The concept of subjective well-being: its origins an application in tourism research: a critical review with reference to China

Abstract

Purpose

The subject of “well-being” has attracted attention from tourism scholars, but differences and misuses in approach have meant that academic contributions and knowledge accumulation to the tourism literature remain relatively little. This paper attempts to clarify the theoretical source of subjective well-being, and critically reflect on the problems existing in the study of well-being when applied to tourism. It is suggested that subjective well-being belongs to the category of “quality of life” and has multiple philosophical foundations and theoretical sources including theories of hedonism, expectation, happiness and various itemised lists of emotions. A hybrid research method is suggested when applying the concept to tourism.

Citation

Liang, Z., Luo, H. and Liu, C. (2021), "The concept of subjective well-being: its origins an application in tourism research: a critical review with reference to China", Tourism Critiques, Vol. 2 No. 1, pp. 2-19. https://doi.org/10.1108/TRC-04-2020-0009

1. Introduction

The purpose of this paper is to trace the evolution and research trend of the concept of subjective well-being (SWB) and critically reviews its use in the tourism literature, and to provide a summary of key issues and research agenda relating to the concept and its application in tourism. It also provides a guidance to the various scales used and makes suggestions referring to the use of concepts of SWB when being applied to tourist experiences. SWB is a psychological construct related to the quality of life (Maddux, 2018; Fang and Feng, 2009) and an important research field in its own right. The scope of quality of life is broad and includes several objective indicators, such as the comprehensive quality of life scale (ComQol) (Cummins, 1997) and subjective indicators, such as SWB, happiness, contentment and self-actualisation (Maslow, 1954). In fact, well-being can be divided into SWB and Psychological well-being (PWB) (Qiu and Li, 2012). The former generally refers to people and their quality of life and its emotional and cognitive evaluation, while the latter emphasises people’s potential and capacity to achieve well-being (Ryff, 1989). SWB is the positive attitude and positive feelings produced by individuals through a comparison of actual and ideal living conditions (Diener et al., 1997; Diener, 2000), and it is a subset of positive psychology. It includes periods of people’s lives at one moment or period (e.g. one year) formed by an event, creating feelings of life satisfaction, achieving a sense of fulfilment as well as meeting desired levels of work, family, marriage and other associated areas of emotional lives, reactions and evaluations (Diener et al., 2003, p. 404). That is, there is a sense of SWB as to the reality of possessing a good life being aware of comparison as well as having an emotional response to enjoyment (hedonic level of affect) (Veenhoven, 1984).

Some classic studies of early SWB remain influential to this day. For example, the Pattern of Human Concerns and the establishment of the “Cantril ladder” (Cantril, 1965) remain highly respected and are used in the annual United Nations’ World Happiness Report as reported by Helliwell et al. (2017). Among the earliest reviews of SWB in multiple subject areas and attempts to propose universal definitions were Wilson (1967) and Campbell et al. (1976). Wilson’s study “Correlates of avowed Happiness” has been hailed as a watershed study of SWB research by Yan et al. (2004) for its considered qualitative and philosophical approach and for its scientific treatment of well-being characteristics (Wang et al., 2010). The article Subjective well-being by Diener (1984) made a significant contribution. The citation rate of this paper provided by Google Scholar exceeds 15,000 at the time of writing (December 2019), and it is one of the most cited papers in this field. Diener has subsequently provided guidance to many students and led research teams that have both inherited and developed further research results (Larsen and Eid, 2008).

In recent years, the concept of SWB has been extensively used in tourism research involving residents (Nawijn and Mitas, 2012; Liang and Hui, 2016; Chi et al., 2017; Ivlevs, 2017; Liang, 2018), tourists (McCabe and Johnson, 2013; Su et al., 2016; de Bloom et al., 2017) and tourism related employees (Liang et al., 2015; Lee et al., 2016). As more tourism SWB studies emerge, so too many problems become more prominent. First, the definition of SWB used in the tourism study is unclear. Some studies confuse different disciplines or philosophical foundation of happiness or well-being. Second, various methods have been used to research SWB with many different measurement scales. There are also derived, simplified, contextualised and even so-called localised or contextualised scales, which easily lead to misuse by researchers. Third, many studies of SWB in the tourism field have not found the “particularity” or specificity of “tourism”, but just provided new evident for existing studies of well-being. There was less theoretical contribution. So, what are the advantages and particularities of tourism in the study of SWB? The discussion of the above issues is the purpose of this article, with a view to providing guidance for the better-informed use of SWB in tourism research.

2. The origin of subjective well-being and its philosophical basis

SWB arises from people seeking a good life, but such thinking is usually premised on moral or ethical theories and studies, involving a series of questioning about values that, as stated by Socrates, teaches “a man how to live.” How one ought to live remains a classic problem for this approach. Although some scholars use the term “SWB” to replace the “Good Life”, there are obvious differences between the two concepts. A “Good life” includes sense of well-being and virtue and even aesthetic or other judgements of the value of “good” and “well-being” but often in essence add nothing of significant worth to the debate. Parfit (1984) notes that well-being theory is divided into three: hedonism theory (the hedonistic), expectancy theory (the desire) and objective list theory. In a similar manner, Haybron (2008) notes five aspects to theories of happiness: hedonism theory, desire theory, authentic happiness theory, eudaimonistic theory and objective list theory (Table 1).

Hedonism, desire and authentic happiness theories have already been used in tourism research. The authentic happiness comprises subjective experience and individual differences as usually mentioned in the hedonism and desire theory of SWB (Sumner, 1996), and uses an alternative indicator – Happiness (Sumner, 1996), and uses an alternative indicator – Happiness to measure (Haybron, 2000, 2003). Happiness is a sense of being happy, has been conceptualised as hedonism (Kahneman et al., 1999) and life satisfaction (Ehrhardt et al., 2000; Wang, 2017) with various scales. Among these, the Life Satisfaction Scale (LSS) (Diener et al., 1985) is the most widely used in tourism research (Neal et al., 2004; Kim et al., 2013; Woo et al., 2015; Wang, 2017). Life satisfaction reflects the relative stability of the individual's overall quality over a period and is mainly based on personal rational judgment and is not easily disturbed by other external factors. It is more suitable for measuring the long-term well-being of residents in tourist destinations.

Reaching back to the classical times of Aristotle, well-being as eudaimonia, was said to be based upon perfection. The virtue required for well-being is the virtue that can best achieve the “ultimate” of being, that is, to be complete and be in a state of perfection (Chen, 2018). Well-being, from this perspective, becomes the doctrine of psychological well-being which underlines the well-cited psychological well-being scale (Ryff, 1989). However, it remains problematic to directly adapt Aristotle's philosophy to the empirical measurement of a scale of well-being, a stance that inhibits its widespread acceptance (Biswas-Diener et al., 2009; Huta and Waterman, 2014). On the other hand, there stands in stark contrast scales based on a list of objective items that lack a solid philosophical foundation.

The definition of SWB is chosen depends mainly on the research question and the value orientation of the researcher. In empirical research, the above-mentioned different definitions of SWB are highly correlated with the measures used in the scales. For example, hedonic and eudaimonistic SWB have been found to be correlated between 0.76 and 0.92 (Waterman, 1993; Waterman et al., 2008; Maddux, 2018). Alternatively, as previously stated, some scholars adopt a different definition of SWB premised on “moral level”. In the case of Kashdan et al. (2008) a moral hierarchy, and for Ryff (2017), a doctrine of well-being as the eudaimonic more than hedonic SWB. There are those who advocate eudaimonistic well-being and consider that a good life is an experience that transcends daily life (Kashdan et al., 2008), which is inherently consistent with concepts of “the way of life” of Xuncius (Sun, 2015). In this sense, SWB is not only a scientific concept but also a social construct (Fried, 2017). There may not be an “absolute” correct and universal SWB, but there is a belief in a defined SWB, and such a choice implies the existence of agreed values, morality and ethics. Therefore, the national, ethnic and regional differences used in the SWB measurements; hence, results are mainly due to the population represented by the samples derived from varying cultural circles (Diener and Diener, 1995; Joshanloo, 2014).

3. Subjective well-being research methods

3.1 Qualitative and quantitative study method

The study of SWB necessarily accepts the “subjectivity” of SWB. This subjectivity is reflected not only in the choice of research methods and measurement scales but also in the characteristics presented by the research design and data. People often respond to questions as to the nature of SWB not only with instantaneous emotional reactions but also with situational judgments (Kahneman et al., 1999). Respondents' answers are thus influenced by a range of contextual factors, and sometimes their replies reflect the unconscious (Leffert, 2016). People's evaluation of SWB is constantly changing and may be far from being constant. This means that short-term SWB is easier and more accurate to measure than long-term SWB (Diener and Larsen, 1984).

Subjectivity is not only an advantage when studying SWB but also a hindrance. It is common that the researcher using in-depth interviews to gain qualitative assessments of well-being or happiness in early studies (Thomas and Chambers, 1989; Wood and Johnson, 1989). Although qualitative materials may better represent the heterogeneity and diversity of interviewees' well-being than scale data, due to the subjective and flexible operation of qualitative research and the difficulty of transforming qualitative materials into comparable dimensions between cases, within tourism research qualitative research methods have gradually given way to more empirical quantitative studies with a few exceptions (Buzinde et al., 2014; Moscardo et al., 2013).

However, it needs to be stated that quantitative research itself cannot avoid subjectivity. Respondents tend to reply instantly and quickly. Most respondents do not accumulate considered judgments about SWB over a long time, yet it is also impossible to give rational answers by invoking long-term memory. The instantaneous answer is significantly affected by the context of the recalled or immediate visit, and the special events and memories of life (Schwarz and Strack, 1999). There are studies that have shown that changes in the measurement of SWB is very sensitive to external factors, and inter-individual measurement results about the same type of event may be quite different, being influenced by personal life events, even in the case of recent events (Diener, 1994). Therefore, to minimise such “noise” any quantitative evaluation of SWB must have clear standards and dimensions, strictly control other contextual factors, and researchers must explain the specific process and context of the investigation in detail (Schwarz and Strack, 1999). However, this is often overlooked in tourism research. Some tourism scholars appear to think that as long as a classic scale is used, any case scenario and group can be selected, and so there is little account provided for the study case scenario and research process.

Although SWB is constantly changing, it is possible that these are fluctuations over a period around something that is relatively stable in the intermediate time frame. At any one time, many people are concerned with only few life domains of their daily lives, despite their immediate external environment and living conditions are relatively stable. Their responses may be shaped by the certain life domains rather than all. This means that respondents’ assessments of SWB is not entirely derived from the review of their current situation but also from the comparison of reference systems such as goals, ideals, past, future and others (Michalos, 1985; Liang, 2018). In a mature and stable society, these life domains of reference rarely change, and SWB is relatively stable. However, in the tourism context of China, development has been rapid and significant changes in socio-economic status, quality of life disparities between residents, and hence changes in SWB have occurred over relatively a small number of years (Liang, 2018). Therefore, the rapid development of China’s tourism provides a special context for the study of SWB.

3.2 Self-report questionnaire research methods

Quantitative data derived from surveys often depends on self-reporting. In the past few decades, academia has developed a series of SWB self-reporting scales based on differing definitions of SWB that have been subsequently introduced into tourism research. Early SWB was measured using a single item. For example, Andrews and Withey (1976) developed the D-T (Delighted-Terrible) scale, containing only one item, namely, “How do you feel about your life as a whole?” Subsequently, the self-report scales ranged from single to multi-item questions and from a single dimension measurement to more complex dimensions. An example of the latter is the scale developed by Lyubomirsky and Dickerhoof (2010, p. 230), based on three underlying continua, of tendency to causal/construal theory and the dispositional and construal approach based on top-down and bottom-up approaches. The top-down approach is based on biological factors and the bottom-up emphases external, environmental factors. Those scholars who adopt the former approach believe that people's overall SWB is mainly affected by physical factors such as physiology, temperament and personality, rather than specific areas of life. Some studies have shown that genetics may explain some five to eight percentage of long-term and SWB (Lykken and Tellegen, 1996). The latter bottom-up thesis is based on the thesis that the main source of well-being depends on the environment in which they live (Diener, 1984) and reinforces the centrality of specific areas of life (such as home and work) in the evaluation scales. A third approach is bi-directional or inter-referential (Lyubomirsky and Dickerhoof, 2010, p. 230), but some suggest this approach should be avoided (Diener et al., 1985). However, some tourism studies have not noticed of this debate.

Measurement scales of SWB using self-report questionnaires can be classified into hedonism and eudaimonistic scales. The SWB scale of hedonism comprised positive affections (PA) and negative affections (NA) and separately assessed impacts on overall life satisfaction (Lucas et al., 1996; Arthaud-Day et al., 2005). The triple SWB model therefore suggests that positive and negative affections are independent variables and affect life satisfaction differently and the model is not bi-polar (Busseri and Sadava, 2011). Lischetzke and Eid (2006) further illustrates a triple SWB model (Figure 1). First, the overall quality of life and a field includes a plurality of measuring specific life domains; second, respondents' answers can change over time. The answer can be instantaneous reaction or accustomed expression. Third, the scale includes life satisfaction, as well as positive or negative affections. In tourism research, McCabe and Johnson (2013) attempted to adopt a logical triple SWB, and used the social survey of the British quality of life in specific areas along with measurements of life satisfaction The Satisfaction with Life Scale (SWLS) was used. The phenomenon of mixing or simplifying the use of scales is relatively common in tourism research. Some researchers measure the emotional, others only use life satisfaction (LS) that measures long-term well-being.

Affect Balance Scale is earlier developed emotional measurement scale, made of ten items (including five PA items and five NA items) composed of “yes” and “no” answers (Bradburn, 1969). Kammann and Flett (1983) developed Affectometer two scales which included ten aspects of emotion, each aspect contained four items, totally amount to 40 items. The most popular are the Positive and Negative Affect Schedule scales (PANAS) (Watson et al., 1988) and the Swedish Core Affective Scale (SCAS) (Västfjäll et al., 2012). The former adopting five-point Likert Scale, comprise ten PA items and ten NA items; the latter has two dimensions: valence and activation. Based on SCAS, some scholars established satisfaction with travel scale (STS), to measure travel-related SWB on the tour (Ettema et al., 2011). Diener et al. (2010) developed a very short scale to assess positive and negative feelings (SPANE) to replace the previous two classic scales. In addition, the Fordyce Happiness Measures (FHM) with only three items (Fordyce, 1997), and the NEO Personality Inventory which can measure the long-term emotional tone of respondents with 30 items (Costa and McCrae, 1980) and the subjective happiness scale possesses only four questions (Lyubomirsky and Lepper, 1999) is also used to measure the emotional dimension SWB.

Life satisfaction is usually measured using the overall life satisfaction scale or the specific life domains satisfaction scale, or both. The most commonly used scales are the Satisfaction with Life Scale (SWLS) (Diener et al., 1985) and the Personal Well-Being Index (PWI) (International Well-Being Group, 2006). Scholars such as Lee et al. (2016), Sirgy et al. (2011) and Lin et al. (2017) have used the SWLS in tourism research. Pavot et al. (1998) proposed the inclusion of timing in the Temporal Satisfaction with Life Scale (TSWLS). This scale contains 15 items and Ye (2007) suggests it prove suitable for Chinese respondents. In addition to the aforementioned Cantril Ladder (Cantril, 1965), the Multidimensional Students’ Life Satisfaction Scale (MSLSS) (Huebner, 1994) which is mainly used to measure children and youths, and the Life Satisfaction Scale (LSS), which is specifically used to measure the elderly (Neugarten et al., 1961) have been used for tourism research. A number of variations of the LSS exist and have been used – such as the Life Satisfaction Index A (LSIA), Life Satisfaction Index B (LSIB) and life satisfaction rating (LSR).

The scales of eudaimonistic SWB mainly includes the Personal Well-being Scale (PWS) and contains six dimensions (with a total of 32 items): self-acceptance, positive relationships with others, personal growth, purpose in life, environmental mastery and autonomy (Ryff, 1989). Many scholars have continuously revised and expanded the dimensions of the PWS scale to make it contextual. Waterman et al. (2010) developed a six dimensions scale with 21 items Questionnaire for Eudaimonic Well-Being (QEWB) which has completely different dimensions. The QEWB scale includes self-discovery, perceived development of one’s best potentials, a sense of purpose and meaning in life, intense involvement in activities, investment of significant effort and enjoyment of activities as personally expressive. Diener et al. (2010) developed a more concise Flourishing Scale, which includes only five dimensions, namely, social relations, meaning of life and purpose, self-esteem and optimism, competence and participation and interest in daily life. In addition, the Oxford Happiness Inventory contains 29 items including those about emotional experience and life satisfaction (Hills and Argyle, 2002) and the Personally Expressive Activities Questionnaire (PEAQ) (Waterman, 1993) are also often used. The scales of eudaimonistic theory often measures not the quality of life itself, but the ability of people to obtain a good life. This reflects the great difference in philosophical foundations of well-being between hedonism and eudaimonistic; therefore, a random mixing of the two scales is problematic.

3.3 Non-self-report scales and mixed research methods

Although SWB scales may be based on different theoretical foundations, the measurement results of different self-report scales are highly correlated, or even consistent. For example, the Life Satisfaction Scale and Life Satisfaction Index A were found to correlate up to 0.81 (Pavot et al., 1991); and the Coefficient of Correlation of Delight-Terrible and Fordyce Happiness Measures is 0.62 (Sandvik et al., 1993). In other words, although a single-item scale such as Delight-Terrible Scale cannot reflect the complexity of SWB, its measurement results are highly correlated with more complex scales. Therefore, the use of single-item scale may be effective proxies and retains scientific validity in some studies.

However, the correlation coefficient between the self-report scale and non-self-report scale has been found to be very low. In a sense, the self-report scale does not fully reveal SWB and still needs other research methods to supplement it Diener (1994). A series of non-verbal records can further help accurately assess SWB. For example, respondents’ gestures, postures, looks, tone, intonation and expressions can provide emotional information. Another example is the use of facial expression analysis, that is by analysing photos or videos of figures to measure the emotional facial expressions has been shown to be useful (Mendes et al., 2012). In addition to external indicators, physiological indicators such as hormones, cortisol, norepinephrine, and heart rate are also used. With the use of medical equipment such as brain motion monitors, skin monitors, and eye trackers, it is easier to obtain non-verbal information. In addition, psychology also provides a number of methods to measure SWB, including peer review or informant reporting by visiting neighbours, colleagues, friends and family (Pavot et al., 1991), observation methods, emotion detection (such as the quick recall of a happy experience) and cognitive evaluation (such as word filling, vocabulary recognition) (Sandvik et al., 1993).

Non-self-report is divided into instantaneous measurements (real-time measures) and recall measure (retrospective measures). Methods for instantaneous measurement include Experience Sampling Method (Csikszentmihalyi and Larson, 2014) and Ecological Momentary Assessment (Stone et al., 1999). Recall measurement methods include the Day Reconstruction (Kahneman et al., 2004) and the U-index (recording the proportion of respondents expressing happiness or dissatisfaction) (Kahneman and Krueger, 2006).

In recent years, the application of new technologies and big data has led to many new methods, but non-verbal records usually only focus on the respondents’ emotions and sentiment measurements, therefore possibly cannot fully reflect SWB. The self-report scales remain mainstream. Self-report scales appear to be reliable and effective (Diener, 1994); and it also seems that a number of SWB evaluations with different definitions are also relatively stable (Schwarz and Strack, 1999), even when compared across different time points and time spans, and in different situations (Diener, 1994). Mixed methods research is increasingly being used in SWB research to develop a more comprehensive understanding of SWB which was proposed by Diener (1994). However, it is still rare to find the use of hybrid research methods to assess SWB in tourism research.

4. Reflection on subjective well-being in tourism research

Nonetheless, research on SWB in the tourism research field has attracted a number of scholars. Many tourism studies have, on the surface, discussed issues of the quality of life (Budruk and Phillips, 2011; Uysal et al., 2012) in ways consistent with an exploration of SWB, including Neal et al. (1999) and Lloyd and Auld (2002) and other studies. SWB research in the tourism field has been mainly divided into two categories: one is to explore the impact of tourism activities and behaviours or experiences on the SWB of tourists (Neal et al., 1999; Lloyd and Auld, 20002; Dolnicar et al., 2012; McCabe and Johnson, 2013), and the second is the impact of tourism development on the SWB of local residents and employees (Liang et al., 2015; Liang and Hui, 2016; Bie et al., 2017; Liang, 2018). However, SWB in other disciplines when compared to tourism remains relatively more mature and tourism possesses a difficult dialogue with the mainstream studies of SWB, partly because it is difficult to accurately assess situations of SWB in tourism. Therefore, it is suggested that the study of SWB in the tourism field urgently needs to solve the following problems.

4.1 Clarify the conceptual basis of subjective well-being

Some studies suggest that the concepts of tourism, quality of life, SWB, life satisfaction and happiness that can be effectively replaced by each other (Kim et al., 2015), or indeed are even identical concepts. This can indeed work in certain situations. As mentioned earlier, there is a high correlation between the results of various hedonistic scales (Pavot et al., 1991; Sandvik et al., 1993). There may also be a high degree of collinearity between life satisfaction, positive emotions and negative emotions within SWB (Pavot, 2008), so that some studies have avoided issues of collinearity and simply used SWB scale with only one aspect of life satisfaction. This ruse can be found in tourism research. Taken together, this may mean that the differences between the results derived from varying scales in many scenarios are not significant. However, the scales based on different SWB measures, as previously discussed, have different philosophical foundations and theoretical origins. Thus, there is a danger, if scales are perceived as being interchangeable, of possibly critical nuances being missed. To this end, studies should include a discussion of the broad and varying definitions of SWB in tourism to, if not avoiding this problem, make the reader be aware of these issues. However, the danger here is that if there are too many differences a lack of comparability between research results may inhibit knowledge accumulation.

Other studies use SWB measures selected from multiple classic scales. This kind of random splicing of self-constructed “scales” neither has a unified philosophical basis nor can the measurements be used for comparative purposes. More research may modify the scale or lead to the development of new scales based on the actual situation of a case or population, as has been undertaken by authors such as Andereck and Nyaupane (2011), Chen et al. (2013), Liang and Hui (2016) and others. This practice is legitimate, it is suggested, if scale adjustment adheres to a consistent and transparent rationale.

4.2 Persistent problem-oriented research design

SWB research in the tourism field presents a certain pragmatic utility: first, part of the research is data – rather than question-oriented. As long as objective data are readily available, research on the “objective” assessment of quality of life may well be legitimate. Measures of SWB can have critical value in the absence of empirical data based on geographical or other similar measures, or when it comes to moving and random groups. In many cases, because of pragmatic managerial questions, many tourism researches mainly depend on the acquisition of data rather than the niceties of the research question. Although the availability of data needs to be considered in any study, unfortunately such a research approach will not benefit long-term conceptual or indeed managerial practice, or indeed possibly future data collection.

Second, SWB often comprises a set of indices based on a certain philosophical foundation. It usually has a stable structure and has undergone long-term empirical tests. Some tourism studies tend to “simplify” items, such as Ivlevs (2017) who used a separate item representing life satisfaction and happiness, and also de Bloom et al. (2017) who also used a single item measure. There are some studies where subjective experience has been reduced to the four subjective happiness scale entry of a title (Subjective Happiness scale) comprising only a measure of being emotional (Lyubomirsky and Lepper, 1999), to other similar shortened versions (Su et al., 2016; Lv and Xie, 2017). More tourism studies focus only on life satisfaction and SWB aspects (Woo et al., 2015), and lack of emotional measurement. Using only the emotional scale or only life satisfaction scale did exist in the field of SWB studies (Pavot, 2008), but can the measuring result of a highly simplified classical scale still reflect real SWB?

Third, SWB also has a temporal aspect. Day reconstruction method (DRM) uses the next day’s memory to form measures of well-being. Neither the use of instantaneous recovery or too long-term memory is thought applicable under this approach. The experience sampling method is usually used to collect transient emotional responses, and generally it is not thought suitable for exploring relatively long-term issues such as the impact of tourism development on the SWB of local residents in some tourist destinations. In the measurement of SWB of residents’ travel, how best to use the techniques of the “usual answer” are among issues considered as in Liang and Hui (2016). Chi et al. (2017) used a criterion of memory of the past two weeks in their study. For travellers, de Bloom et al. (2017) used a plurality of points of measure of SWB, while others argued for measurement scale of the transient response. The problem being addressed is an attempt to ensure that a potential wrong result where “the SWB of tourists before, during, and after the trip does not change significantly” is properly tested. Thus, if one only measures the transient response after, say, swimming, it is not appropriate to use the habitual answer scale if that does not consider an intervening event such as swimming.

Therefore, researchers should choose appropriate research questions for the concept and derived measurement scale, and design research programmes that overcome the difficulties of obtaining data, rather than simply consider the convenience of ease of access to respondents, obtaining data and then consequently engaging in random stitching, adjusting and simplifying scales.

4.3 Causal analysis to avoid random correlation

SWB is a key concept that is affected by many factors and many variables. Genetics (Lykken and Tellegen, 1996), personality characteristics (Strickhouser et al., 2007), extraversion/introversion, affinity, empathy, agreeableness, conscientiousness, etc., have all been shown to affect feelings of subjective happiness (Maddux, 2018). Tourism development or tourism experience is also thought to affect SWB (Liang and Hui, 2016). In tourism research, there are two main relationships between SWB and tourism. One is to treat tourism (sometimes including leisure) as a specific life domain of the quality of life (Chen et al., 2013; Dolnicar et al., 2012; Kim et al., 2015). The second approach is to treat travel as an external or independent variable (or disturbance factor), and to investigate the SWB as producing the impact (Kim et al., 2013; Liang and Hui, 2016; Guo et al., 2014; McCabe and Johnson, 2013; Woo et al., 2015). Different definitions of the relationship between tourism and SWB require different discussions on causal mechanisms.

Research on the effects of SWB on other constructs also requires caution. Discussion exist in the tourism research as to the role of SWB in the construction of residents’ attitudes towards tourism development (Woo et al., 2015; Liang and Hui, 2016; Chi et al., 2017), customer satisfaction (Su et al., 2016), career satisfaction (Lee et al., 2016), place identity (Lv and Xie, 2017), revisit intention (Kim et al., 2015) and other causal relationships between a series of constructs. However, as previously described, based on the well-established theory of SWB is the main measure of the ability of people to obtain a better life, and the use of a hedonistic scale as the main measure of life satisfaction and the emotional. The causal relationship between different definitions of SWB and other constructs is different. Some may be established, while others are not necessarily established. All causal relationships must have a theoretical basis prior to attempts to verify those relationships with data.

5. Conclusions and outlook

5.1 Study conclusions

Research on SWB in the tourism field has accumulated certain results. SWB is a psychological concept and a social construct, with different philosophical foundations and theoretical sources. SWB from different theoretical sources has developed a series of research methods and measurement scales. Each research method has its advantages and disadvantages and scale potential, conditions and caveats. At present, some tourism research may ignore the philosophical basis of the concepts of well-being and how they might be applied, leading to oversimplification, random splicing, arbitrary substitution, misuse, random use of scales, blind development of new scales and even lead to some proposed causal relationships that may have weak theoretical foundations. Many studies have been published and interesting conclusions obtained, but these conclusions may lack comparability and generalisation, and it is difficult to obtain effective theoretical contributions (Qin, 2017).

The shortcomings and problems of the application of SWB research in the tourism field are not unique to tourism. The collinearity of the various components of SWB, the lack of comparability between research results, and research that is too “subjective” or too contextualised are all shortcomings of SWB research (Pavot, 2008). In a sense, due to the diversity of the philosophical foundations of SWB and the non-uniformity of concepts, research on SWB in tourism may be difficult and not yield a widely recognised “perfect” study. The key is, to what extent do you agree with its value, to what extent does it reflect the moral and ethical inherent in the implied concepts, and to what extent are there defective theoretical and technical issues in anyone study. SWB is inherently subjective! For tourism researchers, future SWB research should clarify the conceptual basis of SWB, adhere to the question-oriented research design, avoid causal analysis of random associations, explain the case scenario and research process in detail and control external interference factors – such an approach will create more accurate and reliable research.

5.2 Research prospects

The background of China's rapid tourism development provides a special context for SWB research. On the one hand, the rapidly changing socio-economic environment allows researchers to study the diachronic evolution of SWB over a short period. The shorter the time, the less may be external disturbances, and the researchers can more accurately explore the few disturbance factors (including tourism) on SWB. In China, it has been possible to research emergent SWB and explore related factors for specific groups from points of origin, including students, the elderly and the mentally ill, thereby permitting the construct of a SWB model, identifying SWB mechanism and its determining factors, proposing theoretical stages of explanation, developing cross-cultural and cross-regional comparison and testing the impact of specific interference factors (such as tourism) on SWB. For Chinese tourism scholars, there are several aspects worthy of further studies:

First, the issue of cross-cultural studies of SWB and the construction of Chinese localisation theory. Many of the previously examined scales were developed in the context of European and American countries, and so do not fully conform to the Chinese understanding of well-being. The SWB of the Chinese may be different from that of the West in terms of basic ideas and overall structure (Bian and Xiao, 2014). Examples such as Liang and Hui’s (2016) application of Europe's Tourism Quality of Life Scale (TQOL) to the situation of China showed a need to adjust items that reinforced the importance of personal and family life, thereby improving the TQOL as an explanatory theory of motivation for Chinese residents to travel. In future, attempts to examine the effective explanatory power of SWB, especially with reference to China, it is more likely that new models and interpretations will emerge.

Second, there is a correlation between long-term and short-term SWB. Previous studies have independently discussed long-term or short-term happiness. However, a person's normal travel experience may have an impact on long-term SWB, and a certain travel experience may have an impact on instantaneous SWB. The overall evaluation of SWB by this tourist may then combine both the long and short term. So how do long-term and short-term SWB affect each other in tourism?

Third, to what extent can subjects such as the attitude of residents, tourists’ behavioral intention, and tourist satisfaction link with a series of variables related to SWB, to what extent might there be a causal relationship between travel and SWB, and what might be the nature of that relationship? Although there are reasons to believe there are linkages, and research in this direction has achieved some results, due to the problems and deficiencies mentioned above, there are not, it is suggested, definitive answers have not yet emerged.

Fourth, it does seem that mixed methods research would promise much in this area of SWB, and it is suggested that this is the best means of approach to be adopted to avoid the bias caused by the use of a single research method and a single data source.

5.3 Lack of research

This review may, nonetheless, possess two shortcomings. First, SWB is only the tip of the iceberg in the broader matter of quality of life research. This article is not an exhaustive review of all applications of SWB studies found in tourism-related research. However, it is thought that the key literature and studies of SWB cited in the tourism literature have been included, and the results from any omissions would tend to replicate those included in the review. What has been evidenced is that much of the philosophy, and many of the theories and concepts used in SWB have not been uniformly used and, particularly in China, there is an inconsistency in usage of the concepts and possibly even mistranslation errors due to cultural and linguistic differences. However, such problems are not an excuse for a failure to persevere with the concept as SWB and happiness of people lie at the core of current Chinese official policies, not only with reference to tourism but also with regard to the well-being of the country and its populace.

Figures

Figure 1.

Framework model for the measurement of subjective well-being

Table 1.

Theoretical sources of subjective well-being

Theory	Understanding of well-being	Theoretical flaw	References
Hedonistic theories	Define well-being as pleasure, and well-being consists in a subject’s balance of pleasant over unpleasant experience	The individual differences in joy levels are too large, and the lack of horizontal comparability among individuals makes it difficult to compare the level of happiness horizontally	Sumner (1996), Feldman (2004), Liang and Hui (2016), Liang (2018)
Desire theories	It is a satisfaction based on a comparison of actual and expected living conditions. The theory of informed desire is the most important branch of this approach. It is believed that expectations are based on a thorough grasp of information and environmental conditions	People's expectations are broad, and may even have nothing to do with happiness. It is difficult to conclude that people's satisfaction with expectations is happiness	Griffin (2000), Darwall (2002)
Authentic happiness theories	It is based on the living conditions of their own to fully understand and completely autonomous (autonomous happy to pursue the case) of (being Happy), including hedonism (Hedonism), overall life satisfaction (Life Satisfaction), emotional state (Emotional State) 3 Class scale	Widely used in tourism research. However, there is no emotional measurement of life satisfaction, and the time point and interval of life satisfaction measurement have a great impact on the results	Sumner (1996), Kahneman et al. (1999), Ehrhardt et al. (2000), Haybron (2000, 2003, 2005, 2007), Wang (2017)
Eudaimonistic (or “nature-fulfillment”) theories	It comes from people's self-improvement and self-realization, including not only moral character but also living ability	On the one hand, too much attention to individuals, groups and society as a whole is difficult to reflect the interests of happiness; on the other hand, is a philosophical concept, the lack of generally accepted measurement tool	Ryff (1989), Nussbaum (1993), Haybron (2008), Ruini and Ryff (2016), Chen (2018)
Objective list theories	Integrate the elements of the previous theory of happiness, and think that happiness comes from a series of lists, such as self-evaluation of knowledge, health, happiness, family, wealth, social (friend) relationships, etc	Without a unified philosophical foundation, the most critical philosophical issue is avoided-what is happiness	Murphy (2001), Haybron (2008)

Source: The authors summarised