables are depicted in Table 3. Most participants indicated that the time they worked was less than 2 hours (42%), followed by work activities that last between 2 hours and 4 hours (26%). Almost no one was working more than 10 hours a day (2%). Over half of the work activities (58%) were at least undertaken with someone else.
20 Table 3: Descriptive statistics for work activities
Variable
% of total number of episodes or diary entries (N= 12.406)
Time of work activity
Very short (less than 2 hours) 42%
Short (2-4 hours) 26%
Bit short (4-6 hours) 13%
Medium length (6-8 hours) 7%
Bit long (8-10 hours) 10%
Long (10 hours or more) 2%
Working with at least one person 58%
3.4.4 Activity satisfaction and overall happiness
The descriptive statistics for happiness variables are be depicted in Table 4. The average daily happiness of respondents was 6.23 on a 10-point Likert scale1. This is far below the average
happiness score of people, which is considered to be 7.0 on a 10-point Likert scale (Veenhoven, 2015). In general, leisure satisfaction is considered to be higher (M= 7.92, SD = 2.17) than other activities (M = 7.50, SD = 2.18). Moreover, work related activities are considered to have the lowest satisfaction (M = 7.22, 2.43).
Table 4: Descriptive statistics for happiness and satisfaction variables
Variable N Mean Median Min. Max. SD Average leisure satisfaction 52.449 7,92 8,00 1 11 2,17 Average work satisfaction 12.579 7,22 8,00 1 11 2,43 Average satisfaction of other
activities 218.097 7,50 8,00 1 11 2,18
Average daily happiness 7.651 6,85 7,00 1 11 2,24
3.5 Analyses and Predictions
Field (2009, p. 545) describes that regression can be considered as the most appropriate test, since both model consist of two independent variables and one dependent variable. By using regression analyses, this research wants to identify whether the data fits the conceptual
1 Conversion of 11-point Smiley scale to 10-point Likert scale is as follows:
21 models proposed. Specifically, regression can test whether there is a relationship between variables. If there is a relationship, regression can also test the strength of it (Field, 2009 p. 545).
3.5.1 Conceptual model 1
For this model, four regression analyses are executed. The first regression model tests whether hypothesis 1 is true. To digress, it tests whether the independent variable leisure engagement predicts the dependent variable happiness. According to literature, a positive effect of the independent variable on the dependent variable is expected, which means that people undertaking a leisure activity for a longer period of time are more happy on that specific day compared to the ones who invest less time in that leisure activity.
The second regression identifies whether the second hypothesis can be confirmed. So, if the time spent on leisure activity predicts leisure satisfaction. It is expected that this will be a positive relationship. In other words, the more time spent on a leisure activity, the higher the satisfaction resulting from that particular activity will be.
The third regression model tests whether hypothesis 3 is true. In this case, it tests whether the mediator leisure satisfaction predicts happiness. Again, a positive effect between the variables is expected. To be precise, respondents that indicate that they experienced more satisfaction during the leisure are the ones who report higher levels of happiness during the day.
At last, for testing hypothesis 4, it is tested whether the relationship between the independent variable time spent on leisure activity and dependent variable happiness are better explained if leisure satisfaction, as the mediating variable, enhances happiness. Therefore, it is expected that happiness is enhanced through the leisure satisfaction resulting from leisure engagement.
3.5.2 Conceptual model 2
The second model involves a moderator, instead of a mediator. Therefore, the main effects of the independent variable work satisfaction and the moderating variable leisure satisfaction on the dependent are first executed. These two direct effects are predicted for the first regression model. It is hypothesized that the work satisfaction has a main effect on one’s happiness. Therefore, a positive effect regarding this relation is expected. With the second direct effect it is hypothesized that leisure satisfaction enhances happiness. So, the expectation is made that there is a positive, direct effect of the moderating variable on the dependent variable happiness.
The second step tests whether there is an interaction effect between the independent variable work satisfaction and the variable that moderates (leisure satisfaction) on the
22 dependent variable happiness. It is expected that the score of leisure will positively moderate the correlation between work satisfaction and one’s happiness. Thus, people that are more happy with the leisure experienced are the ones who experience a more positive effect on the relation between the satisfaction of work-related activities on their happiness.
4. Results
4.1 Correlations
Table 5 and Table 6 show the means, standard deviations, and the correlations of the three variables aligned to, respectively, conceptual models 1 and 2. For interpreting the Pearson’s correlations between the variables the suggested ‘rule of thumb’ of assessing the effect size by Cohen (1992) has been used. There is no correlation when r ≤ 0.1, a small correlation when 0.1 ≤ r < 0.3, a medium correlation when 0.3 ≤ r < 0.5, and a large correlation when r > 0.5. The same ‘rule of thumb’ has been applied for the effect size for the R2 found in the regression
outputs, which are shown in the main analysis. In addition, when a correlation is significant, it does not explicitly lead to important or meaningful effects. Especially, when there is a big number of observations (N; Field, 2009 p. 56), which is the case for the dataset observed. Therefore, relying on the found correlations is much more important.
For conceptual model 1 it unexpectedly turns out that time spent on leisure activity and happiness are not correlated, r(5.865) = 0,002, ns. The correlations between time spent on leisure activity and leisure satisfaction and the correlation between leisure satisfaction only turned out to be little correlated, respectively, r(5.865) = 0.144, p < 0.001 and r(5.865) = 0.217, p < 0.001.
For conceptual model 2 it becomes clear that a big correlation has been found, as forecasted in the theoretical framework, for the variables work satisfaction and happiness, r(2997) = 0.711, p < 0.001. However, unexpectedly, the correlation between leisure satisfaction and happiness, and work satisfaction and leisure satisfaction seems to be small, respectively, r(2997) = 0.217, p < 0.001 and r(2997) = 0.171, p < 0.001.
Table 5: Means, standard deviations and Pearson’s correlations of conceptual model 1
Variable Mean SD 1 2 3 1 Happiness 6,85 2,24 2 Leisure satisfaction 7,97 1,99 0,217* 3 Time spent on leisure activity 153,14 129,58 0,002 0,144* Note: N = 5865; *: p < ,001
23 Table 6: Means, standard deviations and Pearson’s correlations of conceptual model 2
Variable Mean SD 1 2 3 1 Happiness 6,85 2,24 2 Leisure satisfaction 7,97 1,99 0,217* 3 Work satisfaction 7,41 2,06 0,171* 0,711* Note: N = 2.997 *: p < ,001
4.2 Main analysis
4.2.1 Conceptual model 1The following section provides the results of the main analysis for conceptual model 1, which contains the direct effects and are covered by, respectively, model 1, 2, 3, and 4. First, model 1, with an explained variance of 0 percent clearly shows that the time spent on leisure activities causes no increase on happiness (β =0,000, ns.). This result indicates that respondents who undertake leisure activities longer are no happier than persons who undertake a leisure activity that lasts less long. The second regression analysis reveals that this model explains, unexpectedly, just variance of 1.6 percent (R2 = 0,016). Consequently, the time spent on a
leisure activity does not have a positive effect on the leisure satisfaction one experiences from that particular activity (β =0,000, ns.). Furthermore, model 3 shows that there is an explained variance of 5,6 percent and it turned out that leisure satisfaction has a weak positive effect on happiness (β =0,224, p < 0,001, R2 = 0.052). This weak effect indicates that respondents who
experience more satisfaction during leisure will also experience a bit more happiness during the day. Finally, Model 4 with 5,6 percent explained variance reveals that time spent on one leisure activity still does not affect happiness (β =-0,31, ns.) while the effect of leisure satisfaction remained significant (β =0,395, p < 0,001, R2 = 0,05). This verification of the made
predictions provides no evidence for the full mediation model. Still, involving leisure satisfaction to predict happiness seems to positively affect happiness. However, this is not caused due to an increase in time spent on that particular leisure activity. All four models can be obtained from table 7.
Table 7: Regression results for all four models.
Model 1 Model 2
Dependent Variable Happiness
Life Satisfaction
Coefficient SE Beta Coefficient SE Beta
24 Time spent on leisure activity 0,000 0,000 0,002 0,002 0,000 0,000 Leisure satisfaction R2 0,000 0,016 Model 3 Model 4
Dependent Variable Happiness Happiness
Coefficient SE Beta Coefficient SE Beta
Constant 5,260* 0,080 4,737* 0,121 Time spent on leisure activity -0,001 0,000 -0.31 Leisure satisfaction 0,224* 0,11 0,229 0,295* 0,016 0,240 R2 0,052 0,056 Note: N= 5.821, *p <0,001 4.2.2 Conceptual model 2
The following section shows the results of the analysis for conceptual model 2. It includes both the direct effects and the interaction effect. This can be depicted in table 8, both effects are covered by, respectively, model 1 and 2. First, the main effect of work satisfaction on happiness was determined. It became clear that, unexpectedly, work satisfaction has a small effect on Happiness (β = 0,042, ns.). However, leisure satisfaction turned out to have a positive effect on happiness (β = 0,215, p < 0,001, R2 = 0,044). So, this indicates that people who experience
leisure satisfaction are also the one’s that tend to be happier. The second model shows expectedly an interaction effect of leisure satisfaction on the relation between work satisfaction and happiness (β = 0,079, p < 0,001, R2 = 0,083). Therefore, the proposed hypothesized
moderating effect of leisure satisfaction on the relation between work satisfaction and happiness is confirmed. In other words, people that are satisfied with leisure activities are the ones that can derive more happiness from the work satisfaction experienced. Furthermore, model 2 (R² = 0,083) explains much more variance compared to model 1 (R² =0,044). To end, the interaction is visualised by creating an interaction plot that shows the direction of the effect, and can be adopted in figure 1. This figure reflects that there interaction effect exists, since the graph lines of both low and high leisure satisfaction are not parallel to each other (Field, 2009 p. 444).
25 Table 8: Regression Results of main (model 1) and interaction effect (model 2) of Work satisfaction
and Leisure satisfaction on Happiness.
Happiness (DV) Model 1 Model 2
Coefficient B SE Beta Coefficient B SE Beta Constant 4,773* 0,182 6,624 0,044 Leisure satisfaction 0,215* 0,031 0,178 0,400* 0,034 0,332 Work satisfaction 0,048 0,029 0,042 0,151* 0,030 0,132 Leisure satisfaction * Work satisfaction 0,079* 0,007 0,302 R2 0,044 0,083
Note: Dependent variable is Happiness. For both models we used the mean centered variables, N=2.997; *: p < ,001
26 Figure 1: Interaction plot of the moderator leisure satisfaction on the effect of the independent variable
work satisfaction on the dependent variable happiness. Dependent variable happiness consist of a10-point Likert scale. ‘‘Low’’ = 1 standard deviation below the mean. ‘‘High’’ = 1 standard deviation above the mean.
4.3 Additional analyses
Besides the tests for the hypothesis, some additional tests are executed. For the hypotheses, unpaid work was excluded from the statistical tests. For leisure, no exceptions were made. However, literature discusses many differences between leisure activities concerning the effect on happiness.
4.3.1 Motivation
Sakkthivel (2011) found evidence that monetary leisure behavioural activities, such as visiting night clubs, shopping and scuba diving, provided less happiness, compared to nonmonetary leisure substitutes (e.g. cycling, fishing and watching television). Moreover, Della Fave and Massimini (2003) stress that structured leisure, or serious leisure as defined by Stebbins (1997), is the crucial aspect of achieving an increase in happiness. Another research found differences in satisfaction among the following four groups: socializing (e.g. go to the theatre), exercising (e.g. play tennis), passivity (e.g. watching television) and home leisure (e.g. reading books). These sub-divided groups are built upon different research results, as the other two examples already mentioned. Still, research reports different size effects on satisfaction de-rived from it. It is therefore worthwhile to create dummies according to the different sub-divi-sions of leisure that are distinguishable in the dataset. Therefore, the next section will analyse if there are any differences in the satisfaction experienced during those leisure activities.
Low work satisfaction
High work satisfaction
6
6,5
7
7,5
Hap
p
in
ess
Low leisure satisfactionHigh leisure satisfaction
27
4.3.2 Results
Table 10 shows the results of specific leisure activities that one undertakes alone or together, which has been proposed by Schmiedeberg and Schröder (2017) as a crucial difference in affecting leisure satisfaction and happiness. First, it is indeed found that satisfaction during leisure is much higher when doing the activity with someone (M = 8,57, SD = 1,53), instead of undertaking alone (M = 7,71, SD = 1,87). Also, it is examined that this difference was the most for relaxation activities, where the mean difference was equal to 0,86 (M = 8,50, SD= 1,52, M = 7,64, SD = 1,86). Moreover, of all activities going out with someone was observed to give the highest satisfaction (M = 8,91, SD = 1,44), where relaxation activities undertaken alone gave the lowest satisfaction (M = 7,64, SD = 1,86).
Another additional test that has been executed is the difference in experienced leisure satisfaction during the day. The outcomes can be depicted in table 11. In accordance to these outcomes it has been found that leisure in the morning was considered with the lowest satis-faction (M = 7,52, SD= 2,41). In contrast, leisure in afternoons were rated the most positive by our respondents (M = 8,01, SD = 2,13). Also, doing exercising activities, such as cycling and sports, in the morning were considered with the best satisfaction of all leisure activities (M = 7,65, SD = 2,67). However, it was found that going out activities were enhancing the leisure satisfaction the most in the afternoon and evening (M = 8,36, SD = 2,22, M = 8,36, SD = 2,50).
Table 10: Means, standard deviations of specific leisure activities, undertaken alone or together.
Participating activity Alone Together
N Mean SDa N Mean SDt Relaxation 18.954 7,64 1,86 18.980 8,50 1,52 Exercising 2.867 8,08 1,90 3.640 8,69 1,61 Going out 329 8,30 1,74 2.568 8,91 1,44 Association 160 8,19 1,89 874 8,59 1,63 Leisure on average 22.658 7,71 1,87 26.536 8,57 1,53
28 Table 11: Means, standard deviations of specific leisure activities, undertaking in the morning, afternoon
or evening.
Time of day Morning
6:00 AM –12:00 PM
Afternoon
12:00 AM –18:00 PM
Evening
18:00 PM –24:00 PM
N Mean SD N Mean SD N Mean SD
Relaxation 5.486 7,47 2,24 12.131 7,93 2,04 19.024 7,90 2,08 Exercising 2.179 7,65 2,67 2.851 8,18 2,30 1.538 8,05 2,50 Going out 170 7,17 3,38 1.553 8,36 2,22 990 8,36 2,50 Association 254 7,34 3,10 394 7,48 3,04 428 7,89 2,72 Leisure in general 8240 7,52 2,41 17061 8,01 2,13 22155 7,94 2,13 Note: for all variables minimum value is 1,00, and maximum value 11,00, activities that covered more than 1 time period are excluded from the analysis.
5. Discussion
5.1 Summary
This study firstly examined if leisure engagement could enhance leisure satisfaction and in turn predict happiness. More specifically, it has been investigated whether the relationship between leisure engagement and happiness is mediated by enhancing leisure satisfaction. For this purpose four regression models were executed. The first regression model was not confirmed at all; undertaking leisure activities longer does not lead to higher levels of happiness. The second regression model could not identify that leisure engagement undertaken by the respondent resulted in higher levels of satisfaction derived from the leisure activities. However, the third model showed that leisure satisfaction had a weak positive effect on happiness. Still, the fourth hypothesis was not confirmed; this means that people are not happier when engaging in leisure activities, and this is also not explained when leisure satisfaction is added to the model as the mediator.
For the main question, it is investigated if satisfaction derived from leisure could positively enhance the relationship between work satisfaction and happiness. Specifically, it is tested whether the relationship between work satisfaction and happiness is moderated by leisure satisfaction. For this, two models were proposed; one for the direct effect of work satisfaction on happiness, and the moderating effect of leisure satisfaction on the relationship between work satisfaction and happiness. Initially, it turned out that work satisfaction had no direct effect on happiness. It appeared, as expected, that leisure satisfaction had a positive interaction effect on the relation between work satisfaction and happiness.
29
5.2 Unexpected results
5.2.1 Leisure engagement and happiness
The expectation that that people are happier when undertaking leisure activities is not confirmed. In contrast to this, the study of Kuykendall et al. (2015) investigated the same relation and found that leisure engagement enhances happiness. Not finding this relationship could possibly be attributed to the fact that Kuykendall et al. (2015) includes diversity and frequency in their model and not only the time spent on the activity as this study does. To explain this, Mannell and Kleiber (1997 p. 55) argue that diversity and frequency are based on activity-based measures and time spent on activity is measured by quantity measures. This has been identified as a key distinction, since engaging in diverse activities frequently is much more prone to decreasing levels of happiness (Kuykendall et al., 2015). To specify, individuals with more diversity in their leisure are more able to substitute particular activities when facing temporary or enduring life changes. It is therefore less likely, by protecting their leisure domain, that someone’s happiness is affected negatively (Hutchinson et al., 2003).
Besides that, Stavrakakis (2015) found that sporting, as a particular leisure activity, was not associated with higher levels of happiness for people with depression or bad mood symptoms. It even appeared that sporting for some respondents was related to a negative effect on their mood (Stavrakakis, 2015). This study observes a sample with an overall happiness below the average of the inhabitants of the Netherlands. Therefore, it is defensible to assess that these respondents have bad moods more often, which could affect the happiness from leisure engagement differently.
5.2.2 Leisure satisfaction and happiness
The finding that leisure satisfaction positively relates to happiness is approved, since the relation was found to be weakly positive. This is in line with the findings from several scholars (Kuykendall et al., 2015; Headey et al. 1991; Diener 1984).
Still, only a weak relation has been found in contrast to the former scholars that found a strong relation between both variables. One clarification for this difference is that this study used the structural approach, where the research defines what can be labelled as leisure. The problem occurred that some groups of activities were not solely based on leisure activities and could therefore not be included in the analyses. For example, there was no possibility to include ‘sex’ in the analysis, since the activity it fell under was sleeping, which also entails other sub-activities, such as ‘being sick’. This can be seen as a limitation, since it appeared that people that have more and better sex are more likely to be happier (Cheng & Smyth, 2015).
Furthermore, Kuykendall et al. (2015) was indeed able to link generalized term leisure satisfaction to happiness. However, this meta-analysis by Kuykendall et al. (2015) used a vast
30 majority of articles that were focusing just on retired individuals. For them, it is possibly easier to find relations between leisure satisfaction and happiness. External negative influences of work, such as the lack of psychological detachment from work, are not present for retired persons.
5.2.3 Work satisfaction and happiness
Another unexpected finding was that there appeared to be no relation between work satisfaction and happiness. This is in contrast with Tait et al. (1989) who found a strong significant relation. It appeared that the work activity was most of the time considered to be less than just 2 hours (42%, table 3). This is an unusual outcome, since the average work week is equal to almost 27 hours (26.95 hours, table 1a). It is therefore likely that respondents have divided work into multiple parts over the day, because of lunch breaks or changing work activities during the day, for example. These breaks can result in distortions in reviewing the satisfaction derived from work as a whole.
5.2.4 Other factors
There is a possibility that other factors have influenced the results of the conceptual models. For instance, 79% of the respondents of ‘de Gelukswijzer’ are women. Prior research conducted by Lunderg and Frankenhaeuser (1999) found that females are less capable to unwind psychologically from work, since they take on more responsibility regarding household tasks (Sonnentag and Bayer, 2005). These findings could have a repercussion on the entire analysis model.
5.3 Interpretation of the findings
As noted, the results have shown that leisure satisfaction contributes to higher levels of happiness. A reason for this is that leisure can enhance happiness by the five psychological mechanisms proposed by Diener and Newman (2013): detachment-recovery, autonomy, meaning, mastery, and affiliation. This study can therefore advice, in accordance with the current literature and our findings, to undertake leisure activities.
Some leisure activities are more satisfying than others. According to the additional analyses one can better investigate in leisure activities that are undertaken together with someone, since this gives much more leisure satisfaction. In addition, to enhance your satisfaction the most one could undertake exercising activities in the morning, and going out activities, such as shopping and cultural activities, in the afternoon and evening.
The latter, from the results it can concluded that leisure satisfaction has a positive moderating effect on the happiness derived from work satisfaction. In other words, if one can enjoy their leisure, the more satisfaction through work is translated into overall happiness,
31 since work constraints, such as emotional exhaustion and poor sleep, will not be translated into one’s happiness in daily life.
5.4 Limitations and future research
On top of the limitations already mentioned in the section of unexpected results there are some more limitations. One point that has not been examined in this study is the causality of the relationship between the variables. This study is cross-sectional, which has a consequence that sequences of events cannot be measured (Levin, 2006) thus with this research design is it not possible to draw conclusions about causality. This is a pity, since many scholars aim for causal relations in the research area of happiness (Lyubomirsky et al., 2005b). To include causality in the model, future research could utilize the data in a longitudinal research design or design quasi-field experiments or lab experiments to, as these are used to establish causal effects among variables (Bryman, 2015). One article that did accomplish to find a causal relationship between leisure satisfaction and happiness was a longitudinal study of Kuykendall et al. (2015). To do this, the following hypothesis was stated: Leisure satisfaction at Time 1, will predict SWB at Time 2.One time aspect that is considered to be interesting in the area of happiness and time is the changing mood of people during the week (Cranford et al., 2006). To digress, one’s happiness is likely to increase when the weekend starts, and at the end of the weekend, the happiness tends to decrease, since work or school are about the begin (Cranford et al. 2006; Helliwell & Wang, 2014). The data provided by ‘de Gelukswijzer’ contains the date variable, which means that the time satisfaction at time point 1, can possibly predict the happiness at time 2. Therefore, subsequent research could make use of the same data in order to research this area of interest. Especially since the dataset can be considered as large. Because of this it is easier to validate the causal relations, which are sought after in happiness studies. (Shmueli and Koppius, 2011; Lyubomirsky et al., 2005b).
Moreover, large samples have the advantage of not losing statistical power in their model when involving multiple control variables (Forman et al., 2008). Future research in this area could therefore include these variables, such as gender and age. This can explain the unexpected results that have been identified regarding gender. Another control variable that should be added is age, since this has been identified as a potential influencer on happiness. Blanchflower and Oswald (2008) identified that happiness was U-shaped through the life cycle. One’s happiness reaches a minimum when they are around middle age and is considered higher among adolescents and older individuals. Taking this into account, subsequent research could focus on just one cohort.
The large sample size of the data is possibly a limitation. This large number of observations of both models (N = 8.118 and N=7.651) results in a large statistical power. The consequence of this is that even small, sometimes negligible effects are still significant. One
32 could question if the small moderation effect found to be statistically significant in this research, is relevant enough to be considered significant in real life. Future research has to establish this.
Lastly, this research defined happiness as the pursuit in life. However, this view on happiness is considered by some research as being insufficient (Baumeister et al., 2013), since it is not linked to help others in need. This so called pursuit of meaning is what sets humans apart from animals. It is important to take this into consideration when interpreting the results and translating the research to practice. Assumptions about what happiness entails should be made very clear.
5.5 Concluding thoughts
According to the SHM, the intentional activities you undertake determine your happiness for 40 percent. The activities you undertake are therefore very important when striving for happiness. However, the results of this study indicate that there is no relation between leisure engagement and happiness, nor was it explained by leisure satisfaction. Besides the time spent on the activity, Kuykendall et. (2015) included the diversity and frequency of leisure activities. They found, in contrast, significant relationships between all variables. Subsequent research should reveal if these other factors can explain the model. Still, the content of activities turned out to be very important, since the satisfaction derived from it was different. This was done by additional tests for the complete sample of 8.188 respondents. It made clear that to arrange leisure satisfaction one should engage in going out and exercising activities. For all activities, the respondents were more satisfied when undertaking leisure activities with someone else. Besides that, research has focused more and more on the outcomes of leisure satisfaction and work satisfaction, since detachment from work seems a requisite to be happy. This study found that more leisure satisfaction could influence the satisfaction derived from work on happiness positively. In conclusion, one should invest in leisure, since it will definitely help to translate satisfaction derived from work into daily happiness.
