Measuring eudemonic well-being at work: a validation study for the 24-item the University of Tokyo Occupational Mental Health (TOMH) well-being scale among Japanese workers.

Although the eudemonic perspective seems to be a promising in considering vocational identity among working population, well-being at work has been discussed primarily in terms of subjective/hedonic well-being. This study aimed to develop a new tool to measure eudemonic well-being at work (The University of Tokyo Occupational Mental Health [TOMH] well-being 24 scale)and investigate its validity in a collectivist culture. Two online surveys were conducted with a total of 1,760 workers in Japan. We created 89 potential items from existing scales. An exploratory factor analysis indicated eight factors for the dimensions of measurement. After item selection based on item response theory, the factor structure with three items from each of the eight dimensions indicated an excellent fit for another sample. Cronbach's α and intra-class coefficients ranged from 0.671 to 0.845. The scores of the tool were more strongly associated with subjective well-being in the work context rather than well-being in general. In addition, the participants in the group demonstrating a higher risk for mental illness and a more stressful work environment indicated significantly lower scores, even after adjusting for general eudemonic well-being. The new measurement may be useful both for academic and practical applications for measuring eudemonic well-being at work, independent from general eudemonic well-being.

Introduction

Well-being has been a central concept for promoting health and positive psychology in recent years. Its importance is recognized not only in psychological and health fields, but also in economics, sociology, management, and education1, 2). Well-being has critically different traits from the absence of negative factors3), having independent impacts on health and mortality4, 5). For working populations, well-being is also an indispensable indicator of positive association with human capital and productivity6).

There are two main conceptual types of well-being: hedonic and eudemonic7, 8). The former refers to emotional aspects of positive psychology. One of the famous theories in this type of well-being, theory of subjective well-being (SWB), suggested three dimensions of well-being: pleasant affect, unpleasant affect, and life satisfaction9). SWB is defined as a person’s cognitive and affective evaluations of his or her life as a whole, including emotional reactions to events and cognitive judgements of satisfaction and fulfillment10). High levels of subjective well-being are beneficial in lowering morbidity, decreasing symptoms and pain, and increasing longevity11). The latter type of well-being focuses on elements of a “good” or valuable life, such as purpose, growth, and meaning. The most well-known example of this type of well-being is psychological well-being (PWB), a theory developed by Ryff12) indicating six dimensions of eudemonic well-being: autonomy, environmental mastery, personal growth, positive relations with others, purpose in life, and self-acceptance. PWB has deeper philosophical roots and captures the essence of ancient Greek imperatives: to know yourself and to become what you are12). While these two types of well-being are correlated, their functions for health and productive outcomes differ2, 13). For instance, Straume and Vitterø13) reviewed functional differences between SWB and PWB, arguing that they are based on different groups of emotions: “pleasure” and “interest,” respectively. They also reported that the former group of emotions is strongly correlated with happiness and satisfaction while the latter is strongly correlated with engagement, involvement, and inspiration. These findings suggest that SWB and PWB play different roles in regulating behaviors and have different antecedents and outcomes.

Practice and research for well-being in various domains and contexts (e.g., school, home, workplace) have become more common14). For example, concepts of positive psychology for children and adolescents require a developmental perspective and unique aspects15). It is reasonable to consider unique structures and outcomes of well-being among children and adolescents since they may be more sensitive to adaptation in school life and academic evaluation, rather than to life in general. Among the working population, domain-specific concepts of well-being (i.e., well-being at work) are also applicable and useful as a major determinant of work-related outcomes such as productivity6). A recent review on positive psychology at work16) indicated various positive concepts at work such as commitment, engagement, resilience, psychological capital, positive teamwork, and leadership. Some of them could be re-conceptualized as unique factors of well-being at work.

Unfortunately, well-being at work has been discussed primarily in subjective/hedonic terms and eudemonic well-being has rarely been investigated. Previous studies have noted that well-being at work consists of emotional (e.g., positive affect at work) and cognitive (e.g., job satisfaction) components17, 18), which are both included in subjective well-being. Indeed, many psychosocial intervention studies in work-related settings have adopted positive affect or job satisfaction as notable outcomes of well-being19,20,21,22). The eudemonic perspective in working lives seems to be more important and useful than the subjective/hedonic perspective when considering vocational identity, career attainment, and relationships with co-workers. A previous study revealed that falling short of early career goals was associated with increased symptoms of depression and lower levels of life purpose23). Another study indicated that scores on life purpose, personal growth, self-acceptance, and environmental mastery were associated with career commitment and work-personality development24). Enhancing eudemonic well-being is also attractive for employers, in terms of less frequent and more fully solved conflicts, loyalty, civility, and increases in innovation at the workplace1). Therefore, creating a concept for eudemonic well-being at work and developing tools for measurement are both important objectives.

There are several previous studies that have integrated both subjective/hedonic and eudemonic well-being at work25,26,27) and attempted to develop measurements for eudemonic well-being at work1, 28, 29). The integrated models of well-being at work, for example, the one by Lent and Brown26, 27), have suggested social cognitive career theory to try unifying theoretical perspective on SWB and PWB. In this model, eudemonic processes serve as key routes by which people achieve and sustain hedonic well-being27). The measurements for eudemonic well-being at work, for example, a recent study conducted in the US29), suggests a new conceptualization for meaning-based job-related well-being beyond job satisfaction. Czerw1) conducted a large validation study in Poland to develop a scale for eudemonic well-being in the workplace context and proposed a scale consisting of 43 items within 4 dimensions: positive organization, positive relations with co-workers, contribution to the organization, and fit and development. However, conceptualization for eudemonic well-being at work is still not enough. Across differing cultural contexts, we should accumulate findings regarding structures, correlations, and similarities and differences in comparison to general eudemonic well-being in working populations. Previous studies suggested that evaluation and predictors for well-being were different between people in individualist and collectivist nations because collectivist cultures give priority to in-group, define the self in relational terms, and sacrifice positive emotions for the achievement of important goals14, 30). In addition, psychometric properties (i.e., reliability and validity), and interpretability for using the proposed measurements should be investigated since previous studies did not check for all aspects of these properties. Furthermore, considering practical use in workplace settings, scales with fewer items will be feasible.

This study aimed to develop a new measurement for eudemonic well-being at work (The University of Tokyo Occupational Mental Health [TOMH] well-being 24 scale) and investigate its structural validity, internal consistency, test-retest reliability, and convergent/known-groups validity. We conducted an exploratory examination of the factor structures of eudemonic well-being at work in Japan, which is predominated by a collectivist culture. We applied item response theory (IRT) to items and selected only three from each of the eight dimensions, with excellent performance for discrimination and difficulty. The developed scale is useful in both academic and practical terms owing to its confirmed psychometric properties and limited number of items. We hypothesized that the structures of the new measurement would overlap with those of general eudemonic well-being (i.e., six dimensions) and those of the previous study for well-being at work1), but also include specific new dimensions because Japanese workers would place more emphasis on interdependence and sacrifice of positive aspects. The new measurement has good internal consistency (Cronbach’s α ≥0.70) and test-retest reliability (intra-class correlation coefficient, ICC ≥0.70). Since the concepts of eudemonic well-being at work would overlap with general eudemonic and subjective/hedonic well-being, sufficient convergent validity (Pearson’s r ≥0.50) would be observed. In addition, higher correlations would be observed with job satisfaction than those with life satisfaction because the new measurement would focus more on the workplace context. Scores of the measurement would be significantly different by known-groups (i.e., levels of psychological distress, job stressors), even after adjusting for general eudemonic well-being. According to previous findings, workers classified as high risk for mental illness and a stressful environment at work—that is, people who experience higher levels of psychological distress, higher job strain, lower social support—would show significantly lower scores on this measurement.

Subjects and Methods

Study design

This was a validation study of a measurement tool comprising two online surveys. The first survey (Survey 1) was conducted in February 2018 and the second survey (Survey 2) was conducted in February 2019. Survey 1 was consisted of a baseline survey and a two-week follow-up to explore the initial factor structure, item characteristics, internal consistency, test-retest reliability, and convergent/known-groups validity. Survey 2 was conducted to collect additional participants to confirm cross-validity for the factor structure by confirmatory factor analysis (CFA). The study protocol was approved by the ethics committee of the Graduate School of Medicine and the Faculty of Medicine, The University of Tokyo (No. 11242-(4)). This article was written according to the standard of development of measurement, Consensus-based Standards for the Selection of Health Measurement Instruments (COSMIN)31). The properties of the measurement (i.e., reliability and validity) were reported according to the COSMIN Risk-of-Bias checklist31).

Participants

For Survey 1, a total of 1,030 workers who lived in all prefectures in Japan were recruited from registered respondents of an Internet survey company32). The survey company had access to a sample of over 2,000,000 participants in Japan and asked them to complete a self-reported questionnaire, based on the eligibility criteria. The criteria for participants were (a) all types of workers, including people engaged in part-time and freelance jobs, and (b) aged 18 or older. There were no exclusion criteria. Eligible participants who registered with the survey company were sent an e-mail with a link to the questionnaire. To obtain informed consent, participants clicked “agree” after reading the terms and conditions of the survey on the first page of the questionnaire and before answering the questionnaire. At the two-week follow-up, the survey company randomly selected 100 participants from the 1,030 workers who completed the baseline questionnaire and invited them to join again. The participants received about 100 points as a reward for each survey, which could be cashed out or used for shopping (one point was equal to one Japanese yen). A response rate could not be calculated because the survey company recruited participants until the number of completers reached the target (i.e., 1,030). Survey 2 was conducted following the same method as Survey 1, recruiting 730 workers.

Development process of an item pool for the new measurement

First, we collected potential items and created an item pool for the TOMH well-being 24 scale. We collected items from previous measurements for general eudemonic well-being based on Ryff’s psychological well-being theory12) and relevant concepts treated as positive outcomes at work16) (e.g., commitment to work and work engagement), modified for the workplace context. A systematic search was conducted to explore existing scales for measuring eudemonic well-being at work by nine researchers (KW, KI, YO, ASh, HE, YK, HAd, HAr, and ASa), and a total of eight scales were selected to be suitable for creating the item pool: CASP-1933), Psychological Well-Being Scale (PWBS)34), Utrecht Work Engagement Scale35), Organizational Commitment Scales36, 37), Job Descriptive Index38), a modified version of Bar-On’s Emotional Quotient Inventory (EQ-I)39), and Workplace PERMA-Profiler3, 40). Second, items were selected from the eight scales which related to components of eudemonic well-being at work, based on discussion among the researchers. Third, sentences of the items were modified into the context of working life. Finally, the item pool for the measurement consisting of 89 items was created (Appendix 1). All items were rated on a seven-point Likert scale (0=strongly disagree, 6=strongly agree).

Measurements

In addition to potential items for the TOMH well-being 24 scale, other variables were also measured in Survey 1 to investigate convergent/known-groups validity. All variables were measured by self-reporting.

General eudemonic well-being

General eudemonic well-being, according to Ryff’s PWB model, was measured with the Japanese version of the PWBS. This scale was developed and used in the Survey of Midlife in Japan (MIDJA), April–September 200841), with enough internal consistency (Cronbach’s α, 0.56 to 0.78). The scale consists of 6 factors, each of which includes 7 items: autonomy (e.g., ‘I am not afraid to voice my opinions, even when they are in opposition to the opinions of most people,’), environmental mastery (e.g., ‘In general, I feel I am in charge of the situation in which I live,’), personal growth (e.g., ‘I think it is important to have new experiences that challenge how you think about yourself and the world,’), positive relations with others (e.g., ‘Most people see me as loving and affectionate,’), purpose in life (e.g., ‘I have a sense of direction and purpose in life,’), and self-acceptance (e.g., ‘In general, I feel confident and positive about myself,’). All items are rated on a seven-point Likert scale (1=strongly disagree, 7=strongly agree). The sum of scores in each factor were calculated and used in analyses. Cronbach’s α of the scale in Survey 1 ranged from 0.647 to 0.843.

Positive and negative affect

As one indicator of subjective/hedonic well-being, positive and negative affect in general were measured using the Japanese version of the twenty-item Positive and Negative Affect Schedule (PANAS)42). PANAS is widely used in many languages to measure positive and negative affect. The two-factor structure, internal consistency, and convergent validity of the Japanese version of PANAS were also confirmed42). We used total scores of 10 items each for positive (e.g., excited) and negative (e.g., afraid) affects measured on a six-point Likert scale (1=not at all, 6=very much so). Cronbach’s α of the scale in Survey 1 ranged from 0.874 to 0.893.

Life and job satisfaction

Another concept of subjective well-being, satisfaction in life and job, were measured with questions from the Brief Job Stress Questionnaire (BJSQ)43). This scale has been widely used to assess stress responses in Japan. Job and life satisfaction measures consisted of one item each: ‘I am satisfied with my job’ and ‘I am satisfied with my family life,’ respectively. The two items are rated on a four-point Likert scale (1=dissatisfied, 4=satisfied), with higher scores indicating higher satisfaction.

Psychological distress

As an indicator of known-groups validity, psychological distress was measured with the Japanese version of the K6 scale44). The scale consists of six items (e.g., ‘About how often did you feel nervous?’), asking respondents how often they had experienced symptoms of psychological distress during the last 30 days. All items were rated on a five-point Likert scale (0=none of the time, 4=all the time). The reliability and validity of the K6 were confirmed in a previous study44). In this study, we divided the participants into three groups based on previous cut-off points: light distress (<5), subthreshold-level distress (≥5)45), and severe mental illness (≥13)46). Cronbach’s α of the scale in Survey 1 was 0.916.

Job stressors

As other indicators of known-groups in the workplace context, three variables of job stressors (job demands, job control, and social support from supervisors and colleagues) were measured by items from the BJSQ43) for job demands (three items, e.g., ‘I have an extremely large amount of work to do,’), job control (three items, e.g., ‘I can work at my own pace,’), and social support from supervisors and colleagues (six items, e.g., ‘How reliable are your superiors when you are troubled?’). All items are rated on a four-point Likert scale (for job demands and job control: 1=not at all, 4=very much so; for social support: 1=not at all, 4=extremely). Higher scores mean higher job demands, job control, and social support. We calculated medians for each variable and divided participants into four groups based on the level of job demands and job control (low strain, passive, active, and high strain)47) and into two groups based on the level of social support (low and high). We supposed that workers experiencing higher job strain and lower social support were more stressful groups. Cronbach’s α of the scale in Survey 1 ranged from 0.745 to 0.868.

Analysis

Of the completers, participants who seemed to answer the items irresponsibly were excluded from analyses based on two criteria: (a) responding to all potential items for the TOMH well-being 24 scale with 3 (neither disagree or agree) or (b) scores between non-reversed and reversed items were extremely different (more than 2 SD). Descriptive statistics (mean, SD, minimum and maximum values) for the 89 items in the item pool were calculated. For selection of the items, item-total (I-T) correlations (r) were calculated. If items had 0.40 or less correlations with total scores of the 89 items, the items were excluded afterward. Exploratory factor analysis (EFA) was conducted to investigate factor structure, using the maximum likelihood method for extraction and oblimin rotation (oblique=0). After exploring dimensionality, IRT analysis was conducted for each dimension using the generalized partial credit model (GPCM). We estimated discrimination (a) and difficulty (b, thresholds) for the items and selected three items per each dimension based on the excellence of these parameters. To confirm factor structure in the selected items, CFA was conducted for participants of Survey 2, using a maximum likelihood estimation. Internal consistency and test-retest reliability were tested by calculating Cronbach’s α and ICC. To examine for measurement errors, the standard error of measurement (SEM) and the smallest detectable change (SDC) were calculated. SEM describes the standard deviation of repeated measures in one participant, and SDC represents the minimal change that one participant must show on the measure to ensure that the observed change is real and not just measurement error48, 49). Convergent validity of the measurement was investigated by correlations (r) with general eudemonic well-being, positive and negative affect, and life and job satisfaction. An analysis of variance (ANOVA) was conducted to investigate mean differences among known-groups (i.e., psychological distress, job strain, and social support). Differences of estimated means, after adjusting for general eudemonic well-being, were also tested by analysis of covariance (ANCOVA). There were no missing values in the data since participants could not finish the survey until they completed all items in the questionnaire. SPSS version 25 (IBM software) for ANOVA and ANCOVA and Mplus version 8.250) for all other analyses were used.

Results

Characteristics of the participants

For both Survey 1 and Survey 2, recruiting continued until 1,030 and 730 participants completed the questionnaire. At the two-week follow-up of Survey 1, a total of 88 participants of the 100 workers randomly sampled completed the questionnaire again (response rate=88.0%). Of those who completed both surveys, 66 participants in Survey 1 and 53 in Survey 2 were excluded from analyses based on the criteria that they seemed to answer the items irresponsibly. As a result, a total of 964, 82, and 677 workers were analyzed. Table 1 shows characteristics of the participants, the majority of which were full-time and day-time shift workers, had university degrees, and engaged in clerical jobs.

Table 1.

Characteristics of the participants

		Survey 1(N=964)		Survey 1 follow-up(N=82)		Survey 2(N=677)
		N	%	N	%	N	%
Sex
	Male	477	49.5	45	54.9	336	49.6
	Female	487	50.5	37	45.1	341	50.4
Mean age (SD)		M=45.09 (SD=13.7)		M=44.39 (SD=13.0)		M=45.14 (SD=14.1)
Educational status
	Junior high school	15	1.6	1	1.2	16	2.4
	High school	269	27.9	22	26.8	172	25.4
	Vocational/technical college	223	23.1	14	17.0	181	26.7
	University	405	42.0	41	50.0	275	40.6
	Graduate school	49	5.1	4	4.9	31	4.6
	Other/unknown	3	0.3	0	0.0	2	0.3
Employment status
	Full-time	529	54.9	45	54.9	360	53.2
	Part-time	235	24.4	23	28.0	189	27.9
	Contract/dispatched	94	9.7	5	6.1	61	9.0
	Freelance	95	9.8	8	9.7	58	8.6
	Other	11	1.1	1	1.2	9	1.3
Employment shift
	Daytime shift	868	90.0	73	89.0	602	88.9
	Rotation/night shift	96	10.0	9	11.0	75	11.1
Occupations
	Managerial	101	10.5	16	19.5	61	9.0
	Professional/technical	173	17.9	13	15.9	110	16.2
	Clerical	218	22.6	14	17.1	162	23.9
	Sales	130	13.5	10	12.2	84	12.5
	Services	144	14.9	15	18.3	108	16.0
	Transport/construction	38	4.0	6	7.3	35	5.1
	Production/skilled	83	8.7	5	6.0	51	7.5
	Agriculture/forestry/fisheries	7	0.7	1	1.2	2	0.3
	Safety	5	0.5	1	1.2	10	1.5
	Other	65	6.7	1	1.2	54	8.0

Factor structure and item parameters

Table 2 shows factor structures, discrimination (a) and difficulty (b) for potential items on the TOMH well-being 24 scale. From the 89 items, 12 were initially excluded owing to low correlations with the total score in the I-T analysis. As a result of EFA, eight factors seemed to be appropriate as dimensions for measurement. The first ten eigenvalues for factors were 32.825, 2.846, 2.421, 2.181, 1.800, 1.373, 1.090, 1.064, 0.921, and 0.899 (A scree plot in EFA is shown in Appendix 2). According to these values and Guttman’s standard, we adopted eight dimensionalities. One of the items (Item No. 65) was excluded due to multiple loading to the factors. The second EFA for the remaining 76 items presented simple eight-factor structures. We named the eight factors, “F1: Role-oriented future prospects,” “F2: Autonomy,” “F3: Role-oriented positive perception,” “F4: Personal growth and development,” “F5: Negative schema,” “F6: Occupational self-esteem,” “F7: Relationship,” and “F8: Meaningful work.” Inter-factor correlations among the eight factors ranged from 0.230 to 0.566. In each dimension, IRT analysis revealed discrimination and difficulty parameters of the items. Some items had low discrimination and reversed transition points for difficulty even if they had strong loadings to the factors. Standard errors for estimating latent factors (θ) ranged from 0.206 to 0.424. A total 24 items (three items for each of the eight factors) were selected as items for the final version of the TOMH well-being 24 scale.

Table 2.

Factorial structures of the items in exploratory factor analysis and item response theory analysis

ItemNo.	Exploratory factor analysis (Survey 1, N=964)								IRT analysis (Survey 1, N=964)
	Factor 1	Factor 2	Factor 3	Factor 4	Factor 5	Factor 6	Factor 7	Factor 8	Discrimination(a)	Difficulty(b, threshold)						SE for theta(θ)	Finalversion
	Role-oriented future prospects	Autonomy	Role-oriented positive perception	Personal growth and development	Negative schema	Occupational self−esteem	Reationship	Meaningful work		b1	b2	b3	b4	b5	b6
14	0.550	0.075	0.103	0.019	−0.036	0.103	0.158	−0.038	0.949	−0.768	−1.426	−1.133	1.227	1.933	2.018	0.341
87	0.417	0.109	0.166	0.112	0.110	0.069	−0.010	0.157	1.652	−2.450	−2.364	−1.950	0.963	2.292	3.850		✔
48	0.395	0.138	0.201	0.150	0.057	0.123	−0.018	−0.033	1.427	−1.947	−2.449	−1.737	1.040	2.040	2.867		✔
88	0.366	0.126	0.120	0.004	0.042	0.253	0.025	0.087	1.242	−1.763	−2.063	−1.479	1.008	2.324	2.909
15	0.339	0.041	0.237	0.084	0.088	0.054	0.186	0.046	1.393	−1.923	−2.121	−1.836	0.955	2.066	2.625		✔
3	0.318	0.190	0.152	−0.109	0.037	0.129	0.159	−0.041	0.753	−1.367	−1.085	−1.218	0.972	1.524	1.677
45	0.306	0.273	0.102	0.165	0.044	0.073	−0.028	0.134	1.244	−1.666	−2.479	−1.857	0.754	1.745	2.563
16	0.059	0.764	−0.084	0.064	−0.067	0.111	0.026	−0.032	1.927	−3.028	−3.122	−2.082	0.785	2.222	3.365	0.424	✔
20	−0.017	0.695	0.019	−0.091	0.208	−0.130	0.079	0.012	0.930	−1.340	−1.780	−1.494	0.806	1.412	1.587		✔
22	−0.060	0.466	0.034	0.177	−0.061	0.059	0.011	0.017	0.721	−1.524	−1.877	−2.042	0.663	1.402	1.568
19	0.130	0.428	−0.040	0.139	−0.031	0.295	−0.112	0.030	1.019	−1.289	−2.242	−1.885	0.870	1.833	2.150		✔
86	0.249	0.256	0.196	0.175	0.089	0.078	−0.055	0.151	0.822	−0.676	−2.445	−1.474	0.596	1.490	1.716
61	0.019	0.019	0.658	−0.029	0.050	0.029	0.104	0.042	1.212	−2.066	−1.764	−1.970	0.813	1.768	2.367	0.206
66	0.057	−0.023	0.653	0.116	0.083	0.036	−0.007	0.110	1.991	−2.757	−2.727	−2.179	0.272	1.933	3.267		✔
7	0.075	0.076	0.574	0.131	0.077	0.159	0.022	−0.080	1.853	−2.704	−2.741	−2.095	0.537	2.551	3.080		✔
59	0.093	−0.025	0.519	0.200	0.041	0.103	0.082	0.007	1.623	−1.869	−2.755	−1.959	0.394	1.986	2.906
29	0.085	0.082	0.488	−0.086	−0.039	0.126	0.288	0.033	1.237	−1.718	−2.100	−1.689	0.593	1.962	2.528
5	0.081	0.088	0.456	−0.075	0.061	0.164	0.120	0.031	0.995	−1.850	−1.522	−1.622	0.672	1.431	2.176
60	0.235	−0.015	0.409	0.150	0.057	0.025	0.007	0.225	1.699	−2.466	−2.421	−2.069	0.507	1.861	3.137		✔
62	0.150	0.039	0.403	0.179	0.072	−0.024	0.063	0.133	1.317	−1.807	−2.033	−1.918	0.515	2.008	2.827
78	0.219	0.059	0.396	0.125	−0.097	0.097	−0.041	0.061	0.949	−1.287	−1.917	−1.300	0.740	1.929	2.121
64	0.158	0.032	0.393	0.127	0.020	0.007	0.057	0.176	1.233	−1.942	−2.166	−1.834	0.678	1.828	2.905
11	0.273	−0.026	0.372	0.121	0.089	0.175	−0.016	0.080	1.663	−2.166	−2.243	−1.700	0.907	2.547	3.387
79	0.344	0.064	0.351	0.182	0.055	0.097	0.029	−0.046	1.638	−2.374	−2.057	−1.874	0.795	2.416	3.187
84	0.146	0.163	0.320	0.223	0.092	−0.080	0.000	0.249	1.497	−2.316	−2.320	−2.275	0.337	1.772	2.609
13	0.230	0.105	0.309	−0.037	0.052	0.113	0.184	0.074	1.068	−1.489	−1.446	−1.587	0.662	1.622	2.201
63	−0.032	0.280	0.292	0.229	−0.052	0.002	0.142	0.055	0.937	−1.534	−2.108	−2.166	0.418	1.207	1.819
58	0.104	0.040	0.286	0.200	0.019	0.198	−0.051	0.237	1.267	−2.170	−1.994	−2.010	0.503	1.381	2.239
67	0.251	0.005	0.257	0.129	0.060	0.048	0.141	0.229	1.489	−2.625	−2.333	−1.743	0.517	1.745	2.786
34	0.169	0.097	0.079	0.597	0.054	−0.014	0.051	0.033	2.218	−3.558	−4.071	−2.857	0.235	1.979	3.563	0.37	✔
31	−0.117	0.110	0.095	0.586	−0.040	−0.015	0.016	0.076	0.936	−1.983	−2.392	−2.307	0.191	1.219	1.386
30	0.048	0.018	−0.005	0.506	0.334	−0.012	−0.006	−0.026	0.730	−1.320	−1.767	−1.506	0.594	0.983	1.034
80	0.117	0.019	0.181	0.404	−0.094	0.100	0.152	0.052	0.889	−0.902	−2.018	−1.369	0.402	1.222	1.403
74	−0.002	0.164	0.244	0.285	0.055	0.204	0.006	0.035	1.083	−1.989	−2.236	−1.764	0.447	1.646	1.988		✔
33	0.127	0.003	0.163	0.272	0.118	0.109	0.074	0.218	1.197	−1.899	−2.242	−2.195	0.191	1.442	2.094		✔
46	0.175	0.117	0.045	0.016	0.446	0.057	−0.022	0.150	0.918	−1.155	−1.899	−1.521	0.888	1.059	1.592	0.331	✔
77	−0.058	0.227	−0.051	0.085	0.438	0.170	0.065	0.083	1.097	−2.168	−2.101	−1.894	0.539	0.991	1.867		✔
28	0.096	0.183	0.207	−0.175	0.429	0.082	0.062	−0.075	0.813	−1.011	−1.833	−1.377	0.952	1.351	1.893
47	−0.021	−0.068	0.083	0.054	0.427	0.022	0.000	0.262	0.639	−1.353	−1.830	−1.457	0.902	1.072	1.041
35	0.229	0.050	0.145	0.189	0.415	−0.042	0.081	−0.050	0.837	−0.786	−1.675	−1.484	0.914	0.956	1.055
55	−0.101	0.043	0.137	0.100	0.410	0.095	0.163	0.152	1.103	−1.790	−2.311	−1.715	0.448	0.764	1.734		✔
36	0.002	0.071	−0.021	0.220	0.405	0.159	0.142	−0.158	0.720	−0.898	−1.818	−1.428	0.848	1.041	1.522
4	−0.051	0.168	−0.026	0.022	0.388	−0.020	0.241	0.225	0.907	−1.435	−1.936	−1.876	0.492	0.978	1.331
32	0.154	−0.016	−0.036	0.136	0.378	0.145	0.075	0.192	0.839	−0.967	−1.999	−1.245	0.586	1.216	1.029
24	−0.154	0.054	0.198	0.071	0.378	0.114	0.110	−0.153	0.541	−0.646	−1.655	−1.343	0.629	0.825	1.073
2	−0.012	0.255	0.099	−0.066	0.366	0.123	0.042	−0.067	0.632	−1.311	−1.357	−1.317	0.938	0.871	1.526
56	0.085	0.026	−0.038	−0.076	0.363	0.319	0.081	−0.032	0.628	−0.757	−1.521	−1.440	1.343	1.345	1.521
9	0.149	0.037	0.194	−0.162	0.299	−0.047	0.060	0.068	0.357	−0.169	−1.022	−1.003	1.075	0.552	0.652
52	0.056	0.117	0.080	0.004	0.069	0.654	−0.006	0.027	2.010	−3.415	−2.623	−2.061	0.933	2.660	4.661	0.295	✔
70	0.079	0.079	0.106	−0.002	0.125	0.601	−0.019	0.104	2.078	−3.041	−2.655	−2.198	0.933	2.855	3.848		✔
76	0.016	0.301	−0.016	0.051	0.009	0.431	−0.005	0.157	1.168	−2.223	−1.797	−1.629	0.714	1.791	2.185		✔
54	0.095	0.061	0.128	−0.007	0.104	0.427	0.106	0.027	1.041	−1.404	−1.712	−1.629	0.744	1.673	2.466
57	0.232	0.088	0.073	−0.075	0.013	0.398	0.117	0.131	1.196	−1.493	−2.251	−1.728	1.302	2.349	2.842
73	0.070	0.013	0.201	0.143	0.006	0.398	0.130	0.008	1.127	−1.697	−2.151	−1.833	0.878	1.807	3.009
72	−0.043	0.102	0.187	0.006	−0.029	0.382	0.237	−0.227	0.534	−0.680	−1.440	−1.167	0.502	1.245	1.320
71	−0.010	0.307	0.047	0.038	0.053	0.377	−0.076	0.213	1.183	−2.236	−2.237	−1.711	0.653	1.780	2.656
82	0.055	−0.100	−0.056	0.096	0.057	0.375	0.260	0.293	0.945	−1.339	−2.299	−1.884	0.782	1.789	2.355
75	−0.031	0.147	0.262	0.188	0.013	0.270	0.120	−0.035	0.984	−1.873	−1.838	−2.022	0.234	1.672	2.093
83	0.096	0.144	0.182	−0.109	−0.128	0.006	0.641	0.024	1.019	−1.391	−1.285	−1.607	0.504	1.593	2.319	0.344	✔
39	0.035	0.007	0.002	−0.001	0.257	−0.074	0.580	−0.006	0.766	−0.950	−1.219	−1.596	1.011	0.696	1.105
43	0.029	0.077	−0.005	0.066	−0.025	0.186	0.511	0.206	1.464	−3.166	−2.470	−2.300	0.541	2.180	3.081		✔
25	0.011	0.014	−0.132	0.113	0.293	0.100	0.501	0.048	0.974	−1.424	−2.002	−1.698	0.754	0.956	1.554
42	0.105	−0.005	0.007	0.144	0.209	−0.077	0.481	0.050	0.883	−1.545	−1.715	−1.527	0.832	0.783	1.449
38	−0.092	0.076	0.163	0.032	0.263	0.025	0.465	−0.056	0.905	−1.160	−2.047	−1.554	0.677	1.139	1.482
81	−0.122	−0.011	0.109	0.251	−0.009	−0.084	0.442	0.159	0.777	−1.586	−1.985	−2.079	0.215	1.019	1.403
40	0.098	0.032	0.036	0.299	−0.029	0.149	0.408	0.010	1.089	−1.626	−1.946	−1.952	0.404	1.394	2.334		✔
51	0.221	−0.030	0.195	0.142	−0.012	0.186	0.271	0.079	0.893	−0.994	−1.922	−1.846	0.653	1.612	1.841
68	−0.055	0.046	0.052	0.020	−0.004	0.179	0.267	0.529	1.554	−2.249	−2.735	−2.507	0.287	1.695	2.418	0.316	✔
69	0.159	0.028	0.107	0.053	0.095	0.072	0.023	0.529	1.540	−2.247	−2.461	−2.169	0.426	1.904	2.870		✔
85	0.110	0.115	0.220	0.136	0.052	−0.017	0.098	0.399	1.576	−2.533	−2.700	−2.056	0.557	2.022	2.754		✔
89	−0.271	0.138	0.192	0.043	0.049	0.246	0.075	0.392	1.055	−1.920	−2.335	−2.245	0.047	1.266	1.926
26	0.117	0.125	−0.049	0.048	0.047	0.332	−0.006	0.349	0.969	−1.753	−1.834	−1.629	0.928	1.698	2.373
23	−0.283	0.252	0.071	0.198	0.050	0.145	0.036	0.339	0.972	−2.327	−2.453	−2.291	0.011	1.389	1.579
37	0.076	−0.008	−0.143	0.248	−0.065	0.174	0.294	0.317	0.939	−1.132	−2.322	−2.288	0.709	1.639	2.129
10	0.071	0.119	0.285	0.275	0.043	−0.127	0.109	0.309	1.295	−1.659	−2.624	−2.174	0.180	1.438	1.984
49	0.268	0.240	0.014	0.011	0.154	0.063	−0.113	0.274	0.626	−0.497	−1.260	−1.747	0.775	1.189	0.922

IRT: item response theory; SE: standard error. Exploratory factor analysis was conducted in maximum liklihood method and obmilin rotation. IRT analysis was conducted in the generalized partial credit model.

The final version of the measurement

Table 3 shows the results of CFA and factor characteristics of the final version of the TOMH well-being 24 scale (The Japanese version of the measurement is shown in Appendix 3). The eight-factor structure with 24 items indicated an excellent fit for the data of Survey 2: comparative fit index (CFI)=0.926, Tucker-Lewis index (TLI)=0.909, standardized root mean square residual (SRMR)=0.046, and root mean square error of approximation (RMSEA)=0.044 (95% confidence interval, 0.039 to 0.049). Factor loadings in the CFA ranged from 0.576 to 0.831. Cronbach’s α coefficients and ICCs ranged from 0.671 to 0.845. SEM ranged from 0.486 to 0.661. SDC ranged from 1.348 to 1.831, indicating that an approximate 1.5-point change of scores implies meaningful change of the concepts.

Table 3.

Items, structures, means, internal consistency, and test-retest reliability of the final version of the scale (TOMH well-being 24)

Confirmatory factor analysis (Survey 2, N=677)					Factor characteristics (Survey 1, N=964 and N=82)
Factor	Name	Item No.	loading	Item in English (not validated)	Min	Max	Mean	SD	Cronbach α	Test-retest Reliability (ICC)†	SEM†	SDC†
F1	Role-oriented future prospects	87	0.761	In my working life, I feel I am making progress towards accomplishing my goals.	0	6	3.13	1.1	0.820	0.738	0.507	1.407
		48	0.671	In my working life, I enjoy making plans for the future and working to make them a reality.
		15	0.751	In my working life, I feel that the future looks good for me.
F2	Autonomy	16	0.716	In my working life, I am not afraid to voice my opinions, even when they are in opposition to the opinions of people.	0	6	3.23	1	0.761	0.749	0.486	1.348
		19	0.576	In my working life, I have confidence in my opinions, even if they are contrary to the general consensus.
		20	0.656	In my working life, It’s difficult for me to voice my own opinions on controversial matters. (R)
F3	Role-oriented positive perception	66	0.831	In my working life, I feel challengingness.	0	6	3.31	1.2	0.863	0.731	0.624	1.731
		7	0.750	In my working life, I can please myself what I do.
		60	0.780	I am proud of my working life.
F4	Personal growth and development	34	0.622	For me, working life has been a continuous process of learning, changing, and growth.	0	6	3.48	1.1	0.802	0.746	0.488	1.353
		33	0.645	In my working life, I have the sense that I have developed a lot over time.
		74	0.736	In my working life, I am generally motivated to continue, even when things get difficult.
F5	Negative schema	55	0.591	I feel disappointed about my achievements in my working life. (R)	0	6	3.43	1.1	0.740	0.671	0.648	1.795
		77	0.587	Before beginning something new in my working life, I usually feel that I will fail. (R)
		46	0.635	I don’t have a good sense of what it is I’m trying to accomplish in my working life. (R)
F6	Occupational self-esteem	70	0.804	I feel sure of myself in my working life.	0	6	3.16	1.1	0.845	0.626	0.661	1.831
		52	0.746	In my working life, I feel confident and positive about myself.
		76	0.612	In my working life, I believe in my ability to handle most upsetting problems.
F7	Relationship	43	0.679	In my working life, I know that I can trust others, and they know they can trust me.	0	6	3.30	1	0.765	0.695	0.604	1.674
		40	0.669	In my working life, I enjoy personal and mutual conversations with others.
		83	0.675	In my working life, I am satisfied with human relations.
F8	Meaningful work	68	0.738	In my working life, I am needed.	0	6	3.40	1.1	0.834	0.781	0.546	1.513
		69	0.753	In my working life, what I do is important.
		85	0.773	I feel that what I do in my working life is valuable and worthwhile.

†N=82. TOMH: The University of Tokyo Occupational Mental Health; ICC: intra-class correlation coefficient; SEM: standard error of measurement; SDC: smallest detectable change.

Convergent validity of the measurement

Table 4 shows a matrix of correlation coefficients between the eight factors of the TOMH well-being 24 scale, general eudemonic well-being measured by the PWBS, and subjective/hedonic well-being measured by the PANAS and BJSQ (life and job satisfaction). Correlations among the eight factors of the TOMH well-being 24 scale ranged from 0.490 to 0.777, indicating moderate-to-strong interrelations between one another. The eight factors of the new measurement also had moderate-to-strong and positive correlations with general eudemonic well-being and week-to-moderate correlations with subjective/hedonic well-being. When compared with correlations between general eudemonic well-being and job satisfaction (0.204 to 0.468), correlations between the TOMH well-being 24 scale and job satisfaction were relatively stronger (0.351 to 0.633).

Table 4.

Convergent validity (r) of the TOMH well-being 24 (Survey 1, N=964)

			TOMH well-being 24								General eudemonic well-being
TOMH well-being 24		Mean (SD)	F1	F2	F3	F4	F5	F6	F7	F8	Autonomy	Mastery	Growth	Relation	Purpose	Acceptance
	F1 Role-oriented future prospects	3.13 (1.1)	1.000
	F2 Autonomy	3.23 (1.0)	0.527	1.000
	F3 Role-oriented positive perception	3.31 (1.2)	0.777	0.506	1.000
	F4 Personal growth and development	3.48 (1.1)	0.708	0.542	0.747	1.000
	F5 Negative schema	3.43 (1.1)	0.600	0.558	0.599	0.616	1.000
	F6 Occupational self-esteem	3.16 (1.1)	0.650	0.660	0.641	0.643	0.617	1.000
	F7 Relationship	3.30 (1.0)	0.615	0.490	0.637	0.621	0.555	0.610	1.000
	F8 Meaningful work	3.40 (1.1)	0.660	0.527	0.699	0.710	0.629	0.653	0.666	1.000
General eudemonic well-being		Mean (SD)
	Autonomy	29.55 (6.0)	0.406	0.749	0.366	0.429	0.500	0.564	0.335	0.390	1.000
	Mastery	30.03 (5.7)	0.604	0.596	0.581	0.601	0.708	0.649	0.580	0.576	0.600	1.000
	Growth	31.17 (6.4)	0.657	0.510	0.617	0.727	0.694	0.584	0.569	0.626	0.498	0.742	1.000
	Relation	30.59 (6.3)	0.524	0.443	0.506	0.577	0.609	0.533	0.682	0.627	0.351	0.676	0.704	1.000
	Purpose	30.09 (5.0)	0.589	0.455	0.534	0.584	0.602	0.512	0.468	0.564	0.445	0.647	0.746	0.588	1.000
	Acceptance	29.12 (6.3)	0.649	0.562	0.585	0.594	0.679	0.695	0.568	0.603	0.526	0.770	0.739	0.662	0.641	1.000
Subjective/hedonic well-being		Mean (SD)
	Positive affect	33.19 (7.3)	0.647	0.519	0.606	0.582	0.521	0.600	0.519	0.567	0.422	0.554	0.581	0.492	0.529	0.605
	Negative affect	29.43 (8.4)	−0.379	−0.431	−0.403	−0.412	−0.545	−0.453	−0.458	−0.397	−0.443	−0.614	−0.492	−0.468	−0.373	−0.531
	Life satisfaction	2.78 (0.8)	0.314	0.203	0.309	0.274	0.267	0.239	0.293	0.250	0.168	0.350	0.325	0.337	0.280	0.386
	Job satisfaction	2.55 (0.8)	0.552	0.351	0.633	0.485	0.454	0.419	0.529	0.486	0.204	0.478	0.444	0.399	0.359	0.468

All correlations were statistically significant (p<0.05). TOMH: The University of Tokyo Occupational Mental Health; SD: Standard deviation.

Known-groups validity

Table 5 shows descriptive and estimated means by levels of psychological distress, job strain, and social support. The scores of the TOMH well-being 24 scale were lower in the groups of severe psychological distress, high job strain and passive for job strain, and low social support. The mean differences in most factors of the TOMH well-being 24 scale among the known-groups were still significant even after adjusting for scores of general eudemonic well-being. Workers with severe distress and low social support had lower scores, and those with active for job strain scored higher; however some of the differences in scores were not statistically significant: F1, F2, F4, or F7 for psychological distress; F5 for job strain or social support.

Table 5.

Known-groups validity of the TOMH well-being 24 (Survey 1, N=964)

		Descriptive statistics								Estimated means (Adjusted by Ryff’s 6 factors of PWB)
Psychological distress (K6)		Light(<5, N=348)		Subthreshold(≥5, N=460)		Severe(≥13, N=156)				Light		Subthreshold		Severe				p forANCOVA
		Mean	SD	Mean	SD	Mean	SD			Mean	SE	Mean	SE	Mean	SE
	F1 Role-oriented future prospects	3.53	1.1	3.07	0.9	2.41	1.2			3.10	0.05	3.18	0.04	3.06	0.07			0.174
	F2 Autonomy	3.63	1.0	3.18	0.9	2.51	1.2			3.25	0.04	3.26	0.03	3.13	0.06			0.127
	F3 Role-oriented positive perception	3.78	1.2	3.25	1.1	2.48	1.3			3.34	0.05	3.36	0.04	3.12	0.08			0.033
	F4 Personal growth and development	3.88	1.0	3.38	0.9	2.87	1.2			3.47	0.04	3.49	0.03	3.46	0.06			0.855
	F5 Negative schema	4.00	1.0	3.30	0.9	2.57	1.1			3.54	0.04	3.41	0.03	3.27	0.06			0.003
	F6 Occupational self-esteem	3.59	1.1	3.14	1.0	2.28	1.3			3.10	0.05	3.24	0.04	3.07	0.07			0.007
	F7 Reationship	3.76	0.9	3.18	0.9	2.62	1.2			3.38	0.04	3.27	0.03	3.20	0.07			0.097
	F8 Meaningful work	3.80	1.1	3.36	0.9	2.64	1.4			3.38	0.05	3.47	0.04	3.26	0.07			0.025
Job strain (BJSQ)		Low strain(N=251)		Passive(N=212)		Active(N=233)		High strain(N=268)		Low strain		Passive		Active		High strain		p forANCOVA
		Mean	SD	Mean	SD	Mean	SD	Mean	SD	Mean	SE	Mean	SE	Mean	SE	Mean	SE
	F1 Role-oriented future prospects	3.32	1.1	2.72	0.9	3.60	1.0	2.87	1.2	3.10	0.05	3.00	0.05	3.34	0.05	3.08	0.05	<0.001
	F2 Autonomy	3.53	0.9	2.87	0.9	3.67	1.0	2.87	1.1	3.27	0.04	3.14	0.05	3.42	0.04	3.12	0.04	<0.001
	F3 Role-oriented positive perception	3.52	1.2	2.81	1.0	3.93	1.1	2.99	1.3	3.29	0.06	3.09	0.06	3.68	0.06	3.20	0.06	<0.001
	F4 Personal growth and development	3.62	1.1	3.02	0.9	3.93	1.0	3.30	1.1	3.41	0.05	3.33	0.05	3.67	0.05	3.48	0.05	<0.001
	F5 Negative schema	3.73	1.0	3.09	0.9	3.80	1.1	3.11	1.1	3.45	0.05	3.41	0.05	3.51	0.05	3.37	0.04	0.162
	F6 Occupational self-esteem	3.35	1.1	2.81	1.0	3.65	1.1	2.84	1.2	3.06	0.05	3.10	0.05	3.38	0.05	3.12	0.05	<0.001
	F7 Reationship	3.56	0.9	2.97	0.9	3.65	1.1	3.01	1.1	3.35	0.05	3.23	0.05	3.45	0.05	3.18	0.05	<0.001
	F8 Meaningful work	3.56	1.0	2.88	1.0	3.87	1.1	3.27	1.2	3.34	0.05	3.18	0.06	3.62	0.05	3.45	0.05	<0.001
Social support by supervisor and colleagues (BJSQ)		Low support(N=419)		High support(N=545)						Low		High						p forANCOVA
		Mean	SD	Mean	SD					Mean	SE	Mean	SE
	F1 Role-oriented future prospects	2.80	1.1	3.39	1.1					3.02	0.04	3.21	0.03					<0.001
	F2 Autonomy	2.98	1.0	3.43	1.0					3.14	0.03	3.31	0.03					<0.001
	F3 Role-oriented positive perception	2.89	1.2	3.64	1.1					3.12	0.05	3.46	0.04					<0.001
	F4 Personal growth and development	3.16	1.0	3.72	1.0					3.40	0.04	3.54	0.03					0.006
	F5 Negative schema	3.13	1.0	3.67	1.1					3.39	0.04	3.47	0.03					0.099
	F6 Occupational self-esteem	2.85	1.1	3.40	1.1					3.09	0.04	3.22	0.03					0.012
	F7 Reationship	2.84	0.9	3.65	1.0					3.08	0.04	3.47	0.03					<0.001
	F8 Meaningful work	3.03	1.1	3.69	1.0					3.27	0.04	3.50	0.04					<0.001

TOMH: The University of Tokyo Occupational Mental Health; PWB: psychological well-being; BJSQ: the brief job stress questionnaire; SD: standard deviation; SE: standard error; ANCOVA: analysis of covarience.

Discussion

Our study suggested eight dimensions for eudemonic well-being at work among Japanese workers. Some of them were similar to previous findings and the others might be unique in the workplace context and/or collectivist cultures. Other indicators for validity and reliability of the final version of the measurement, which consists of the selected 24 items based on the IRT analysis, were enough-to-excellent. Scores were more strongly associated with subjective well-being in work contexts (i.e., job satisfaction). In addition, participants in the higher risk for mental illness and stressful environment at work indicated significantly lower scores, even after adjusting for general eudemonic well-being. The TOMH well-being 24 scale may be useful both for academic and practical use to measure eudemonic well-being at work, independent from the general concept of eudemonic well-being.

Among the suggested eight factors, autonomy (F2) and relationship (F7) were very similar to factors in general eudemonic well-being since Ryff’s PWB scale has the same dimensions. The previous scale also indicated relationships with co-workers as one of the main dimensions1). These two factors may also be important at work. The other three factors, role-oriented future prospects (F1), personal growth and development (F4), and occupational self-esteem (F6), were similar but slightly different with those in general eudemonic well-being and the previous scale. While items in F1 included overlapping concepts of achievement, purpose in life, and self-realization, the most important concept of this factor might be “prospects” of working life in the future. Career development and management across working lives are closely related to health and well-being23, 24). In addition, safe and positive status control during occupational life are well-known as rewards that evoke a person’s sense of mastery, efficacy, and esteem51). Thus, positive evaluation of their future prospects may be a central concept for eudemonic well-being among workers. Items in the factor of personal growth (F4), which is also a dimension of general eudemonic well-being, might contain expanding experiences and skills, as well as continuation of working even when the work is stressful and difficult. Items in F6 represented occupational self-esteem rather than self-acceptance and optimism. These factors might overlap but express different aspects of well-being than eudemonic well-being in general. The extracted factors were consistent with eudemonic well-being in terms of self-determination theory (SDT)52), especially in the work context53). According to this theory, human beings possess three basic, innate needs: relatedness, competence, and autonomy. In the work context, mastery and meaningfulness are additional basic psychological needs. Fulfillment of these basic needs by work can trigger intrinsic motivation and would affect work performance53).

Role-oriented positive perception (F3), negative schema (F5), and meaningful work (F8) might be unique dimensions in the workplace context. Items in F3, named as role-oriented positive perception, might represent meaningful and challenging work and might be indispensable for discussing positive aspects. This factor may be similar to existing concepts of work engagement54) and engagement in the Seligman’s PERMA model55). Our measurement suggested one more unique and important dimension, meaningful work (F8). When compared to daily personal life, working life may place more emphasis on how their work and work roles are meaningful, which reveal each worker’s strengths and values as distinguishable from other employees. Previous studies also supported this finding, indicating that role conflict and unclarity for roles at work are adverse job stressors56). It is worth investigating whether dimension of negative schema (F5) is unique only in collectivist cultures. This factor comprised only reversed items, which seemed to indicate an absence of negative cognition for working lives. Given the previous finding that East Asians tend to sacrifice positive emotions for achievement of important goals14, 30), the negative aspects might stand out when discussing their well-being. Absence of negative cognition about their working lives might therefore be important for eudemonic well-being at work among Japanese workers.

Results for reliability have supported our hypothesis, indicating good internal consistency and test-retest reliability. The measurement also indicated reasonable values for SEM and SDC and can detect meaningful change of eudemonic well-being at work, at around a 1.5-point change in scores.

The hypotheses for convergent/known-groups were also supported. Scores of the eight factors of the TOMH well-being 24 scale had moderate-to-strong and positive correlations with general eudemonic well-being and subjective/hedonic well-being. Relatively strong correlations were observed between autonomy (F2) and autonomy of the PWBS, and between relationship (F7) and relation in the PWBS. These results might be based on the similarity of these concepts. In addition, positive correlations between the factors and job satisfaction, which represents subjective well-being at work, were stronger than those between general eudemonic well-being and job satisfaction. The results might suggest that concepts measured by the TOMH well-being 24 scale are close to concepts in the workplace context. The most interesting result was that negative schema (F5) displayed a relatively strong correlation with general eudemonic well-being. This could have been owing to the basic traits of the target population; that is, East Asians tend to emphasize negative aspects when discussing well-being. The negative schema score (F5) significantly differed by the level of psychological distress but not by the level of job stressors (i.e., job strain and social support by supervisors and colleagues). Therefore, in collectivist cultures, the negative aspects of eudemonic well-being at work could be associated with general health status and not with domain-specific factors.

For mean scores of the new measurement by levels of psychological distress, differences between workers who had severe and light distress ranged from 1.01 to 1.43. Thus, a change in scores of over 1-point might be clinically meaningful. In addition, lower scores of the TOMH well-being 24 scale might be related to adverse psychosocial factors at work: high job strain and passive jobs, and low social support. Furthermore, results of the ANCOVA indicate that variances explained by the levels of psychological distress and job stressors were significantly different with those relating to general eudemonic well-being. The concepts measured by the TOMH well-being 24 scale might be distinguished from concepts measured by PWBS. High well-being scores in active jobs may be explained by existing findings that associations between job resources (including job control) and positive outcomes at work are exaggerated when levels of job demands are high57).

The study has several limitations. First, because the response rate could not be calculated and because an online survey was used, selection bias might exist. For instance, participants who were unhealthy and had low eudemonic well-being at work may have been reluctant to participate in the survey. Secondly, there may have been errors in measuring assessment of the standards of convergent validity. Third, other confounders not measured in this study, such as psychological capital (e.g., self-efficacy, optimism, and intrinsic motivation), might have distorted the results of the correlation analyses. Forth, some of the measurement properties, such as content validity and responsiveness, could not be discussed. Finally, the generalizability of the results to workers from other cultural backgrounds could be questioned owing to the sampling method. Compared to a recent national Labour Force Survey in Japan58), our sample included more workers engaged in managerial jobs (1.9% in the national survey) and more workers who had graduated from university (27.7% in the national survey). These workers could be related to higher levels of job stressors and higher scores of well-being.

Conclusion

The newly developed measurement, named the TOMH well-being 24 scale, indicated good reliability and validity. It may be a useful measurement tool for eudemonic well-being at work, as an independent concept from general eudemonic well-being. Regarding practical implications, this scale can be used for preventing psychological distress; assessing key indicators for improving work performance and productivity; and considering vocational identity, career commitment, work-personality development, and relationships with colleagues.

Funding

This work was supported by Health and Labour Sciences Research Grants (Research on Occupational Safety and Health) 2017 (H27-Rodo-Ippan-004) from the Ministry of Health, Labour, and Welfare and a Grant-in-Aid for Scientific Research from the Japan Society for the Promotion of Science (JP18H06435).

Conflict of Interest

The authors declare that they have no conflict of interest.