Psychometric properties of a workplace

method: Using a cross-sectional survey design, 789 public servants completed measures on workplace spirituality and perceived employee performance. A three-stage process was followed to investigate (1) the basic internal assessment qualities of the measure, (2) the added-value of the subscales (i.e., alignment with organisational values, meaningful work, sense of community) to the model, and (3) the assessment of the external validity of the measure in relation to an external variable (in this case perceived employee performance). Different indices are consulted during the three stages to determine (1) whether the measurement in question is essentially unidimensional in nature and (2) whether the estimated group factor scores are better predictors of the criterion than the estimated general factor scores. Main findings: The various indices suggested that the measure of workplace spirituality could be treated as multidimensional and essentially unidimensional in nature. The three primary factors (alignment with organisational values, meaningful work and sense of community) resulted in a significant (yet small) increase in accuracy of predicting the estimated scores associated with perceived employee performance compared with the prediction based on the estimated scores for the general factor (workplace spirituality). However, the presence of a strong general factor cannot be ignored, pointing to a measure that is also essentially unidimensional and to be scored accordingly – allowing for quick and accurate assessment of individuals’ levels of workplace spirituality. Practical/managerial implications: Applied researchers and practitioners should take note of the theoretical and statistical value associated with the subdimensions of workplace spirituality to better understand why these dimensions are predictors of employee performance. The results of the study emphasise the important role of both personal-level and organisational-level variables associated with workplace spirituality in relation to perceived employee performance. Contribution/value-add: The study suggests that this measure of workplace spirituality can be treated as essentially unidimensional and multidimensional in nature – depending on the purpose of the assessment (fine-grained versus general). In addition, the present study provides evidence of the usefulness in employing novel procedures to determine the dimensionality of an instrument using external validity evidence.


Introduction
Spirituality is regarded as a human desire for interconnectedness, transcendence, introspection and a sense of purpose and meaning greater than oneself (King & Boyatzis, 2015;Peterson & Seligman, 2004). This search for spiritual meaning extends from beyond the personal to the professional arena (Giacalone & Jurkiewicz, 2010), where it allows for a connection between 'the own self and the workplace' (Rathee & Rajain, 2020, p. 27). Workplace spirituality serves as a source of self-worth (Gotsis & Kortezi, 2008), which allows people to feel that they are serving a higher purpose, whilst finding meaning in their work (Göçen & Özğan, 2018). Eginly (2017, p. 83) stated that workplace spirituality enables employees to excite their 'life force and energies towards work-related matters', and when met with obstacles, it 'enables them to act with a fighting spirit to overcome the problems'. Therefore, workplace spirituality is viewed as a positive resource that assists employees in dealing with stress, enhancing creativity and solving business problems (Daniel, 2015;Karakas, 2010). It is also said to facilitate resilience and assist with the management of emotional problems (Gangadharan & Welbourne, 2016) whilst improving the physical, emotional, psychological, social and spiritual well-being of employees (Krahnke, Giacalone, & Jurkiewicz, 2003;Pawar, 2016). Gotsis and Kortezi (2008, p. 576) regarded workplace spirituality as the 'missing attribute of organisational life' and argued that our understanding of corporate reality will remain limited and incomplete in the absence thereof.
Despite these claims regarding the role of spirituality in the workplace, the complexities in defining workplace spirituality are widely recognised in the literature (Eginly, 2017;Karakas, 2010;Milliman, Czaplewski, & Ferguson, 2003;Petchsawang & McLean, 2017) as is the controversy around the measurement thereof (Case & Gosling, 2010;Dent, Higgins, & Wharff, 2005;Giacalone & Jurkiewicz, 2010;Karakas, 2010;Liu & Robertson, 2011;Miller & Ewest, 2013). Kinjerski and Skrypnek (2006) concluded that the field has lacked a widely accepted, psychometrically sound measure of spirituality at work, thus hindering research and further understanding of the construct. This is supported by others (Liu & Robertson, 2011;Miller & Ewest, 2013), who argued that the development of workplace spirituality scales is still in an exploratory phase and that it is only recently that existing scales have started to assess the specific variables relevant to workplace spirituality. Also, the construct validity of these measures either has not been tested or the scales show poor reliability and weak validity. Thus, the credibility of the results pertaining to workplace spirituality is mostly 'undecided' (Göçen & Özğan, 2018, p. 74).
However, to enhance the applicability of spirituality in the work context, organisational scholars encourage researchers to provide contextualised definitions of workplace spirituality and to conduct empirical studies (Giacalone & Jurkiewicz, 2010;Krahnke et al., 2003;Miller & Ewest, 2013;Milliman et al., 2003;Panday, 2017;Rathee & Rajain, 2020). These authors emphasise the importance of validating existing workplace spirituality measures on different work samples, as well as the importance of developing and testing rigorous, valid and generalisable workplace spirituality measures. They also suggest the establishment of basic relationships between workplace spirituality and workplace outcomes through empirical analysis (Rathee & Rajain, 2020).

Aim and objectives
Considering these arguments, the aim of the present study is to investigate the psychometric properties of a measure of spirituality at work developed by Milliman and his colleagues (2003) amongst a group of managers employed in local government. To achieve this aim, the following two objectives have been formulated: 1. To determine whether this measure of workplace spirituality is multidimensional or essentially unidimensional in nature. 2. To determine whether the estimated primary factor scores for the three sub-dimensions of this measure (alignment with organisational values, meaningful work, sense of community) are better predictors of the estimated scores for perceived employee performance compared with the general factor estimated scores (workplace spirituality).

Literature review Workplace spirituality: The construct
Research on spirituality was mostly executed in the fields of psychiatry and psychology (Palmer, 2018). During the 1990s, an awareness of the role of spirituality in the workplace became evident (Garcia-Zamor, 2003;Mohan & Uys, 2006). This was followed by the first empirical study in 1999 (Mitroff & Denton, 1999), which indicated that senior human resource executives and managers yearn for more spiritual openness within the workplace. Allowing the definitions of religion and spirituality to emerge from the respondents, it became clear that a distinction was made between religion -being viewed as a more institutional and dogmatic phenomenon and spirituality -viewed as a more individualistic phenomenon (see Göçen & Özğan, 2018;Soder, 2016). Respondents related spirituality to interconnectedness -'a feeling of being connected with one's complete self, others and the entire universe' (Mitroff & Denton, 1999, p. 83). It also included finding a life purpose (meaning) and the importance of the alignment between an individual's 'core values and those of the organisation they belong' to (Soder, 2016). Whilst spirituality is viewed as an appropriate and even essential topic for discussion amongst colleagues, the subject of religion is generally seen as unfitting in the same context (Mitroff & Denton, 1999).
Since 2000, many studies relating to workplace spirituality have been conducted within the field of organisational behaviour (Van der Walt & De Klerk, 2014). However, because of the richness and broadness of the construct (Vasconcelos, 2013), there is no consensus on how workplace spirituality should be described or defined (Ashforth & Pratt, 2003;Kolodinsky, Giacalone, & Jurkiewicz, 2008;McKee, Mills, & Driscoll, 2008). After reviewing 140 publications, Karakas (2010) identified 70 different definitions of spirituality at work and introduced three perspectives on its role in organisational performance. These include a human resources perspective (how spirituality enhances employee well-being and quality of life), a philosophical perspective (how spirituality provides a sense of purpose and meaningfulness) and an interpersonal perspective (how spirituality provides a sense of community and interconnectedness).
Various authors (Ashmos & Dunchon, 2000;Duchon & Plowman, 2005;Eginly, 2017;Milliman et al., 2003) have structured their definitions of workplace spirituality around three aspects: the inner lives of employees (related to an individual identity which is part of the person's self-concept and a social identity, occurring through group membership), the need to belong (sense of community and connectedness) and the need for meaning and purpose achieved through meaningful and purposeful work. For example, Ashmos and Dunchon (2000, p. 137) defined workplace spirituality as 'the recognition that employees have an inner life that nourishes and is nourished by meaningful work that takes place in the context of community'. There is this notion that workplace spirituality allows employees to perceive things more holistically and to feel more connected to 'a wider, deeper and richer whole' (Samul, 2020, p. 4) whilst assisting them to move beyond self-interest (Corner, 2009). Other definitions of workplace spirituality include transcendence (moving beyond self-actualisation), joy, energy, spiritual energy (Dehler & Welsh, 2003;Eginly, 2017;Petchsawang & McLean, 2017), personal values and the alignment thereof with the organisation's values (Neal, 1997;Neck & Milliman, 1994;Petchsawang & McLean, 2017). Petchsawang and McLean (2017, p. 218) described workplace spirituality as the 'individual's experience of energy, joy and awareness of alignment between one's values and one's meaningful work, a sense of connection to others, something larger than self and transcendence'.
However, from an overview of the workplace spirituality literature, little recognition has been given to previous work and theories relating to the conceptualisation of workplace spirituality. A long history of theory on meaningful work (Chalofsky, 2003) and related concepts, such as psychological safety and personal engagement (Kahn, 1990), job enrichment (Hackman & Oldman, 1980), experienced flow (Csikszentmihalyi, 1990), critical psychological states relating to personal and work outcomes (Renn & Vandenberg, 1995) and relational cohesion (Lawler & Yoon, 1996), dating back to as far as Immanuel Kant (1724Kant ( -1804 has played an important role in informing the various conceptualisations of workplace spirituality. By following and building on the premises of previous theories (e.g. Alderfer, 1972;Hackman & Oldham, 1980;Maslow, 1954;May, Angel, & Ellenberger, 1958), Kahn (1990) studied how people's psychological experiences and their work context influence their personal engagement and disengagement and described three psychological conditions of personal engagement and disengagement at work: meaningfulness, safety and availability. Kahn (1990, pp. 703-704) argued that meaningfulness is experienced when people feel useful, valuable and worthwhile and describe psychological meaningfulness 'as a feeling that one is receiving a return on investments of one's self in a currency of physical, cognitive or emotional energy'. During this time, Csikszentmihalyi (1990, p. 1) introduced the term 'flow'referring to the psychology of 'optimal experience' and stated that the creation of meaning involves 'bringing order to the contents of the mind by integrating one's actions into a unified flow experience' (p. 6). He argued that finding a purpose by itself is not sufficient. Harmony can only be found when 'an important goal is pursued with commitment and focus and all the varied activities fit together into a unified flow experience' (Csikszentmihalyi, 1990). Therefore, 'purpose, resolution and harmony unify life and give it meaning by transforming it into a seamless flow experience' (p. 6). Aktouf (1992), with his focus on humanism and radical humanism, pleaded for a more global view of humankind. Instead of viewing employees as a resource that should be monitored and potentially exploited, management practices should allow employees to use their intelligence in serving the organisation, assisting them in developing a sense of belonging (through collaboration, synergy and closeness), whilst providing them with more control over their work environments. Research on meaningfulness per se has a strong empirical and theoretical base in the psychology of work. The concept of a 'search for meaning' in non-work domains and its role in survival have been emphasised by Viktor Frankl (1985) in his observations and discussions on life in the German concentration camps. In an organisational context, human resource development and organisational development practitioners and scholars have for many decades concerned themselves with research and theories relating to meaningful work in an effort to create a deeper understanding of its development and operationalisation. Chalofsky (2003), Rosso, Dekas and Wrzesniewski (2010), and Mercurio (2016) conducted literature reviews, respectively, on the theoretical and historical assumptions underlying meaningful work and how it has developed over time. Chalofsky (2003) conducted several literature reviews in 1996, 1999 and 2003 in order to develop a construct for the meaning of work. Theories such as those of Maslow (1954), Herzberg, Mausner and Snyderman (1959), McClelland (1965), Alderfer (1972), Rogers (1961) and Ackoff (1981) explained how people move from basic to higher-order needs and how people find purpose when they experience the freedom to be who they are. May, Gilson and Harter (2004) extended the work of Kahn (1990) by exploring the influence of job enrichment, work-role fit and co-worker relations on psychological meaningfulness and the mediating effects of the three psychological conditions of meaningfulness on employee engagement. Steenkamp and Basson (2013, p. 1) introduced the concept of a 'meaningful workplace' in reaction to what they called 'the alienation and resultant loss of meaning by individuals in the workplace'. They concluded that a meaningful workplace is: (1) an environment (which is defined and characterised as meaningfulness at work) where people (2) perform meaningful work (which is described and defined as the meaning of work), and (3) where employees experience meaningfulness as all the requirements for psychological safety, psychological availability and community are complied with. (p. 8) Mercurio (2016) distinguished between a single source perspective and a multiple-source perspective on meaningful work. Literature that focuses on single sources of meaning has mainly dominated the work and theories underlying the present understanding of meaningful work and focuses mainly on self, cognition, spiritual life, values and motivation as sources of meaning. This research often includes a focus on Maslow's (1954) higher-order needs theory (working towards a higher purpose), motivation-hygiene theory (the degree to which motivational factors are built into the work itself) by Herzberg et al. (1959) and Hackman and Oldham's (1980) job characteristics model (where experienced meaningful work is a critical psychological state in order to become motivated at work). Others look at meaningful work from a more holistic, multidimensional perspectiveintegrating various sources of meaning (Pratt & Ashforth, 2003;Rosso et al., 2010), including a focus on externally orientated mechanisms of meaning, such as belongingness (identification and involvement with a specific group), sensemaking (interpersonally constructed meaning), purpose and transcendence (connecting to something beyond the material and greater to oneself) (Mercurio, 2016;Rosso et al., 2010).
As various disciplines have contributed to the body of knowledge relating to the meaning and meaningful work (e.g. spirituality, psychology, theology and sociology) and because of the complex nature of this concept, Mercurio (2016) concluded that this field of study lacks integration, consensus and a common understanding. However, the contributions made by these authors in the various efforts to define the construct of workplace spirituality cannot be underestimated. Also, with regard to the conceptualisation of workplace spirituality, there are concerns relating to the lack of an 'interdisciplinary, intersubjective discussion' about a spiritually oriented workplace (Gotsis & Kortezi, 2008, p. 581). The various approaches to workplace spirituality (i.e. exploratory, contextual and consequential) are mostly context specific. Therefore: [S]pirituality is conceived, and thus, theoretically constructed in a very specific way, always in accordance with the corresponding theoretical framework. Namely, spirituality is 'interpreted', and consequently, defined under the influence of the worldviews implied by specific religious, cultural and philosophical traditions, and in some instances, with the use of the analytical tools of a given scientific theory. (p. 581) Case and Gosling (2010) also argued that the meaning that workplace spirituality has for employees is potentially impoverished because scientific investigation of the concept necessitates the measurement thereof -subsequently leading to a narrow conception of spirituality.
On the individual level, there is the notion that when employees integrate their spiritual qualities, values and ideals with their jobs, it allows them to bring their complete selves to work and to deploy their emotions, intellect and creativity to work processes more fully (Eginly, 2017;Mitroff & Denton, 1999). These spiritual values and qualities also influence employees' interactions, interpretations, responses and outcomes at work (Kolodinsky et al., 2008, p. 466). Mitroff and Denton (1999) concluded that employees found their work to be more meaningful and satisfying when they were allowed to express their spirituality through work, subsequently enhancing their performance.
At the organisational level, workplace spirituality is embedded in and promoted by organisational culture (Palmer, 2018) as it embraces a sense of meaningful purpose and belonging that 'creates transcendence and unity within the diversity' (Göçen & Özğan, 2018, p. 75). Giacalone and Jurkiewicz (2003) defined workplace spirituality at this level as: [A] framework of organizational values evidenced in the culture that promotes employees' experience of transcendence through the work process, facilitating their sense of being connected to others in a way that provides feelings of completeness and joy. (p. 13) Such spiritual organisational cultures recognise both the mind and spirit of employees, as well as their need to find meaning and purpose in their work, whilst wishing to connect with other employees and be part of a community (Indradevi, 2020). Organisations with a strong sense of spirituality harness the 'whole person' and 'the spiritual energy that is at the core of everyone', which enables employees to apply stronger ethics and values in the workplace, whilst allowing for more flexibility and empowerment (Mitroff & Denton, 1999, p. 84). In such organisations, employees do not only experience trust and connectedness but they also work collectively towards a motivational organisational culture characterised by reciprocity and goodwill, resulting in higher levels of performance and profitability (Marques, 2006;Mitroff & Denton, 1999), whilst attracting and retaining the best employees (Eginly, 2017).
Despite these various arguments and definitions, it seems from the recent literature that 'conceptual convergence' has started to emerge, as definitions and the conceptualisation of workplace spirituality mainly include the recurring themes of sense of meaning, sense of purpose, sense of community and transcendence (Palmer, 2018, p. 28). Gotsis and Kortezi (2008) concluded that: [T]here are good reasons to believe that workplace spirituality is more than an impermanent trend; on the contrary, the concept carries a much more substantial meaning and its potential contribution to a more rounded understanding of human work, of the workplace and of the organisational reality in general is worthy of examination. (p. 575) For the purposes of this research, workplace spirituality is conceptualised as involving: [T]he effort to find one's ultimate purpose in life, to develop a strong connection to co-workers and other people associated with work, and to have consistency (or alignment) between one's core values of their organization. (Milliman et al., 2003, p. 427)

Measurement of workplace spirituality
Various authors express doubt whether workplace spirituality can be measured and subsequently question the credibility of results in this domain. They point out that it is: [A] difficult task to measure the transcendent and/or mystical side of people with a fully supported empirical consensus, since spirit, by definition, is beyond the body and the things we see around us. (Göçen & Özğan, 2018, p. 74) According to Case and Gosling (2010, p. 16), there is a concern that measuring spirituality in a quantitative manner and reducing it to a set of hypotheses and statistical relationships 'risks trivialising the subject and, indeed, offending the sensibilities of those whose beliefs and values are being scrutinised'.  & Kortezi, 2008;Milliman et al., 2003;Rego, Cunha, & Souto, 2007;Sheng & Chen, 2012). These measurement efforts attempt to create a common language and normative understanding of what work spirituality entails (Krahnke et al., 2003). The purpose is to make this area of research more relevant to the field of organisational studies so that it can be incorporated as an independent variable in conceptual models and its positive effects on work outcomes such as performance can be demonstrated (Karakas, 2010). Miller and Ewest (2013, p. 39, 40) classified workplace spirituality measures into three categories: development scales (e.g. level of development within the employee in reference to a range of 'mature versus immature behaviour'); adherence scales ('authentic adherence to religious, spiritual, or traditional beliefs') and manifestation scales. Manifestation scales disclose specific manifestations and phenomenological experiences 'without regard to specific traditions and expressions of a person's values and corresponding motivations' (Miller & Ewest, 2013, p. 39). This typically includes measures assessing specific dimensions of workplace spirituality. The literature review by Rathee and Rajain (2020) showed that these dimensions mostly include inner life and meaningful work (Albuquerque et al., 2014;Ashmos & Dunchon, 2000;Pawar, 2009;Rego et al., 2007); a sense of community (Albuquerque et al., 2014;Ashmos & Dunchon, 2000); as well as alignment with organisational values and contribution to society (Rego et al., 2007).
Some researchers, such as Rego and Cunha (2008), included five dimensions for the measurement of work spirituality (opportunities for inner life, enjoyment at work, sense of contribution to society, alignment with organisational values and the team's sense of community). Others include four dimensions, such as compassion, mindfulness, meaningful work and transcendence (e.g. Petchsawang, 2008;Petchsawang & Duchon, 2009;Petchsawang & McLean, 2017) or engaging work, sense of community, spiritual connection and mystical experience (Kinjerski & Skrypnek, 2006).

The measuring instrument
Based on the work of Mitroff and Denton (1999), Ashmos and Dunchon (2000) developed the Spirituality at Work Scale. Miller and Ewest (2013) classified this scale as a manifestation scale, as it includes seven dimensions at three levels: individual, work unit and organisational levels, whilst it measures both individual and organisational perception. The purpose of this instrument is to understand how spirituality can contribute to more productive organisations (Miller & Ewest, 2013, p. 39, 40). However, items representing the individual level (inner life, conditions for community and meaningful work) produced the cleanest factor structure whilst those representing work unit and organisational levels were not as promising as significant overlap, as well as convergent and discriminant validity issues were evident (Ashmos & Dunchon, 2000). Milliman et al. (2003) strived to find a more precise way of measuring the different dimensions of workplace spirituality to address limitations in previous workplace spirituality research. Based on the Spirituality at Work Scale (Ashmos & Dunchon, 2000), as well as other prior studies, Milliman et al. (2003) focused on three dimensions: purpose in one's work or 'meaningful work' (individual level); 'having a sense of community' (group level) and 'being in alignment with the organisation's values' and mission (organisational level). Six items from Ashmos and Dunchon's (2000) Spirituality at Work Scale were used to measure 'meaningful work' and eight items to measure 'alignment with organisational values'. Milliman et al. (2003) then developed seven new items to measure 'sense of community', as they felt the items developed by Ashmos and Dunchon (2000) focused on the enabling conditions or outcomes of community and not on the sense of community.
Meaningful work in this questionnaire relates to the enjoyment of work, being energised by work and work that provides personal meaning and purpose. Examples of items that represent this dimension of workplace spirituality are 'I experience joy in my work', 'My spirit is energised by my work', 'Work is connected to what I think is important in life', 'I understand what gives my work personal meaning' and 'I see a connection between work and social good' (Milliman et al., 2003, p. 437). The way in which meaningful work is defined in this subscale relates to definitions of 'meaningful work' found in organisational psychology literature where it refers to whether employees experience their work as significant and serving an important purpose (Ashforth & Pratt, 2003), and it holds a positive meaning for an individual (Rosso et al., 2010) whilst merging professional and personal selves (Chalofsky, 2003). However, although some research on meaningful work focuses more on the person who is performing the work (including the importance of work in a person's life, work values and what is regarded as important in the job), other researchers focus more on what work is being performed (e.g. job design, job crafting) (Michaelson, Pratt, Grant, & Dunn, 2014). Therefore, meaningfulness is sometimes described in terms of job characteristics that are regarded as highly valued (Grant, 2007). Chalofsky (2003) stated that three themes emerged in the literature regarding meaningful work: a sense of self (e.g. the contribution that work makes to the sense of self), the work itself (e.g. challenging work, being able to contribute to the organisation's effectiveness) and a sense of balance (e.g. the merging of professional and personal selves). According to Rosso et al. (2010), employees' experience of meaningfulness can originate from four different sources: the self, spiritual life, other persons and the work context.
In Milliman et al.'s (2003) workplace spirituality questionnaire, having a sense of community relates to a sense of connection to co-workers, being linked to a common purpose and supporting one another. Examples of items measuring this dimension include 'Working cooperatively with others are valued', 'I feel free to express opinions', 'I think employees are linked with a common purpose', 'I believe employees genuinely care about each other' and 'I believe people support each other' (Milliman et al., 2003, p. 437). The way Milliman et al. (2003) defined this dimension is in line with how sense of community has been defined in organisational research literature. Organisational research scholars distinguish between a psychological sense of community and a structural sense of community (Garrett, Spreitzer, & Bacevice, 2017). Whilst a psychological sense of community in the workplace refers more to employees' sense of membership and identification with the work-related group or network (small or large), a structural sense of community is based on the view that the workplace is a community that is formed for a purpose (Klein & D'Aunno, 1986). Therefore, the focus is beyond the value of the social connections and more on the instrumental functions that the workplace structure serves, for example, knowledge-sharing, collaboration and innovation. So, what makes a workplace group a community is not as much the 'quality of relationships amongst members' but rather a set of shared practices' to accomplish some intended purpose' (Garrett et al., 2017, p. 823).
Alignment with organisational values includes feeling connected to an organisation's goals, identifying with the organisation's values and mission and feeling that the organisation cares for its employees (Milliman et al., 2003). Examples of items include 'I feel positive about the values of the organisation', 'I feel connected to the mission of the organisation' and 'the organisation cares about all its employees' (Milliman et al., 2003, p. 437). Milliman et al. (2003) empirically assessed these dimensions at the individual level of analysis by involving 200 part-time MBA students attending a business school in the south-west USA. The measure was based on a seven-point scale from 1 ('disagree strongly') to 7 ('agree strongly'). The reliability for each multi-item scale was measured using Cronbach's alpha, whilst the validity was assessed using confirmatory factor analysis (CFA). However, only the factor loadings of the items used in the measuring instrument were reported. Coefficient alphas range from 0.82 to 0.94. Unfortunately, Milliman et al. (2003) did not report on the goodness-of-fit statistics associated with the three dimensions of workplace spirituality. They only reported the goodness-of-fit statistics for the overall model (including workplace spirituality and the five attitudinal variables. Milliman et al. (2003, p. 428) acknowledge the limitations of workplace spirituality measures in general, as well as the limitations of their workplace spirituality measure. They point out that 'workplace spiritualty is a complex and multifaceted construct' and 'using a survey instrument to measure something as deeply personal and abstract as workplace spirituality presents numerous challenges'. However, Gotsis and Kortezi (2008, p. 583) argued that the 'limited knowledge and understanding of the field' do not diminish the importance thereof and should not inhibit the further examination of workplace spirituality. The challenge is to ensure the selection of relevant 'proxies' of workplace spirituality (such as a sense of community, meaningful and purposeful work, connectedness) 'that come extremely close to revealing the phenomenon of spirituality' (Dent et al., 2005, p. 639). This serves to sufficiently substantiate a further inquiry into the validity of the Milliman et al. (2003) measure. Pradhan and Jena (2016) conceptualised employee performance as consisting of three aspects: task performance, adaptive performance and contextual performance. Task performance relates to the effectiveness with which employees execute their assigned tasks and job responsibilities, as captured in their job descriptions and facilitated through relevant technical knowledge, skills and cognitive abilities. Adaptive performance relates to the ability of employees to deal with dynamic, changing and uncertain work environments (e.g. technological and job changes, as well as the ability to adjust their interpersonal behaviour when needed) (Griffin, Neal, & Parker, 2007;Pradhan & Jena, 2016). Contextual performance is viewed as prosocial or extra-role behaviour that includes elements such as assisting others, creating teamwork, taking on extra work, perseverance and ethical conduct (Coleman & Borman, 2000;Motowidlo & Schmit, 1999).

Employee performance and workplace spirituality
Performance and effectiveness have also been investigated in relation to workplace spirituality (Daniel, 2010;Duchon & Ploughman, 2005;Novitasari et al., 2018;Rego et al., 2007;Sharma & Singh, 2020). The majority of studies relating to performance focus on organisational performance (Albuquerque et al., 2014;Krahnke et al., 2003;Novitasari et al., 2018) and work unit or team performance (Daniel, 2010;Duchon & Ploughman, 2005), whilst a limited number of studies focus on individual performance (Rego et al., 2007;Petchsawang & Duchon, 2012). Several of these studies conclude that if organisations encourage workplace spirituality, it leads to enhanced organisational performance (Albuquerque et al., 2014;Eginly, 2017;Giacalone & Jurkiewicz, 2010;Krahnke et al., 2003;Novitasari et al., 2018). Although studies on workplace spirituality and employee performance are sparse, there is some evidence that workplace spirituality enhances individual performance. Mitroff and Denton (1999) found that employees who could express their spirituality through work performed better and found their work more meaningful and satisfying. A study by Rego et al. (2007) showed how the perceptions of people regarding the five dimensions of workplace spirituality (a team's sense of community, alignment with organisational values, sense of contribution to society, enjoyment at work and opportunities for inner life) predict self-reported individual performance. Petchsawang and Duchon (2012) concluded that workplace spirituality does relate to employees' performance, as measured by their supervisors.
According to Jhajharia and Gautum (2015), spirituality enhances employee performance and encourages employees to be more honest, courageous and compassionate in their workplace. Moreover, the philosophical point of view of spirituality enables understanding of an individual's meaning and purpose, improving the ability of the employee to perform better and integrate innovation and creativity into their work. In contrast, the interpersonal viewpoint provides an employee with a sense of interconnectedness and community, which intensifies the attachment, loyalty and sense of belonging to the organisation and drives employee performance (Jhajharia & Gautum, 2015). Meaningfulness and professional companionship enhance employees' commitment to improving their performance levels (Pradhan & Jena, 2016). A review of the literature (Karakas, 2010) indicated that spirituality benefits employees and supports organisational performance because it provides employees with a sense of purpose and enhances their well-being. It also allows employees to experience meaning, a sense of community and interconnectedness in the workplace.
Despite the potential positive value of workplace spirituality, as debated amongst these various researchers, many others warn against the misuse thereof within organisations. The concerns relate mostly to how spirituality can be mismanaged for managerial control and material gain or using spiritual means to achieve economic goals ( Milliman et al. (2003) share the concern that organisations may attempt to manipulate and use the concept of spirituality at work as a tool to increase productivity. Schutte (2016) recommended that several questions regarding workplace spirituality need to be investigated, such as whether spirituality is about materialistic or non-materialistic concerns, whether the way of viewing workplace spirituality is not depriving it from the real meaning of spirituality and whether a theological or philosophical approach to the study of spirituality can be integrated with a social scientific approach to management and business. Other concerns relate to the role of leadership in workplace spirituality (Raaj & Gunaseelan, 2017;Tourish & Tourish, 2010). The results of the study by Raaj and Gunaseelan (2017), who examined the role of destructive leadership on workplace spirituality, show there is a significant positive and negative relationship between the variables of destructive leadership and workplace spirituality, whilst Tourish and Tourish (2010, p. 209) argued that not enough acknowledgement is given to the differences in power and priorities that exist between leaders and employees in the discussion of workplace spirituality. Therefore, workplace spirituality can rather advance the leadership agenda and be used as a vehicle to promote dictatorial and controlling leadership instead of the liberation of employees -leading workplace spirituality to: [B]ecome yet another attempt to establish monocultural workplace environments, in which employee dissent is demonized as the sinful antithesis of pure spiritual values, to which only morally deficient individuals could object, and which organizational leaders are uniquely qualified to articulate. (p. 209) The present authors acknowledge the views regarding the positive application and the possible excessive use or misuse of spirituality in the workplace. However, we take the viewpoint of eCunha, Rego, Simpson and Clegg (2020), who emphasise that for spirituality to be a virtuous strength, both its insufficient use (e.g. spiritual emptiness and http://www.sajip.co.za Open Access purposelessness) and the excessive use thereof (e.g. proselytism) need to be avoided.

Research design
A cross-sectional, quantitative survey research design was employed for the present study. The top three metropolitan municipalities in South Africa were identified based on results extracted from the Municipal IQ (a web application comprising of statistics and intelligence services that monitors and assesses South African municipalities) (Ilan & Heese, 2016). Municipalities are ranked in terms of performance and efficiency drawn from the Municipal Productivity Index (MPI) (Statistics South Africa, 2011), which includes financial and non-financial data. Factors that were included in this research were capacity requirements, reporting, basic planning and financial management. Once the three top-performing metropolitan municipalities were identified (i.e. the City of Johannesburg, City of Tshwane and City of Cape Town), they were approached and invited to participate in the study. The metropolitan municipalities provided the authors with a list of employees who formed the sampling frame from which simple random sampling was employed to select participants. These lists comprised employees employed in the service-orientated departments.
The list included all levels of employees in these departments and excluded general workers and technical staff (working on roads, potholes, streetlights, etc.). Simple random sampling was employed to select participants. Participation was voluntary.

Participants
Data were collected from managers working in the three selected metropolitan municipalities. The responses of 789 respondents were used in the data analysis. The majority of the sample was identified as female (51%). In terms of age, the majority of the sample was between 41 and 50 years (34%), followed by 27% aged between 31 and 40 years. With regard to tenure, the majority of the sample had between 6 and 10 years of experience (26%), followed closely by those with more than 20 years' experience (25%). Lastly, 58% of the respondents had at least a Bachelor's degree.

Measuring instruments
For the present study, two measuring instruments were used. Workplace spirituality was measured using the scale developed by Milliman et al. (2003). The instrument consists of three dimensions: meaningful work (six items), sense of community (seven items) and alignment of organisational values (eight items). The instrument has proven reliable (meaningful work: α = 0.88; sense of community: α = 0.91; alignment of values: α = 0.94). The present study found the following reliability estimates for the three dimensions: meaningful work: α = 0.911, sense of community: α = 0.905 and alignment of organisation values: α = 0.941. Although Milliman and his colleagues investigated the goodness-of-fit, they did not determine the goodness-of-fit for this measure of workplace spirituality separately from the measurement model used to evaluate their structural model.
Employee performance was measured using a scale developed by Pradhan and Jena (2016), which consists of three dimensions: task performance (6 items), adaptive performance (7 items) and contextual performance (10 items). Pradhan and Jena found the instrument to be reliable, with estimates ranging between 0.80 and 0.90. The present study found the following reliability estimates for the three dimensions: task performance: α = 0.839, adaptive performance: α = 0.735 and contextual performance: α = 0.907.

Data analysis
When investigating the dimensionality of measuring instruments, scholars are of the opinion that the goodness-offit associated with competing measurement models should be supplemented with and evaluated by using different approaches and indices (Ferrando & Lorenzo-Seva, 2018a). More specifically, emphasis should be placed on the quality of the score estimates that were derived from the solution (Ferrando & Lorenzo-Seva, 2018b). To this end, Garrido, González, Lorenzo-Seva and Ferrando (2019) suggested a three-stage process to investigate (1) the basic internal assessment qualities of a measure, (2) the added-value of the subscales (i.e. alignment with organisational values, meaningful work, sense of community) to the model and (3) assessment of the external validity of the measure in relation to an external variable (in this case perceived employee performance). The first two stages enabled the present study to investigate research objective one, while the third (and last stage) enabled the present study to investigate research objective two. It should also be observed that this last stage could also be viewed as a form of concurrent validation given that the data is cross-sectional.
Different indices are consulted during the three stages to determine (1) whether the measurement in question is essentially unidimensional in nature and (2) whether the estimated group factor scores are better predictors of the criterion than the estimated general factor scores. To implement this three-stage process, Garrido and colleagues (2019) suggest the use of two software programs: FACTOR (for the first two stages) (Ferrando & Lorenzo-Seva, 2017) and the unival package (Ferrando, Lorenzo-Seva, & Navarro-Gonzalez, 2019) available in the R 4.0.5 statistical package (R Core Team, 2021).

Ethical considerations
Before data collection commenced, ethical clearance was granted by the research ethics committee of the Faculty of Economic and Management Sciences at the University of the Free State (UFS-HSD2017/0670). Participants were informed of the purpose of the study, as well as their right to withdraw if they so wished. As part of the informed consent process, it was explained to participants that their responses would be anonymous.

Results
The following sections report the results associated with the workplace spirituality measure developed by Milliman and his colleagues (2003) following the three-stage process suggested by Garrido and colleagues (2019).

First-stage results
From Table 1, it is evident that the goodness-of-fit statistics (as obtained from FACTOR) associated with the original three-factor model has a better fit compared with the unidimensional model. The rotated loading matrix is reported in Table 2 whilst the inter-factor correlation matrix is reported in Table 3. It is evident from Table 2 that items 7 and 8 (associated with a sense of community) have cross loadings on organisational values. It should also be observed that applicability of the unidimensional model is supported by the results associated with the parallel analysis, suggesting a single factor explaining 53% of the variance. The factor loadings associated with the unidimensional model ranged between 0.628 and 0.822. The value associated with the explained common variance (ECV) should also be consulted when determining the dimensionality of a measure. Values ranging between 0.70 and 0.85 for ECV are indicative of a measure that can be treated as essentially unidimensional (Garrido et al., 2019). Although it is evident that this measure of workplace spirituality could be treated as essentially unidimensional in nature (ECV, marginal reliability, expected percentage of true differences [EPTD]), the competing threefactor model is still the better option (goodness-of-fit statistics).
Given the conflicting evidence, the present study continued to investigate the dimensionality of the measure of workplace spirituality by consulting the values associated with (1) marginal reliability and (2) EPTD. These two indices provide information about the consistency of person ordering along the factor continuum (Garrido et al., 2019).
It is evident, from Table 4, that both the marginal reliability and EPTD associated with the estimated general factor are (slightly, but not much) higher than those associated with the three estimated primary factors. In addition, both the marginal reliabilities and EPTD associated with the three primary factor score estimates are high. More specifically, a minimum value of at least 90% is required if researchers want to use factor score estimates for individual assessment . This implies that it is possible to consistently order and differentiate individuals based on the score estimates obtained from the measure of workplace spirituality (Garrido et al., 2019). If the marginal reliability and EPTD (associated with the estimated general factor score) were much higher than those associated with the estimated primary factor scores, there would be clear evidence of unidimensionality.
Based on the results associated with Stage 1, it is evident that there is no clear indication whether the measure of workplace spirituality should be treated as essentially unidimensional or multidimensional.

Second-stage results
Given the conflicting evidence, the present study continued to stage two of the three-stage process. When the primary estimates can more accurately predict their corresponding factors than the general estimates can, they have 'added value' (Garrido et al., 2019). The proportional reduction of mean square error of prediction (PRMSE) is an index that can be used by researchers to determine whether the primary     (Garrido et al., 2019). In short, if the three primary factors (alignment with organisational values, meaningful work, sense of community) have 'added value', then it would be advisable to treat the measure of workplace spirituality as multidimensional in nature.
From Table 5, it is evident that the values associated with the PRMSE for all the three primary factor score estimates are higher than PRMSE values associated with the general factor score estimates. This implies that all three of the primary factors (alignment with organisational values, meaningful work, sense of community) have added value. It should also be observed that the lower limits of the confidence intervals for each of the primary factor score estimates are higher than the corresponding value of the PRMSE associated with the general factor score estimates. This implies that if we use the general factor score estimates (i.e. workplace spirituality), we will not be that accurate in our prediction of individuals' true levels of workplace spirituality in the primary factors.
Based on the results associated with Stage 2, it is evident that the measure of workplace spirituality should be treated as multidimensional in nature.

Third-stage results
The third and final stage of the process requires researchers to determine how the different factors in the solution (primary factors versus general factor) relate to an external variable (in our case, perceived employee performance). More specifically, firstly, this stage allows researchers to determine whether the three primary score estimates are related to the external variable (perceived employee performance) in the same way as they relate to the general factor (workplace spirituality) (Garrido et al., 2019). Secondly, this stage allows researchers to determine whether the prediction of the external variable (perceived employee performance) based on estimated scores of the three subscales of the measure in question (alignment with organisational values, meaningful work, sense of community) outperforms the prediction that can be made from the general factor score estimates (workplace spirituality) Garrido et al., 2019).
From Table 6, it is evident that the zero value falls outside the confidence intervals in all three cases. This implies that the three primary factors do not relate to the external variable (perceived employee performance) in the same way as they relate to the general factor (workplace spirituality). The maximum difference score (0.198) and the associated confidence interval (0.0838-0.3053) were obtained by subtracting the extreme value (0.643) from the median value of the three primary factor scores (0.445). Given that 0 is not included in the confidence interval (hence statistically significant), it can be concluded that there is indeed evidence of differential validity. This implies that some of the given factors (e.g. meaningful work and sense of community) are more strongly related to perceived employee performance (external variable) than others (i.e. alignment with organisational values).
The incremental value estimate was 0.0976 (0.023; 0.1411) and significant. This implies that there is a significant (yet small) increase in the accuracy of the prediction of an individual's perceived performance factor score estimates based on the primary factor score estimates (i.e. sub-dimensions of workplace spirituality) compared with the prediction based on the general factor score estimates. In short, the prediction of estimated scores for perceived employee performance based on the estimated primary factor scores outperforms the prediction that can be made from the estimated general factor score (i.e. workplace spirituality).
Based on the results associated with Stage 3 (see Table 6 and  Table 7), it is evident that the primary factors of the measure of workplace spirituality (alignment with organisational values, meaningful work, sense of community) are better predictors of perceived employee performance than the general factor (workplace spirituality) -supporting the use of the three fine-grained scores. However, the high value associated with the ECV (0.87) cannot be ignored -suggesting that this measure of workplace spirituality can also be treated as essentially unidimensional (and scored accordingly). Given that the incremental value estimate was fairly small (0.0976), the loss of predictive power when general estimate scores are to be used is likely to be negligible.
Thus, the analyses provide support for treating this measure of workplace spirituality as essentially unidimensional and multidimensional in nature -depending on the degree to which researchers and practitioners are interested in finegrained or general scores of workplace spirituality. The results showed that the three primary factors (alignment with organisational values, meaningful work, sense of community) have added value over the general factor (i.e. workplace spirituality), allowing researchers to use the fine-grained factor scores when required. The correlated multidimensional model, therefore, provides a more refined  assessment of the workplace spirituality construct. However, when treating this measure of workplace spirituality as essentially unidimensional, the loss in predictive power is small and likely to be negligible.
In summary, researchers and practitioners need to be guided by the intended use of the factor scores derived from this measure. If the purpose is to better understand the underlying nature and/or working of workplace spiritually, researchers will do well to treat this measure of workplace spirituality as multidimensional. However, if researchers and practitioners are interested in obtaining an individual's general level of workplace spiritually (as well as studying the relationship between the general construct of workplace spirituality and other related constructs), it is theoretically and empirically justified to use single-factor scores.

Outline of the results
Regarding the dimensionality of the measure of workplace spirituality developed by Milliman and colleagues (2003), the present study confirmed the relevance and usefulness of the original three-dimensional conceptualisation for workplace spirituality based on the various indices reported during the three stages as suggested by Garrido and colleagues (2019). There seems to be ample evidence that when trying to measure workplace spirituality, researchers should incorporate meaningful work, a sense of community and an alignment with organisational values. It is, therefore, clear that when incorporating these three factors, researchers will be able to measure workplace spirituality at an organisational level (alignment of values) and a personal level (meaningful work and sense of community) (Milliman et al., 2003;Rathee & Rajain, 2020).
Very little information is available regarding the goodness-offit associated with measures of workplace spirituality. The greatest limitation is that a number of studies either (1) do not report the goodness-of-fit statistics associated with their measure (Ashmos & Dunchon, 2000) or (2) report the goodness-of-fit statistics associated with the measurement model used to test a structural model (Bashkar & Mishra, 2019;Milliman et al., 2003;Naseer et al., 2020;Otaye-Ebede et al., 2020). However, the results of the present study are comparable to those reported by Petchsawang and McLean (2017). These authors evaluated a four-dimensional measure of workplace spirituality (compassion, mindfulness, meaningful work and transcendence) consisting of 20 items. In addition to the consultation of goodness-of-fit statistics, the present study explored the usefulness of additional indices (ECV, marginal reliability, EPTD). Although the ECV suggested the presence of a general factor (workplace spirituality), both the marginal reliability and EPTD suggested that this measure of workplace spirituality could also be multidimensional in nature (as suggested by the goodness-of-fit associated with the three-factor model). It should be observed that none of the studies mentioned here employed the process and indices suggested by Garrido and colleagues (2019) to their measures of workplace spirituality.
The three primary factors (alignment with organisational values, meaningful work and sense of community) allowed for better predictions of perceived employee performance compared with the general factor (workplace spirituality). More specifically, the study found that these three primary factors are related to perceived employee performance in ways that cannot be accounted for by the general factor (workplace spirituality). It is also interesting to find that alignment with organisational values and meaningful work were strongly related to perceived employee performance. Given this discussion about the conceptualisation of workplace spirituality, especially as it relates to personal and organisational levels (Milliman et al., 2003;Rathee & Rajain, 2020), it appears that both personal-level and organisationallevel variables (Milliman et al., 2003;Rathee & Rajain, 2020) are drivers of employee performance. In contrast, Duchon and Plowman (2005) found that only a sense of community and meaningful work were significant predictors of organisational performance. One can, therefore, conclude that workplace spirituality is beneficial to both employees and the organisation.

Practical implications
Researchers and practitioners who are interested in obtaining an accurate, yet general, indication of workplace spirituality, can treat this measure of workplace spirituality developed by Milliman and colleagues (2003) as essentially unidimensional, using composite scores to correlate individuals' levels of workplace spirituality with other relevant constructs. However, treating workplace spirituality as a multidimensional construct does have some value. From a practical point of view, applied researchers and practitioners cannot ignore the theoretical and statistical value associated with the subdimensions of workplace spirituality to better understand why these dimensions are predictors of employee performance. The results of the study emphasise the important role of both personal-level and organisationallevel variables associated with workplace spirituality in relation to perceived employee performance. For example, the meaningfulness of work is not only embedded in work itself but also in how employees think about their work. Therefore, where needed, it is important to change employees' perceptions of the importance of their work. Organisations can cultivate a sense of meaningfulness through leaders who not only constantly remind employees of the importance of their work in relation to the higher purpose of the organisation but also encourage open discussions where employees can express their inner values and beliefs and make suggestions on how their work can be made more meaningful. To create a sense of community, an organisation needs to develop an organisational culture that acknowledges and welcomes diversity, whilst allowing for freedom of expression of different beliefs, values and feelings without feeling alienated. Although the virtual work environment and conditions pose challenges to building a sense of community, new ways should be found to overcome these barriers. As it is easier to foster a sense of community in smaller size workplaces or units (Ditzel, 2017), the focus could be on developing co-working spaces (virtual or physical) (Garrett et al., 2017), where smaller teams can either work together, sharing certain practices and achieving a shared purpose together or as individuals in the presence of others with the purpose of feeling part of a community. Finally, organisations would do well to show their employees that they care about them and clearly communicate the mission and vision of a particular organisation. This will allow employees to evaluate the extent to which their own values are aligned with those of the organisations they work for, thus ultimately they feel more connected to organisational goals (Milliman et al., 2003).

Limitations and recommendations
This study focused mainly on high-performing municipalities. Future researchers are encouraged to consider other contexts, such as poor-performing organisations or municipalities. Given the fact that workplace spirituality has been shown to influence employee and organisational performance, it would be necessary to understand the role of workplace spirituality (or the possible lack thereof) in settings where performance is poor or below standard. In addition, future researchers could investigate the influence of workplace spirituality on other important individual (e.g. engagement) and organisational (e.g. commitment) outcomes. For example, previous studies have found a significant relationship between workplace spirituality and commitment (Indartono & Wulandari, 2014) but not within an African context. In terms of measurement, future studies would do well to use hard criteria (concrete evidence) associated with employee performance within municipalities and organisations. The present study employed a self-report measure for perceived employee performance that may not be as an accurate indicator as that obtained from organisations and/or municipalities.