Introduction

In an attempt to provide cross-cultural understanding of well-being indices, several validation studies have been carried out in different contexts but predominantly in Western and Asian settings (Disabato et al. 2016; Petrillo et al., 2015; Prilleltensky et al. 2015). There has also been a surge in the translation of instruments into different languages to increase cultural validity of the instruments. In the African context, the Mental Health Continuum (MHC), Satisfaction with Life Scale (SWLS), and Positive Affect and Negative Affect Scale (PANAS) have been validated (Keyes et al. 2008; Wissing and Temane 2008; Schutte and Wissing, 2017). There is also evidence of scholarly work comparing well-being indices in terms of levels and conceptualisation across different countries (Bojanowska and Zalewska 2016; Delle Fave et al. 2011; Delle Fave et al., 2016). Despite these advancements, two critical areas of inquiry remain unexplored. First is the minimal representation of countries with different socio-cultural histories that could affect the understanding of well-being. In addition, current cross-cultural studies do not adequately capture the cultural depth necessary for understanding the components of well-being. In an attempt to address these research gaps, the present study focuses on exploring the factorial validity of the SWLS and Meaning in Life Questionnaire (MLQ). In addition, we determined whether high levels of satisfaction and meaning in life do dovetail with Ghanaian adults’ conceptualisations of well-being, providing some indication of the extent to which hedonic and eudaimonic dimensions can be clearly distinguished in the Ghanaian context. We argue that conceptualisations of well-being matched against levels of satisfaction and meaning in life will provide a new understanding of the distinctions in hedonic and eudaimonic dimensions of well-being.

Hedonic and Eudaimonic Well-Being

Although still debatable, a common framework for distinguishing well-being conceptualisations, is hedonic versus eudaimonic perspectives (Ryan and Deci 2001; Sheldon 2013). Hedonia represents the maximisation of pleasure and minimisation of pain while eudaimonia refers to human flourishing and self-actualisation (Ryan and Deci 2001). Psychologists have extended definitions of hedonia to include how satisfying an individual evaluates their life to be. For instance, Diener’s (1984) and Diener et al. (2018) model of subjective well-being has been categorised into the experience of life satisfaction and the balance between positive and negative affect (see also Busseri and Sadava 2011). In terms of eudaimonia, two common components that most researchers agree on are personal meaning and growth (Ryan and Deci 2001). A widely referenced representative model of eudaimonia is the psychological well-being model of Ryff (1989) that emphasises positive functioning. Ryff and Keyes (1995) further argued that the structure of psychological well-being should include self-acceptance, environmental mastery, positive relations, purpose in life, personal growth, and autonomy. Other eudaimonic well-being perspectives include self-determination theory (Deci and Ryan 2008), eudaimonic identity theory (Waterman and Schwartz 2013), and human strengths (Linley 2013).

A criticism levelled against the distinction of well-being into hedonia and eudaimonia is the lack of discriminant validity between these constructs. Most studies have shown large correlations between the two constructs (Gallagher et al. 2009; Fredrickson et al. 2013). In addition, Disabato et al. (2016) highlighted that although hedonia and eudaimonia are generally expected to correlate differentially with other well-being constructs (e.g. curiosity, gratitude), there is a lack of strong evidence showing this. Their international study showed that a one-factor model presented acceptable model fit indices albeit weaker than the two-factor model of well-being with large correlations between hedonia and eudaimonia, implying that these dimensions might be capturing one overarching construct. However, Oishi and Diener (2014) argued that these dimensions might be clearly distinguishable depending on the context. Their study indicated that levels of meaning in life and life satisfaction were significantly different across wealthy and poorer nations with wealthier nations having higher levels of life satisfaction and poorer nations doing better on meaning in life. The distinction between hedonic and eudaimonic perspectives are supported by the findings of Joshanloo (2016). Our heuristic distinction between hedonic and eudaimonic perspectives for purposes of this paper is also informed by studies finding differential links for these perspectives with biological and neurological components (e.g. Fredrickson et al., 2013; Luo et al., 2017). This study intends to show whether contextually-embedded expressions of ideas on well-being link with high scores on satisfaction with life and meaning in life. To do this it is necessary to ensure that the tools of research, SWLS and MLQ, are applicable to the specific context of study.

Cross-cultural research is usually aimed at either exploring the same question in several cultures or determining differences across cultures (Epstein et al. 2015). In a similar fashion, newly created well-being items are tested in different contexts to determine whether they measure the same construct in different settings. It is also argued that validating existing questionnaires is more desirable than creating new ones as cross-cultural adaptation is faster and allows for the use of equivalent measure (Epstein et al., 2015). However, one of the common problems with the transference of measurement tools is the differences in meaning of items across varying cultural backgrounds. As a result, cross-cultural validation is needed to ensure that the new questionnaire functions as intended and shares similar properties as the original in the new context (Mokkink et al. 2010).

In the African context, with the exception of South Africa, well-being research is in its infancy and as in any growing field requiring empirical evidence, there is a need for appropriate research tools. We set out to determine whether existing well-being tools are valid measures of well-being in a historically, socially, and politically different context such as Ghana. As part of the Eudamonic Hedonic Happiness Inventory (EHHI) project, we have chosen two commonly validated scales: the SWLS and MLQ, and explored its psychometric properties in the Ghanaian context as a first phase of this paper. In the second part of the study, we sought to deepen our understanding of the dimensions of well-being by exploring the meaning of well-being as well as the extent to which these meanings reflect separate (hedonia and eudaimonia; Deci and Ryan 2008) orientations to well-being. Previous empirical work on these two measures are discussed in the following sections.

Satisfaction with Life Scale (SWLS)

Subjective well-being (SWB) is typically regarded as comprising satisfaction with life, positive affect, and infrequent negative affect (Pavot et al. 1991). As the cognitive component of SWB, satisfaction with life is referred to as people’s global assessment of their life. This assessment is based on ideas of an ideal life circumstance and tends to reflect the different domains of life including family, career, and health (Pavot and Diener 1993). The plethora of evidence on the factorial validity of the SWLS has emerged mainly from the West (Diener et al. 2013; Glaesmer et al. 2011; Zanon et al. 2014) and Asia (Mansoureh et al. 2011; Sachs 2003), with few emerging from South Africa (Khumalo et al. 2010; Wissing et al. 2008; Wissing et al. 2010).

The SWLS is a single factor, five-item measure of global life satisfaction (Diener et al. 1985). The initial validation report demonstrated the convergence of the SWLS with other measures of well-being and good internal consistency as well as applicability to a wide range of age groups (Diener et al. 1985). Further evidence of the scale’s utility in western contexts showed that the SWLS has high reliability scores (Diener et al. 2013; Eid and Diener 2004; Glaesmer et al., 2011; Slocum-Gori et al. 2009). Given that the SWLS is a short self-report inventory that taps into overall life satisfaction, correlates negatively with ill-being measures, and has been used in a variety of contexts, it is regarded as a useful instrument for measuring life satisfaction (Sachs 2003). A recent validation study in Brazil revealed that although SWLS was a valid tool with measurement invariance across gender, it did not show invariance when compared with groups from the US (Zanon et al. 2014). Among Hong Kong university students, the SWLS showed an over fit in a modified two-factor model and an acceptable fit with a modified one-factor model (Sachs 2003).

In the African context, the SWLS has been validated both in English (Wissing and Eeden 2002; Wissing and Temane 2008) and Setswana (Wissing et al. 2010). The English version was administered in urban areas and findings revealed that the SWLS was a component of general psychological well-being. The Setswana version was also found to be reliable and valid for use in Setswana-speaking groups (Wissing et al. 2010). Westaway and Maluka (2005) found that the scores on the SWLS demonstrated internal reliability among Black South African adults but there was however a problem with the fifth item on the scale (If I could live my life over, I would change almost nothing) which had a low factor loading. A study in Togo (French-speaking West African country) showed that although the scale demonstrated good psychometric properties, item one (In most ways my life is close to my ideal) had a poor factor loading (Sovet et al. 2016). This was explained as being due to the ambiguity of the wording for this item. Although the SWLS is a well-established instrument, which has been validated in a number of contexts, some of the problems identified in previous studies necessitates that we explore its psychometric properties in the Ghanaian context. Moreover, the authors are not aware of any existing literature that explored the psychometric properties of this scale in the Ghanaian context.

Meaning in Life Questionnaire (MLQ)

Experiencing meaning in life has been linked to optimal psychological functioning (Baumeister 1991) and is defined in terms of coherence, understanding of life, and purposefulness (King et al. 2006; Martela and Steger 2016; Steger 2012). Meaning in life comprises three components: cognitive (understanding who we are), motivational (pursuit of goals), and affective (feeling that the world makes sense) (Reker and Wong 1988). Steger (2012) further postulated that meaning in life includes presence of meaning and search for meaning. Presence of meaning explores the extent to which an individual perceives their life as significant, purposeful, and valuable (Steger et al. 2008). This is compared with search for meaning which measures the intensity of efforts directed towards understanding the meaning or purpose of one’s life. Whereas presence of meaning is a highly desired psychological quality, search for meaning is the “process” that gets one there (Steger and Kashdan 2013). In comparison to search for meaning, presence of meaning has been found to correlate positively with psychological well-being (Brassai et al. 2012; Steger et al. 2011). There had been inconsistent findings on the relationship between search for meaning and psychological functioning. Some evidence regards this dimension as dysfunctional (Steger et al. 2011), while others argue that its effect on well-being depends on the underlying motivation (Reker and Chamberlain 2000).

The MLQ was created because existing measures of meaning had limited construct conceptualisation, inconsistent factor structure, poor distinction between presence of meaning and sources of meaning, and excluded the search for meaning component (Steger et al. 2006). The MLQ has good psychometric properties with robust ability to study the dynamics of meaning in its interaction with other psychological constructs (Temane et al. 2014). For example, Doğan, et al. (2012) found that both search and presence of meaning were related to subjective well-being of Turkish students. The MLQ has also been found to have high reliability scores and emerged as a valid measure of meaning in Korean (Ju et al. 2013), Japanese (Steger et al. 2008), and Romanian (Brassai et al. 2012) samples. In a multicultural South African group, the MLQ was noted to be reliable and the two-factor structure was supported by the data (Temane et al. 2014). In a Rasch analysis of the MLQ among participants from South Africa, Australia, and New Zealand, Schutte et al. (2016) found the Presence of meaning subscale is insensitive at high levels of meaning while most participants were in this range. They also found item nine (my life has no clear purpose) to be problematic. However, there is still limited evidence for cross-cultural applicability as previous studies have been carried out in the West and Asia, with South Africa as the only African sample. The Ghanaian context presents a different socio-cultural context in which the experience of meaning in life might differ from the South African context.

Well-Being in the Ghanaian Context

Delle Fave et al. (2016) argued that previous studies on well-being are still lacking in clarity on the definitions of happiness (which is considered as a hedonic concept of well-being). In some studies, this term is used interchangeably with life satisfaction. Delle Fave et al. (2016) indicated that apart from conceptual ambiguity associated with the term happiness and the construct of well-being, cultural awareness in the evaluation of well-being is still underexplored. Cultural awareness is necessary because of the need to promote respect for diversity and to prevent the imposition of concepts from one context to another (Christopher et al. 2014).

Although the individual’s sense of SWB has been highlighted as critical to understanding the effects of the social, political, and economic landscape, this has not been given adequate scientific attention in the Ghanaian context (Pokimica et al. 2012). More than a decade ago, Ghana saw a relatively stable economic growth with macro-level indicators indicating improved living conditions (World Bank 2007) but this has not been evenly spread (Gyimah-Boadi and Mensah, 2003). Deeply ingrained cultural practices and values are also increasingly giving way to urbanisation and modernisation accompanied by an increased emphasis on religion. Given the lack of safety-net programs, individuals tend to find a haven in religious organisations in times of social and economic crisis (Pokimica et al., 2012). In the end, all of these affect the individual’s conception and experiences of well-being.

In a nationally representative sample of Ghanaian adults, the Afrobarometer survey revealed that religious affiliation was related to absolute SWB, while importance placed on religion predicted relative SWB (Pokimica et al. 2012). In other studies, similar findings on religious participation and SWB emerged (Addai et al. 2014; Asamoah-Gyadu 2005). Apart from religion, other contextual indices have emerged as predictors of SWB in Ghana. For instance, using the Core Welfare Indicators Questionnaire of Ghana, family reciprocity emerged as important for happiness (Tsai and Dzorgbo 2012) due to the economic security the family provides. A related theme, social capital, has also been linked to happiness in the Ghanaian context (Sulemana 2015). More specifically, data from the Afrobarometer survey showed that interpersonal and institutional trust predicted SWB (Sulemana 2015). Also worth mentioning is that underlying these predictors is the socio-economic landscape in Ghana where trust and dependency was a necessary buffer for economic challenges. However, these studies have been limited to only predictors of SWB and did not explore eudaimonic well-being. Glozah (2015) explored conceptions of well-being among Ghanaian adolescents and found that physical health and the presence of social support were important for well-being. Wilson and Somhlaba (2016) argued that underlying adolescents’ experience of life satisfaction was the realisation of goals, positive self-image, parental support, and positive peer relations. Common to these studies on well-being is the importance of religion, social relations, and physical health.

It is against this backdrop that we deemed it necessary to explore the underpinnings of well-being in the Ghanaian context and to determine whether various conceptions of well-being would reflect high scores on either hedonic and eudaimonic well-being, suggesting a clear distinction between these dimensions in the Ghanaian context. The aim of this study is thus two-fold: First, determining the psychometric properties of the SWLS and MLQ. Second, to match conceptualisations of well-being against levels of SWLS and MLQ to determine how contextual understandings of well-being are linked to and distinguish between hedonic and eudaimonic well-being in the context of this study.

Research Objectives

  1. 1.

    To determine the factorial validity of the SWLS and MLQ.

  2. 2.

    To qualitatively explore conceptualisations of well-being in the Ghanaian context.

  3. 3.

    To explore how high scores on the SWLS and MLQ as measures of hedonia and eudaimonia, respectively dovetail with Ghanaian adults’ expressed conceptualisations of well-being and provide distinctions between the two dimensions of well-being.

Method

A concurrent mixed-method design, which involves the gathering of quantitative and qualitative survey data simultaneously, was employed. The present study is linked to a larger international project: EHHI aimed at exploring dimensions of well-being across different countries (Delle Fave et al. 2011). However, the qualitative responses from the semi-structured questions reported on in this paper was not part of the original battery of questions for the EHHI and the sampling frame is larger than the scope of the EHHI.

Sampling and Participants

A method of convenience sampling was used to select 420 participants between the ages of 24 and 62 in the Greater Accra region of Ghana. Participants included employed individuals living in urban areas in Accra. Participants’ mean age = 41.32 (SD = 9.59) and included 230 men and 178 women, 12 participants did not indicate their gender. The majority of the sample were married or cohabiting with a partner (47.4%), 38.6% were single and 10% were separated, divorced, or widowed. Given the specific quota sampling used in the present study some groups were under-represented including unemployed individuals, those with no formal schooling and the aged.

Procedure

After necessary ethical clearance had been obtained from the University of Ghana Ethics Committee for Human Research (ECH 086 16–17), three research assistants and a project coordinator were recruited and trained in the administration of the questionnaire and data gathering. Participation was voluntary in all instances base on full informed consent obtained in written form. Research assistants who were trained fieldworkers recruited participants through face-to-face interaction. Questionnaires were self-administered and took approximately 60 min to complete. The research assistants removed the informed consent from the answer sheets, and numbered them correspondingly. Questionnaires were thus handled anonymously in coding and analyses. The coded responses were stored on password-protected computers. Both quantitative and qualitative data were collected. The qualitative semi-structured open-ended question on what participants see as ‘well-being’ was included in the battery of questionnaires. Data collection for the present study took place in 2017.

Measuring Instruments

The basic EHHI study (Delle Fave et al. 2011) included a battery of questionnaires that measured various facets of well-being and included scales assessing symptoms of psychopathology. The battery used for this study includes the EHHI Inventory tapping into definitions of happiness as well as most important valued things and goals as well as the reasons thereof, a checklist of the degree of happiness and meaningfulness experienced in different life domains. Other scales included measured basic psychological needs, depression and anxiety, and demographic information. For the present study, an open-ended question on well-being (what does well-being mean to you), SWLS, and MLQ were used.

Satisfaction with Life Scale (Diener et al. 1985)

The SWLS is a five-item scale that has been designed to tap into an individual’s evaluation of their life (see appendix 3). Each item was rated on a 7-point Likert scale ranging from 1 (strongly disagree) to 7 (strongly agree). Greater scores indicated higher levels of satisfaction with life. Reliability analysis showed that the SWLS scores were internally consistent with a Cronbach’s alpha coefficient of .88 (Guhn et al. 2018). In the present study, we obtained a Cronbach alpha of .84, which is considered as good.

Meaning in Life Questionnaire (Steger et al., 2006)

The MLQ is a 10-item scale designed to tap into the extent to which individuals assess their lives as meaningful. This instrument comprises two subscales: Presence of Meaning and Search for Meaning with five items on each scale. Examples of items on the scale includes “I understand my life’s meaning” and “I am seeking purpose or mission for my life” for Presence and Search subscales, respectively (see Appendix 3). Each item was rated on a 7-point Likert scale ranging from 1 (absolutely untrue) to 7 (absolutely true). In the original study, Steger et al. (2006) found reliability indices of .81 for MLQ-P and .84 for MLQ-S at Time 1 with similar indices at Times 2. In the present study, reliability indices were .86 for MLQ-P and .84 for MLQ-S.

Interview Schedule

An open-ended question tapping into coneptualisation of well-being was included. Respondents were asked to explain what well-being meant to them. The question asked was “What does well-being mean to you?”

Data Analysis

The quantitative data were analysed by implementing Mplus (Muthén and Muthén, 1998–2018) and structural equation modelling was performed to determine the factorial validity of the SWLS and MLQ. A confirmatory factor analysis (CFA) using robust maximum likelihood (MLR) estimation was conducted to test the hypothesised factor structure of all the scales. Missing data (missing at random) were handled by full information maximum likelihood estimation. The following model fit indices is presented: Comparative Fit Index (CFI; Bentler 1990). Tucker-Lewis Index (TLI; Schermelleh-Engel et al., 2003); Root Mean Square Error of Approximation (RMSEA; Steiger 1998); and the Standard Root Mean Square Residual; (SRMR). The criteria for an excellent model fit for these goodness-of-fit indices were considered to be CFI > .95, TLI > .95, RMSEA < .08 and SRMR < .08 (Hu et al., 1995). Reliability was evaluated using Cronbach’s alpha.

In order to further understand well-being in the Ghanaian context, thematic analysis was performed on the qualitative responses. All cases with missing data for the variable describing the meaning of well-being were removed from the datasets, resulting in a total sample of 336 cases. Sixty cases were randomly selected from the datasets using a random number generator function in Excel. In cases where the response was invalid, the next valid response was included. Selected cases were exported into an Excel spreadsheet for further analysis. The descriptions of well-being were then coded by one author and verified by another. All discrepancies were noted and addressed in a consensus discussion. We used a data-driven coding system where we looked for ideas and concepts on well-being from participants’ responses. The guiding coding rule was to determine how the data related to the research question on conceptualisation of well-being. After initial observations, we generated a coding scheme reflecting several dimensions of well-being including “good health”, “freedom”, “life satisfaction”, “availability of resources” “leisure” just to mention a few. Following the assigning codes, two of the authors searched for possible themes across the codes. Themes represent an overarching idea that explains two or more codes (Clarke & Braun 2013). After the initial assigning of themes (see Appendix Table 5), we reviewed these themes to determine whether further categories could be created and then a final list of themes was presented in the results section. To ensure trustworthiness of our findings we have provided background information on the context of study, a detailed account of the data analysis as well as excerpts from participants’ responses to validate the themes presented.

Using the Statistical Package for Social Science (SPSS, version 25.0), descriptive statistics were performed to determine the mean levels of satisfaction with life and meaning in life (presence subscale) from the SWLS and MLQ data. To investigate the extent to which participants’ conceptualisations of well-being reflect or distinguish levels of hedonia or eudaimonia as measured by life satisfaction and presence of meaning in life, mean scores for life satisfaction obtained from the SWLS and presence of meaning in life obtained from the MLQ were sorted into descending order. The top 60 scoring participants for each of the two measures (n = 120) with available data for the question “What does well-being mean to you?” were exported into an Excel spreadsheet for further analysis. Sampled responses of the participants’ descriptions of well-being were then coded by one of the authors into one of three categories (coding scheme), hedonia (well-being means a life full of happiness), eudaimonia (well-being is living a life full of peace and harmony), or an overlap (both hedonic and eudaimonic well-being and/or need fulfilment, i.e. being happy always and peace of mind). The codes were then verified by another author and all discrepancies were highlighted and discussed during a consensus meeting (see Appendix Table 4). The principle behind the coding was based on the application of the two broad categorisations of well-being (hedonic and eudaimonic) or an overlap of both which are two well defined constructs of well-being in existing literature and supported by empirical studies as indicated above. Discrepancies were recoded accordingly and the analysis were conducted using the final code descriptions as agreed upon by the two authors.

Ethical Considerations

Ethics clearance was obtained from the University of Ghana Ethics Committee for Human Research (ECH 086 16–17). Consent was sought through trained fieldworkers in order to ensure that participants do not feel coerced to participate in the study. In addition, participants were informed that participation was voluntary and they had the right to withdraw at any point in time during the study. Participants’ personal and identifiable information were excluded from the results presented. Any individual who needed to access the data for example research interns signed a confidentiality form.

Results

Validation of SWLS and MLQ

The SWLS Factor Structure

For the purposes of investigating the construct validity of the SWLS, a measurement model was tested for goodness of fit. The theoretically intended unidimensional factor structure was tested. We found that CFI = .971; TLI = .94; RMSEA = .075 [90% CI: .037, .116]; SRMR = .030 indicating an acceptable model fit. The standardised factor loadings ranged between .834, for item 2 and .501 for item 5. The model explained 63% (highest) of the variance for item 3 and 25% (lowest) for item 5 (see Table 1 for summary of factor loadings). We found support for convergent validity with presence of meaning subscale (R = .40) and the mental well-being scale (R = .37).

Table 1 Factor loadings for SWLS and MLQ
Full size table

MLQ Factor Structure

A two-factor structure including presence and search for meaning was tested. We found a poor model fit with CFI = .842 and RMSEA = .106. Further exploration of the factor loadings showed that item 9 had very low factor loading and the least variance explained (R2 = .12). A two-factor structure with the exclusion of item 9 was then tested which showed acceptable model fit indices: CFI = .921; RMSEA = .079 [90% CI: .062, .097]; SRMR = .052 (see Table 2). All items had factor loadings greater than .60. Item 8 had a largest amount of variance explained by the model (64%) (see Table 1 for summary factor loadings).

Table 2 Modification Indices for SWLS and MLQ
Full size table

Qualitative Findings on Conceptualisations of Well-Being

Participants’ expressed conceptualisations of well-being were qualitatively explored by coding their responses on what well-being means to them and organising the various codes into meaningful themes and sub-themes. The grouping of the codes into themes and sub-themes were informed by existing concepts in well-being literature in addition to being data-driven. Six themes were identified to provide a picture of which areas of well-being were most important in this particular context. The main themes included being free from worries, a positive state of being, material soundness, having meaning and purpose in personal life, being socially connected with family and friends, and experiencing well-being on multiple levels. Each theme and related sub-themes are discussed next.

Absence of Ill-Being and Stress

The absence of ill-being refers to well-being as a stress free state without problems such as stress, ill-health and material needs. Participants provided responses such as “well-being means living well devoid of problems” (participant 57B, male, age 35) and that well-being means “living a comfortable life, a life free of illness, and free of stress” (participant 4B, male, age 33). Participant responses also suggested that the availability of material needs and good health are aspects that enable one to be free of stress.

Emotional and Psychological Stability

This theme depicts well-being as a positive state of being, happiness, and psychological soundness. Well-being as a positive state of being was described as a positive emotional and cognitive state, for example “…being happy and sound in mind” (participant 55A, female, age, 30) and “Well-being to me means to be happy always and peace of mind” (participant 160A, female, age, 31). The responses comprising this theme demonstrate the relationship that health could have on happiness and peace of mind. Collectively, this sample described their well-being in both hedonic and eudaimonic terms tapping into the importance of soundness of mind as well as positive emotional experiences and a balanced orientation. Thinking of well-being in these terms reflect current theoretical notions and in a way blurs existing distinctions between the hedonic and eudaimonic dimensions.

Life and Sustenance

The theme life and sustenance portrays well-being in terms of material well-being, life satisfaction, and sound physical health. Material and financial well-being demonstrate the importance of having access to resources, being healthy and financially able to maintain a good standard of living and experience a quality life on multiple levels. Some responses combined aspects of health and finances for example, “when you are healthy and have money” (participant 190A, male, age, 39) and “living in good health without financial problems” (participant 62B, male, age, 30). Participants also described well-being as having a good standard of living which ranged from “…living a luxurious life” (participant 115A, female, age, 30) to “comfortable living” (participant 9B, male, age, 35), and “attaining the maximum standard of living” (participant 36B, male, age, 59). Closely related to this is the idea of well-being as the fulfilment of various needs. Responses representing well-being as need fulfilment demonstrate that this could imply more than merely material needs. For example, one participants indicated that “well-being to me is when you have everything you need complete” (participant 107B, female, age, 32). Salient in these findings is financial viability both for subsistence and for improved quality of life. Also noteworthy is the link drawn between maintaining good physical health and availability financial resources.

Meaning and Purpose

Well-being for some participants meant having meaning and finding purpose. This was communicated in responses such as “well-being is a meaningful life” (participant 147B, female, age, 31) and “well-being to me is finding a purpose in life” (participant 171A, male, age, 32). Other aspects that were captured in this theme revolved around the sense of freedom (“to be free to be yourself”; participant 21B, male, age 41), accomplishment (“to be fulfilled and satisfied in all areas of life”; participant 178A, female 60), as well as fulfilment (“to be healthy in order to fulfil my goal”; participant 174B, female, 53).

Social and Relational Well-Being

This theme represents the importance of interpersonal relations and social connectedness as a sense of well-being. For some, family health, happiness, and need fulfilment was the primary meaning of well-being. For example, one participant said well-being for them “is when I wake up each morning to see my family in good health and happiness” (participant 126B, female, age, 41) while another mentioned “well-being for me is when I have everything I need to make me and my family happy” (participant 126A, female, age, 52). The responses “to be at peace with God, others, and yourself” (participant 181B, female, age, 50) depict the meaning of well-being as harmony and the importance of interpersonal relationships. This finding dovetails with the notions of interconnectedness in a horizontal and vertical sense. Noteworthy, is the emphasis on maintaining relationships and being able to meet the needs of significant others and not just the presence of social support.

For some participants, the ability to make social contributions was the key meaning of well-being in terms of affecting and impacting others. This was expressed in the following response: “Well-being means a state of fulfilment in personal life and in the lives of others” (participant 67B, male, age, 42). Well-being is located both in the individual’s experience of fulfilment but also influencing positive change in the lives of others. This theme suggests that well-being in terms of need fulfilment extend beyond the individual to the social.

Multidimensional Well-Being

Well-being as a multidimensional concept comprises a combination of emotional, mental, psychological, spiritual, physical, financial, and social facets. Example responses depicting well-being as multidimensional include, “being psychologically, physically, and socially fit” (participant 24A, male, age, 30) and “when an individual is prospering in all facets of life, work, health, and even spiritual growth” (participant 86A, female, age, 31). In this conceptualisation of well-being, the emphasis is not on one domain or the other but multiplicity of factors that come into play when individuals assess how well their lives are. Well-being seems to be also presented in terms of interrelatedness on an internal intra-psychological level, which manifest across life domains. This view of well-being indicates that it will be erroneous to only employ a single or bifocal lens in understanding well-being.

Distinguishing Hedonia and Eudaimonia in Conceptualisations of Well-Being

Data analysis was conducted to examine how participants’ conceptualisations of well-being reflected their mean scores on the SWLS as a measure of hedonic well-being and the MLQ as a measure of eudaimonic well-being. Conceptualisations of well-being of a sub-sample of participants (n = 120) who scored highest on the SWLS (n = 60) and MLQ (n = 60), were categorised into one of three categories: hedonic, eudaimonic, or overlapping conceptualisations of well-being. Conceptualisations which referred to notions of happiness and need fulfilment, were coded as hedonic. Conceptualisations pertaining to a sense of purpose were coded as eudaimonic. Conceptualisations that included notions referring to both constructs were coded as overlapping. It was expected that individuals who scored high on life satisfaction were more likely to think of well-being in hedonic terms while those with higher scores on experiences of meaning, will have a eudaimonic orientation towards well-being. Those grouped as overlapping were an unexpected group whose definitions covered both hedonic, eudaimonic and associated needs fulfilments in conceptualisations.

As shown in Fig. 1, the majority of participants conceptualised well-being from a hedonic perspective (52.5%). This comprised 58% of participants scoring highest on the SWLS, (with a mean of 6.2), and 47% of participants scoring highest on the MLQ-P (average mean of 6.74). The eudiamonic category comprised the smallest overall percentage (22%). This comprised 23% of MLQ scores (average mean score of 6.75) and 20% of SWLS scores (average mean score of 6.22). The overlap category which represents conceptualisations of well-being which included both hedonic and eudaimonic descriptions, made up a total of 25.5% of which 30% comprised scores from the MLQ (6.81 average) and 22% comprised scores from the SWLS (6.22 average).

Fig. 1
figure 1

Associations between high scores on the SWLS and MLQ on the one hand, and categories of lay conceptualizations of well-being indicated as hedonic, eudaimonic and overlapping

Full size image

The results show that conceptualisations of well-being of participants scoring high on the SWLS largely reflected a hedonic description (n = 35), but conceptualisations of well-being of participants who scored high on the MLQ-P did not clearly represent a eudaimonic description (n = 14), although the number of cases that defined well-being from a eudemonic perspective among the high MLQ-P was negligibly greater (n = 14) than the number cases from high SWLS group (n = 12). Participants scoring high on the MLQ mostly described well-being in hedonic terms (n = 28) or in ways that included an overlap of hedonic and eudaimonic conceptualisations (n = 18). In addition, the fact that the overlapping definitions is higher for MLQ-P compared to SWLS suggests that the eudaimonic orientation of well-being to some extent include some hedonic experiences, but not the other way around.

Although there were no large differences in participants’ means scores on the SWLS and the MLQ, the mean scores for all three categories were slightly higher for the MLQ. (see Table 3). In other words, mean scores for meaning in life was generally higher although on face value (but not statistically significant) there were more cases of definitions of hedonic well-being from these participants. This buttresses the point that individuals who are hedonically orientated might still be concerned about existential issues such as meaning in life, blurring the distinction between the two forms of well-being.

Table 3 Mean scores for MLQ-P and SWLS for the Hedonia, Eudaimonia and Overlap categories
Full size table

Discussion

The overall aim of the present study was to explore the measurement and conceptualisation of well-being within the Ghanaian socio-cultural setting. In order to achieve this aim we tested the structural validity of two commonly used well-being scales, the SWLS and MLQ, and analysed qualitative descriptions of well-being to unearth unique understandings of the construct in the Ghanaian context. We also compared participants’ lay conceptualisation of well-being categorised as hedonic, eudaimonic or overlapping for participants with high scores on the SWLS and MLQ to determine to what degree these qualitative categorisations of lay Ghanaian people’s ideas reflect their levels of hedonic or eudaimonic well-being as supposedly being captured by the SWLS and MLQ. In other words, to determine if qualitative categorisations of well-being would reflect high scores on either forms of well-being, indicating the extent to which this categorisation is applicable in the Ghanaian context. Our findings indicate that the two well-being scales have satisfactory psychometric properties in this sample. The thematic analyses indicate that multidimensional well-being, psychological and emotional stability, social and relational well-being, meaning and purpose, as well as the absence of ill-being are important conceptualisations of well-being in the present sample. The results of the study also indicate that the current sample of Ghanaians’ lay conceptualisation of well-being based on their SWLS and MLQ scores does not distinguish clearly between hedonic and eudaimonic dimensions. These are discussed in more depth in the following sections.

In line with previous findings, the SWLS was found to have good psychometric properties. We found support for the unidimensional factor structure of the scale. Research in Western contexts have pointed to the factorial validity of this tool (Diener et al. 2013; Glaesmer et al. 2011; Zanon et al. 2014) as well as in South Africa (Khumalo et al., 2010). In addition to the evidence from South Africa and Togo, our study shows that the SWLS is a valid tool for measuring how satisfied individuals are with their lives in the Ghanaian context. Apart from the construct validity, we did find convergent validity with other well-being scales.

We further tested the two-dimensional factor structure of the meaning in life questionnaire and found that the two-factor structure was the best fitting, however, it required some modification to the model. In order to improve the model fit we removed one item (the only negatively formulated item to be reversed scored “my life has no clear purpose”) as it seemed to be inadequate in tapping into the presence of meaning in the Ghanaian context. Contrary to our findings, previous research in South Africa indicated that the MLQ was reliable and the two-factor structure was supported (Temane et al. 2014). The reason for the poor factor loading of item 9 requires further exploration in Ghanaian context. It is possible that items tapping into the presence of meaning should only address such and not include to be reversed scored items on absence of meaning. Schutte et al. (2016) also argued that the poor performance of item 9 could be due to the disadvantages inherent in negated items such as cross-cultural differences in response including acquiescence. Schutte et al. (2016) suggested the removal of this item.

In our qualitative analysis we found that well-being in the Ghanaian context went beyond “feeling good” and “optimal functioning” to include domains of life such as physical health, material well-being, absence of ill-being and social and relational well-being. The emphasis on good health as a conceptualisation of well-being points to the current physical needs of the context of study. Glozah (2015) found that adolescents in Ghana described well-being in relation to the absence of diseases such as malaria. The current challenge with physical health care affordability and accessibility might explain the primacy given to this dimension of well-being.

With regard to material and economic well-being, there is extant research to show that there is a relationship between income levels and experience of well-being (Diener et al. 2015). In addition, research in deprived contexts have noted the importance of material resources for well-being (Wilson et al., 2018a, b; Wilson and Somhlaba 2016). It is interesting to note that economic well-being was linked to the absence of ill-being, sustenance, and health. This finding illustrates that financial viability was not only linked to the “good life” but the experience of balance in life. The slow socio-economic progress of the country has left most individuals with insufficient material resources to meet their daily needs. In addition, the absence of safety net programs has further exacerbated financial constraints (Pokimica et al., 2012). The need for material resources was accompanied by need to be free from stressors. This supposition of well-being slightly reflects Keyes’ proposition of well-being being the absence of ill-health (Keyes 2005).

A surprising finding was the seemingly minor emphasis on social relationships. When participants related the importance of relationships, it was in order to be able to provide for the needs of the family. A likely explanation is that the sample comprised individuals who were above 30 and were more likely to be married and have children. As a result, relational well-being was mainly defined in terms of being able to meet family demands and expectations. Wilson et al. (2018a, b) found that among aged individuals in Africa, earning enough to take care of the family was a key aspect of relational well-being. Tsai and Dzorgbo (2012) also argued that family reciprocity was key for happiness in the Ghanaian context. In the work of White (2017), well-being was conceptualised as encompassing broader relational networks and the capacity to cater for members of the network. Although previous studies on well-being included positive relations with others, as seen in the work of Ryff (1989), it does not tap into unique dimensions of reciprocity and generativity that is emphasised in studies in the African context. The focus is not primarily on ‘the good life’ in the sense of what they can get from others in relationships, but on ‘living the good life’ based on values assuming the collective norms of importance of the other and behaving in a responsible manner according to social norms. Relationships are thus implied in all behaviours without mentioning them as such.

With regard to social and relational well-being there was an emphasis on harmony with both God and others. Delle Fave et al. (2016) argued that across different countries, harmony was identified as a key component of happiness. In our study, participants were concerned about maintaining relationships and providing support rather than just the experience of support. This speaks to the notion of social contribution as suggested by Keyes (2005). It is worth mentioning at this point that harmony with God and the experience of spiritual well-being were the only mention of religion in connection to well-being. We would have expected more participants to refer to religion and religious experiences in the conceptualisation of well-begin as is typical in the African context (Wilson et al., 2018a, b; Wilson and Somhlaba 2016) but this was not the case. This might be due to increasing urbanisation and the need for self-reliance that comes with it.

These findings resonate to a great extent with the three dimensional categorisation of well-being proposed by the ESRC Research Group of Well-being in Developing Countries (WeD) (White 2008). White (2018) indicated that in a rural Zambian context where the “understandings of wellbeing are fundamentally grounded in the material and relational”– being able to provide for one’s family and having something to share with others who ask, requires a contextual and cultural grounding in our interpretation of what it means to be well. A similar notion is reflected in the words of a participant in the present study as indicated above: “Well-being means a state of fulfilment in personal life and in the lives of others”.

The WeD research group (White 2008) contended that the understanding of well-being is socially and culturally constructed. They conceptualised well-being as found in developing countries in terms of three main dimensions with unique descriptions that differ somewhat from definitions of the terms in mainstream positive psychology. The three integrated and interdependent dimensions are the subjective, material, and relational. The subjective dimension refers to values, experiences and perceptions linked to amongst others the moral order, what is seen as good, how people feel about their lives, the desires they have, what they see as meaningfulness and their level of satisfaction with life. Social and cultural norms play a role in the conceptualisation of this dimension, which is different from the individualist notions of “subjective well-being” defined only in terms of satisfaction with life and positive affect.

The relational dimension includes both close and more distant relationships linked to care, mutual support and obligations, but also links to cultural, and societal identities in which politics and scope for personal and collective action. The material dimension refers to standards of living, finances, employment, services, practical welfare, quality of environment and livelihood, and physical health – this is often expressed as deeply linked with the ability to care for family and others. In addition to these three components, it is worthwhile to emphasise the psychological states that were included in the conceptualisations of well-being for our present sample.

In determining whether lay conceptualisations of well-being categorised as hedonic, eudaimonic or overlapping reflect scores on life satisfaction and meaning in life, we found a non-distinct pattern in the difference in the number cases from both the high MLQ-P (the presence subscale) group and the SWLS group that described well-being as either hedonic, eudaimonic or an overlap between these two. Earlier scholarly work has indicated that there is a lack of evidence for the absolute distinction between hedonia and eudaimonia (Disabato et al. 2016; Fredrickson et al. 2013), providing some support for our current findings of blurred expression in this specific cultural group). Moreover, a number of perspectives do not support a total separation between hedonic and eudaimonic well-being, but argue that they do also overlap, function complementarily or show differential correlates in many studies (Luo et al. 2017). Our findings seem to point to the fact that among urban employed individuals in Ghana, who think of well-being in terms of hedonia or eudaimonia do not necessarily also score high on measures distinctly measuring these dimensions. The clear distinction between these dimensions is less apparent among individuals doing well on quantitative measures reflecting assumedly hedonic and eudaimonic dimensions. This leans towards notions of balance and equipoise in being well as experienced in this particular context.

Conclusions and Implications for Future Research

Our findings indicate that second order categorisations of well-being phenomena need to be revisited, perhaps especially in an African context. The reasons therefore are in particular that, firstly, lay conceptualisations of well-being in the Ghanaian context indicate the understanding of well-being as multidimensional and a more or less holistic phenomenon (emotional, mental, psychological, spiritual, physical, financial, and relational/social). Secondly, lay conceptualisations of well-being are not to be neatly categorised as only hedonic or eudaimonic or strictly a combination of both, but also showed a unique overlapping category where the simultaneous fulfilments of needs of the individual and others are intertwined with hedonic and eudaimonic notions.

Future research must concern itself with further theorising of well-being based on lay conceptualisations in order to interrogate current patterns of thought that might be limited in scope for other populations. In addition, we suggest that that the enhancement of well-being outcomes be designed to tap into the emotional, mental, spiritual, and relational. Exclusive focus on either component might not achieve intended effects in the Ghanaian context. There is also the need the large-scale qualitative studies on well-being across different groups in Ghana, given that the current sample was limited to urban employed individuals. Policy-makers and practitioners need to consider a model of well-being that is sensitive to context, and recognises people as psychological, social, and moral beings that need to navigate through quality engagements and fulfilment of their needs.