Evidence of construct validity for work values using triangulation analysis Open Access

From a semantic perspective, personal and work values are highly related constructs (Sharabi, 2017). However, relative to personal values, the literature on work values remains stagnant in business research. For example, over the last five years, papers published in the Euromed Journal of Business have linked the construct of personal values with topics ranging from aspects of entrepreneurial culture (Falaras and Moschidis, 2023) and CEOs and other high-ranking executives’ traits (Alhmood et al., 2023) to consumers’ engagement with new technologies (Ajayi et al., 2022) and the impact of board diversity on dividend payout policy (Khan et al., 2022). Nascimento and Loureiro (2022) also used personal values as part of their conceptual framework to explain green purchasing behaviour. In contrast, work values were only used by Sharabi and Yanay-Ventura (2022) in an examination of the construct’s effect on the promotion to managerial positions.

The near isolation of work values research from streams of research on personal values has recurrently been cited as a reason for this inconsistency (see Busque-Carrier et al., 2022). Within the personal values stream of research, the refined theory of basic individual values (Schwartz, 2017) provides a well-supported definition and, more importantly, a dynamic structure of motivational types that allows for classification, which has long been validated across nearly 100 countries and well over 500 samples (Sagiv and Schwartz, 2021). In contrast, until recently, work values research has relied on typologies derived using inductive approaches, which are fundamentally data-driven rather than guided by an a priori theoretical stance (see Papavasileiou et al., 2017). Consequently, there is still a lack of consensus regarding the underlying structure of the various types of work values (Abessolo et al., 2017; Busque-Carrier et al., 2022). Thus, it is argued in this paper that a theory – rather than results – driven approach “should generate a less contorted, more coherent set of predictions that emanate from or build upon the underlying perspective” (Shaw, 2017, p. 821).

Within this context, the aim of this paper is two-fold: first, to illustrate an emerging typology that theoretically links work values with personal values, and second, to provide evidence of construct validity for this typology. To accomplish this aim, we initially develop a hypothesis that links work values with personal values following the theory of basic individual values as refined by Schwartz (2017) and then provide evidence of construct validity using triangulation analysis with data from Japan. Hailing from topography and based on the metaphor of a triangle, triangulation was first introduced in navigation and surveying as a precise method for locating the exact position of an object. Within the social sciences, Campbell and Fiske (1959) the term is introduced as a means of examining the convergent and discriminant validity of measures. The implicit assumption in triangulation analysis is that, by using a combination of perspectives, investigators, methods, and/or data sources, the researcher can provide a more accurate and valid interpretation of the concept under investigation. Triangulation is, therefore, considered an innovative method of analysis in business research, capable of developing a consolidated picture of the topic under investigation (del Vecchio et al., 2022; Nizamidou, 2023).

The remainder of this paper is organised as follows. Section 2 presents the theoretical framework of the study, which offers a working definition of work values and a description of the most accepted work values typologies that facilitate the development of a formal hypothesis. Section 3 describes the data used to test the hypothesis, with an emphasis on the sample and aspects of work employed in the assessment tool. Section 4 presents the triangulation analysis performed to test the proposed hypothesis. First, the results of the principal component analysis (PCA) are presented using data from the public sector. Second, the findings from the hierarchical cluster analysis (HCA) were revealed from data derived from parapublic ^[1] employees. Third, the outcome of the smallest space analysis (SSA) is offered using data from the private sector. Finally, the concluding section (Section 5) determines the theoretical, methodological, and practical implications, acknowledges the study’s limitations, and presents avenues for future research.

The theoretical basis of this endeavour lies in the Theory of Work Adjustment (TWA; Dawis and Lofquist, 1984). The logic underlying the TWA is that (1) all individuals have a set of preferences for specific aspects of work, which is reflected by the high importance they place on them; (2) each context of work enforces high importance differently; and (3) individuals develop job satisfaction and subsequently increase their tenure when the specific aspects of work on which they place high importance are fulfilled by the specific context of work. The greater, therefore, the agreement between an individual’s high importance of aspects of work and the reinforcement of those work aspects in any work context, the more satisfied the individual becomes, and as such, the more likely this individual is to stay in this work context. Thus, the relative importance that individuals place on aspects of work, can be understood as an important prism for examining the interface between individuals and different contexts of work across sectors and cultures (Papavasileiou and Lyons, 2015; Sharabi and Kay, 2023).

The literature is replete with references to the conceptualisation of work values as the relative importance that individuals place on aspects of work (Sharabi, 2017). From this perspective, it is difficult to distinguish work values from concepts such as job orientations, work goals, work norms, and/or work outcomes (Lyons et al., 2010). To address this confusion, Lyons and colleagues (see Lyons et al., 2010; Papavasileiou et al., 2017) proposed the term work aspects to refer to individual attributes of work (e.g. schedule, salary), working conditions (e.g. supervisory relations, job security), and work outcomes (e.g. intellectual stimulation, status), reserving the term work values as a higher-order construct that is understood by assessing the underlying patterns of the relative importance that individuals place on a variety of work aspects. We concur with the above and define work values as implicit organisers of the relative importance that individuals place on aspects of work. For example, an individual who places high importance on aspects of work related to personal fulfilment, accomplishment, and independence can be said to hold cognitive work values.

Figure 1 provides a schematic representation of the most accepted work values typologies over the past decade, using evidence from a variety of disciplines, including business, management decisions, human resource management, gerontology, industrial psychology, applied psychology, nursing, vocational behaviour, career assessment, and group dynamics. The first dichotomises work values into intrinsic and extrinsic types. Intrinsic work values are associated with the work itself and represent the relative importance placed on aspects of work such as variety, challenge, and intellectual stimulation. Extrinsic work values pertain to the consequences of work and reflect the importance placed on aspects of work such as job security and pay.

Over the past decade, the intrinsic/extrinsic typology has been used to develop the work values profiles of students in Canada (e.g. Chow et al., 2017) and the United States (e.g. Fukasawa et al., 2020), freelancers in Russia (e.g. Shevchuk et al., 2018), and worker co-operatives in South Korea (Park, 2023) (see Figure 1). Multi-sectoral studies have also utilised the typology across a variety of cultural contexts, including the Netherlands and Belgium (e.g. Schreurs et al., 2014), Finland (e.g. Sortheix et al., 2015), India (e.g. Fouad et al., 2016), the United Kingdom (e.g. Furnham et al., 2021), Peru (e.g. Fuchs et al., 2022) and Gesthuizen et al.'s. (2019) work across 37 countries (see Figure 1).

The second trichotomises the construct into instrumental, cognitive, and affective types. Instrumental aspects are concrete (e.g. pay) and have direct practical consequences (e.g. benefits and flexible schedule). In addition to the instrumental aspects that are more salient, there are aspects that reflect considerations, beliefs, and opinions, such as achievement and independence, which are considered cognitive. The third type includes aspects that stem from the expression of feelings, mainly in the context of interpersonal relations, such as a fair supervisor and esteem from coworkers, and can be acknowledged as affective.

Over the past 10 years, this typology has been employed to construct the work values profiles of students in Romania (e.g. Cardoș and Mone, 2016), Taiwan (e.g. Cheng and Lai, 2018), Turkey (e.g. Inanc and Ozdilek, 2019) and Spain (e.g. Cano-Escoriaza et al., 2019), employees in toy factories in China (e.g. To and Tam, 2014), clerical workers in Hong Kong (e.g. Yeung et al., 2015), and nurses in South Korea (e.g. Lee and Yi, 2023) (see Figure 1). Furthermore, Sharabi (2017) and Sharabi and Yanay-Ventura (2022) adopted this typology in multi-sectoral studies in Israel whereas Brinck et al. (2020) assessed the changes in these types of work values in Germany between 1989 and 2016 (see Figure 1).

The third typology, is in agreement with the existence of the instrumental and affective types posited in the trichotomisation approach, but claims that the existence of a fourth distinctive prestige or power type could better classify the relative importance that individuals place on aspects of work. This can be accomplished by dividing the cognitive region of work aspects into intrinsic and prestige regions. The intrinsic region reflects a sense of personal growth with respect to aspects of work, such as the use of one’s abilities, responsibility, and meaningfulness. Their attainment derives directly from the nature of their work experience. In addition, the aspects of work included in the prestige region are outcomes whose attainment includes a comparison of self with others, projecting personal superiority such as achievement, status, independence, and advancement.

During the past decade, the intrinsic, extrinsic (preferred to instrumental because of the use of intrinsic labelling), prestige, and affective typology has been implemented to comprehend the work values profiles of students in Greece (e.g. Papavasileiou and Lyons, 2015), Portugal (Cortés-Pascual et al., 2014), and Lithuania (e.g. Tūtlys et al., 2023) (see Figure 1). Furthermore, the typology has been favoured to develop the work values profiles of workers in the automobile sector in Iran (e.g. Jalalkamali et al., 2016), the public sector in China (e.g. Zhong et al., 2021) as well as nurses and counsellors in the United States (e.g. Burns et al., 2020; Hampton and Welsh, 2019) (see Figure 1). Multi-sectoral studies have also selected this typology to develop work values profiles among workers in South Africa (Jonck et al., 2017) and French-Canadians (e.g. Busque-Carrier et al., 2022) whereas Rani and Samuel (2016) conducted a comparison between public and private sector employees in India using the typology (see Figure 1).

The last typology is favoured to summarise most of the reinforcement and satisfaction individuals seek to acquire through working according to TWA for the following reasons: First, in Busque-Carrier et al.’s. (2022) review of the work values literature from 1970 to 2017, it is argued that the content of the intrinsic, extrinsic, prestige, and affective typology is aligned with the four types of work values that have been consistently identified in previous studies. Second, this typology allows for a link with mainstream research on personal values – the purpose of our study. In this emerging theoretical model, work values comprise four types conceptual parallel to four higher-order types in the theory of basic individual values (Schwartz, 2017).

• The intrinsic type, which refers to the pursuit of personal growth (e.g. accomplishment, fulfilment, and independence) is parallel to openness to change values.

• The affective type, which relates to feelings, emotions, and social experiences (e.g. relationships with coworkers, feeling of belonging, and life-time employment), is in parallel with self-transcendence values.

• The prestige type, which reflects aspects of dominance over others and personal success (e.g. authority, status, and directing others) and is parallel to self-enhancement values.

• The extrinsic type, which captures more concrete work outcomes (e.g. understanding boss, comfortable environment, and salary) and is in parallel with conservation values.

Within this context it is argued that work values are projections of the basic individual values at the context of work (see Figure 2). Thus, it is expected that the four work values types similar to those mentioned above would underlie the relative importance that individuals place on aspects of work.

Data from the Working Person Survey (WPS) were used to test the conceptual validity of this hypothesis. The WPS is a high-quality Japanese biannual cross-sectional workforce survey that was initiated in 2000 by the Recruit Works Institute (Suga, 2020). Considering that the conceptual framework underlying the measure has not been explicitly articulated by the Recruit Works Institute, 12 work aspects were selected and compared against other popular measures of work values that were employed in the previous validation studies mentioned above, including SWVI, WVQ, and LWVS. As shown in Table 1, it was judged that they covered a sufficient range of work aspects to facilitate the proposed intrinsic, extrinsic, prestige, and social typology. It should be noted that work values are an organising construct and, as such, the precise nature of the items in any work values item is not particularly important, as long as the set of items is representative of the work values domain (Papavasileiou et al., 2017).

Based on Question 6 of the survey, which lists 68 employment sector options, the following three sectoral subsamples were chosen: (1) public (from survey option 66, those working in public administration); (2) parapublic (from survey option 53, those working in health services such as hospitals and clinics); and (3) private (from survey option 29, those working in information technology). Table 2 describes the characteristics of the sample in terms of gender, marital status, education, and employment type.

Prior to conducting the triangulation analysis, the data were examined for missing values and univariate and multivariate normality. Three cases were identified in which the option “no answer” was selected for at least one of the 12 aspects of work, and these cases were excluded from further analysis. The skew and kurtosis indices were greater than the cut-off values of |3| or |8| recommended by Kline (2011), suggesting univariate normality of data. To accomplish multivariate normality, cases with significant values (p < 0.001) of squared Mahalanobis distance were deleted. Following Hair et al. (2014), the final sample size of 676 (i.e. N of 100–400) and the ratio of participants per measured item (above 55) were adequate to perform multivariate data analysis.

First, PCA was performed on data derived from public sector employees (N_public = 229). PCA considers total variance and derives principal components that contain small proportions of unique variance and, in some instances, error variance (Hair et al., 2014, p. 105). With this technique, units (values of 1.0) are inserted in the diagonal of the correlation matrix so that full variance is introduced into the resulting pattern matrix (Hair et al., 2014, p. 105). The Bartlett test of sphericity (748.330; p = 0.000) and the Kaiser–Meyer–Oklin measure of sampling adequacy (KMO = 0.785) were in line with Hair et al.’s (2014, p. 103) recommendations for sufficient correlations among items to proceed with this type of analysis. As such, oblique rotation was performed, and the eigenvalue criterion (>1.0) was used to derive the pattern matrix of public sector employees. The results revealed the relative importance of public sector employees placed on the 12 aspects, comprised of 4 principal components, explaining 64.3% of the total variance (Table 3). All items loaded greater than |0.60| on their designated principal component, with communality scores exceeding 0.50 (Hair et al., 2014, p. 115). In addition, internal consistency, as measured with Cronbach’s α coefficients, was above the lower acceptable limit (0.60–0.70; Hair et al., 2014, p. 90), ranging from 0.62 to 0.81.

The next step in the triangulation analysis comprised Hierarchical Cluster Analysis (HCA) with data collected from parapublic sector employees (N_parapublic = 244). The HCA calculates the proximity between each variable (i.e. work aspects) and then applies a clustering method to form sets of variables that express similar patterns or correlations over the group of people under investigation into a hierarchy or treelike structure (Hair et al., 2014). Ward’s minimum variance clustering method was employed, with the Squared Euclidean distance as the criterion for similarity. Studies comparing alternative clustering methods have indicated that Ward’s method exhibits the best overall performance in the recovery of true clusters (Dulebohn and Martocchio, 1998). Moreover, the advantage of using the squared Euclidean distance is that the clustering solution obtained can be directly related to those obtained from other multivariate techniques such as factor analysis (Pryor, 1983). Accordingly, the resulting dendrogram shown in Figure 3, when compared to the pattern matrix derived from public-sector employees, provides a structural similarity with four major and corroborating clusters.

From top to bottom, Cluster I represents the importance attached to social work values (i.e. coworkers, belonging, and lifetime employment), Cluster II reflects the preference for intrinsic work values (i.e. fulfilment, accomplishment, and independence), Cluster III describes the significance of extrinsic work values (i.e. salary, workplace conditions, and understanding boss), and Cluster IV indicates the importance attached to prestige work values (i.e. authority, job status, and directing others).

The third step of triangulation includes the Smallest Space Analysis (SSA), a non-metric multidimensional scaling procedure that plots variables (i.e. work value items) as points in an Euclidian space with inter-point distances corresponding to proximity measures based on similarity coefficients (such as Pearson correlations) among the variables (Lyons et al., 2010). In particular, two-dimensional SSA was conducted using the MINISSA module of the NewMDSX multidimensional scaling software. The analysis was performed on matrices of Pearson correlations derived from data from private sector employees (N_private = 203). The findings produced an acceptable fit (stress = 0.12; coefficient of alienation = 0.15) (Shye et al., 1994).

As shown in Figure 3, the two-dimensional space was divided into four clear and non-overlapping regions corresponding to the structure of the work values types observed previously. Work aspects related to personal fulfilment (FUL), accomplishment (ACC), and independence (IND) form one region that reflects an intrinsic type of work. An adjacent region includes the work aspects of workplace conditions (WKPL), understanding boss (BOSS), and salary (SAL), refer to an extrinsic type of work values. The other region adjacent to the intrinsic type contained work aspects related to prestige: authority (AUT), job status (STA), and directing others (DIR). The final opposing region contains three aspects related to affective work values: coworkers (COW), belonging (BEL), and lifetime employment (LTE). Each of the derived regions had Cronbach’s α coefficients above the lower acceptable limits (Figure 4).

Values are socially desirable, sense-making systems that people use to gain cooperation in pursuing their own goals by communicating about them with others (Sagiv and Schwartz, 2021). As such, values offer a powerful and significant effect in democratising and humanising the corporate setting by motivating actions in relation to the public fabric of a particular work context (Stokes et al., 2016). Based on the public fabric of the Japanese work context, this paper offered evidence of construct validity for a work values typology, providing the following theoretical, methodological, and practical implications.

This paper presents an authentic, a priori theory development and testing perspective for business scholars interested in work values, following a hypotheticodeductive inquiry. A theory, rather than a result-driven perspective, can generate a less contorted and more coherent typology of work values that emanates from an underlying motivational perspective. The “start with theory” approach should allow scholars interested in work values to offer more refined and comprehensive operationalisations to compare and integrate different aspects of work and types of work values. Moreover, the proposed typology has the advantage of corresponding to Schwartz’s prominent model of universal values and motivational types, opening avenues for further value-based investigations between the specific context of work and the more general context of personal life.

Furthermore, the typology lends support to those interested in extending the extant under contextualised boundary conditions of unpacking generational identity in the work settings (see Ng et al., 2022). It is well acknowledged that “generational cohorts focus on cataclysmic events that bring about a change in the value structure of society and bring a new set of values to those coming of age during those events” (Moore et al., 2022, p. 4). The set of values that result from successive entry into adulthood is distinctive of this generation group and continues to shape work-related attitudes and expectations of this cohort group in later years (Joshi et al., 2010). Thus, the relational process between generations in the workforce is based on different sets of work values.

Methodologically, this study adds to the emergent literature on business research that stresses the importance of triangulation analysis to enhance the reliability and validity of the findings (e.g. Nizamidou, 2023). In this sense, the study is innovative because the construct validity and reliability of the proposed work values typology have been accomplished using a multiple triangulation approach that combines both data and within-method triangulation. Until recently, most validation studies within business research have focused on a single analytical method, such as PCA (e.g. Fait and Sakka, 2020; Sefiani et al., 2018), exploratory factor analysis (e.g. Sefiani et al., 2018), confirmatory factor analysis (e.g. Nair and Shams, 2020) and their combination (e.g. Chatzoudes et al., 2022; Ehsani and Hosseini, 2023; Taamneh et al., 2022). The methods employed in our research covered all commonly applied techniques for exploring the structure underlying the data, and provided inductive, deductive, and spatial evidence to corroborate the observed structure. The methods and samples used in our study go beyond the norms reported in validation studies in business research. Drawing on this innovative approach, the results demonstrated acceptable internal consistency and a coherent structure that fits the theoretical model. These findings lend strong support to the use of the intrinsic, extrinsic, affective, and prestige typology for studying work values. This will hopefully encourage field scholars to adopt the proposed typology in future values-based explorations in the context of work.

Given that work values are a construct that defines the relational process between generations in the workforce (Joshi et al., 2010; Papavasileiou and Stergiou, 2023), this study also contributes to practitioners involved in diversity and inclusion. Recently, it has been suggested by business researchers that a “one size fits all” human resource management policy with the purpose of attracting, engaging and retaining the post 80s generation – the Millennials – is rather inadequate because national contexts, lifestyle choices and ageing (maturation) play a critical role in affecting what Millennials desire and prioritise from their employers and in their work (Ng et al., 2022). The intrinsic, extrinsic, affective, and prestige typology, through its simpler structure, provides practitioners with a considerable understanding of generational work values and an opportunity to not only assess and explain the preferences for specific aspects of work but also to identify the work values profiles of each generation and compare them. Therefore, important questions could be addressed, such as: Are Millennials for example a subgroup of the workforce with distinct combinations of work values profiles? What was the prevalence of these profiles? How do these profiles relate to the various outcomes of importance?

The typology is also critical for broader practical societal reasons, particularly those related to Japan, the cultural context of the paper. The Japan Business Federation, with a vision to advance to Society 5.0 – a transition to a human-centred society that leaves behind the deeply rooted and long-standing predominant marketized, capitalistic, and functionalistic models – has inter alia called for the development of human resources that can lead to diverse teams (Ashta and Stokes, 2023a). Work values are important elements in building diverse teams (see Klein et al., 2011) and the typology offers human resource managers in Japan a rigorous tool to assess the extent to which members of a team are dissimilar from one another in specific aspects of work as well as developing work values profiles for each team.

This paper provides an innovative paradigm for a multiple triangulation approach that combines both data and within-method triangulation. However, there is still potential to refine the methodology and expand the scope of research in future studies along three avenues. The first is the expansion of the aspects of work. It is well acknowledged that within the social sciences, “no single measure may be considered perfect or optimal” (Baruch, 2014, p. 16), and the 12 aspects of work included in the study, while reliable, do not exhaustively cover this domain. Future research could benefit from incorporating a wider range of work aspects. This approach allows for a more comprehensive understanding of work values, as the number of conceivable work aspects is almost limitless. The rationale is that an ever-increasing range of work aspects can be incorporated, taking into consideration the a priori classification across the intrinsic, extrinsic, prestige, and affective typology. This incorporation would preferably use an equal number of work aspects for each type, in line with Durvasula et al.'s. (2006) recommendations for multi-type constructs.

The second is the length of the survey. An important point for consideration in future studies is the balance between the comprehensiveness and length of the survey. As employees face an increasing number of surveys, their willingness to participate in lengthy surveys also decreases. Therefore, future studies should consider using a relatively short work values inventory to maintain high response rates. Studies employing a wider range of work aspects should avoid including more than six aspects per type, to ensure survey brevity and effectiveness.

The third is the conduct of qualitative instrumental case studies. Given the deductivist approach followed in this study, the inherent subjectivity of the interpretivist paradigm could provide an extension by signalling, deepening, and enriching the perception that different points of view are collected (Stokes, 2011). For example, (Warner et al., 2022; Schmitz, 2019) used this approach to seasonal and blue-collar employees, respectively, to provide insights into the importance of several work aspects corresponding to intrinsic, extrinsic, affective, and prestige work values. Considering that the context of the study was Japan, future studies could employ qualitative instrumental case studies to offer insights into the types of work values that influence self-initiated expatriates’ adjustment in an inter-Asian context, extending the work of Ashta and Stokes (2023b) on traditional spiritual values.

We would like to acknowledge the Center for Social Research and Data Archives located at the University of Tokyo, Institute of Social Science for providing the 2008 WPS database, and Dr Akiko Sakamoto, Dr Kei Tsui, and Yoshihide Yanagino for their help with the translation. We are also grateful to Professor Sean Lyons for his invaluable feedback on a previous version of the manuscript. Lastly, but not least we are thankful to the participants of the 78th Annual Meeting of the Academy of Management for their constructive comments on an earlier version of this manuscript.