Abstract
Objectives:
Although well-being at work is important for occupational health, multi-dimensional workplace well-being measures do not exist for Japanese workers. The purpose of this study was to investigate the validity of the Japanese version of the Workplace PERMA-Profiler.
Methods:
Japanese workers completed online surveys at baseline (N = 310) and 1 month later (N = 100). The Workplace PERMA-Profiler was translated according to international guidelines. Job and life satisfaction, work engagement, psychological distress, work-related psychosocial factors, and work performance were measured as comparisons for convergent validity. Cronbach's alphas, Intra-class Correlation Coefficients (ICCs), and measurement errors were calculated for the reliability, and the validity of the measure was tested by correlational analyses and confirmatory factor analysis.
Results:
A total of 310 (baseline) and 86 (follow-up) workers responded and were included in the analyses. Cronbach's alphas and ICCs of the Japanese Workplace PERMA-Profiler ranged from 0.75 to 0.96. Confirmatory factor analysis indicated that the 5-factor model demonstrated a marginally acceptable fit (χ2 (80) = 351.30, CFI = 0.892, TLI = 0.858, RMSEA = 0.105, SRMR = 0.051). Overall well-being and the five PERMA domains had moderate-to-strong correlations with job satisfaction, psychological distress (inversely), and work-related factors.
Conclusions:
The Japanese version of the Workplace PERMA-Profiler demonstrated adequate reliability and validity. This measure could be useful to assess well-being at work, promote well-being research among Japanese workers, and address the problem of definition for well-being in further studies.
Keywords: Flourishing, Japanese workers, PERMA, Psychometrics, Well-being, Workplace
Introduction
The importance of well-being has been recognized not only in academic fields but also in public policy and economics1,2). Multiple well-designed cohort studies and meta-analyses have reported that well-being correlates with lower mortality risk2-4). Within occupational settings, a positive perspective, including a focus on well-being, has also been recognized as important for fostering human capital and productivity5,6).
The conceptualization and definition of well-being is a difficult problem and topic of active discussion among researchers, and currently focuses on a diverse array of dimensions or descriptions rather than definitions7). Perhaps the most well-defined trait of well-being is the separation of positive and negative dimensions. A systematic review indicates that effects of well-being are independent of negative affect8). Well-being can thus be critically distinguished from the absence of negative factors (e.g., negative affect, depression, anxiety, and distress). The other distinction that has been made is between hedonic (emotion, pursuing pleasure, avoiding pain) and eudaimonic (the good life) dimensions9). A cognitive evaluation of one's life (satisfaction with life) provides a third dimension. Most of the proposed well-being models utilize a combination of hedonic, eudaimonic, positive, negative, and evaluative dimensions2).
For instance, Diener's subjective well-being theory (SWB)10) suggested three dimensions of well-being: pleasant affect, unpleasant affect, and life satisfaction. Ryff's psychological well-being (PWB)11), focuses on six eudaimonic dimensions: autonomy, environmental mastery, personal growth, positive relations with others, purpose in life, and self-acceptance. Well-being at work has been discussed primarily in emotional (e.g., positive affect at work) and cognitive (e.g., job satisfaction) dimensions12). Alternatively, well-being has been conceptualized more holistically as flourishing, which combines multiple hedonic and eudaimonic dimensions. For example, Seligman's PERMA model13) consists of five domains: positive emotion (P), engagement (E), relationships (R), meaning (M), and accomplishment (A).
The problem of defining these concepts of well-being should be addressed through operationalization, using established measures7). Most of the well-being and flourishing models have corresponding measures. For instance, the PERMA-Profiler developed by Butler and Kern14) allows individuals to monitor their well-being. This tool can also be useful for integrating the dimensions of well-being, compared with other previously developed measures15,16). Kern developed a workplace version of the measure (the Workplace PERMA-Profiler)17), which adjusted the questions to the workplace context to measure well-being at work (Table 1). Mirroring the general version of the PERMA-Profiler, the workplace measure consists of five factors (positive emotion, engagement, relationships, meaning, and accomplishment) across 15 items, along with 8 additional items to measure happiness (1 item), negative emotion (3 items), health (3 items), and loneliness (1 item). The measure is freely available for individual use (www.permahsurvey.com).
Table 1.
The English Workplace PERMA-Profiler
| Label | Question | Response Anchors |
|---|---|---|
| The PERMA domains, negative emotion (N), and physical health (H) are computed as the average across the three items. Overall happiness is the average of the 15 PERMA items and the overall happiness (Hap) item. Loneliness (Lon) is a single item. Copyright Kern (2014), used by permission from the author. | ||
| A1 | How often do you feel you are making progress towards accomplishing your work-related goals? | 0 = never, 10 = always |
| E1 | At work, how often do you become absorbed in what you are doing? | |
| P1 | At work, how often do you feel joyful? | |
| N1 | At work, how often do you feel anxious? | |
| A2 | How often do you achieve the important work goals you have set for yourself? | |
| H1 | In general, how would you say your health is? | 0 = terrible, 10 = excellent |
| M1 | To what extent is your work purposeful and meaningful? | 0 =not at all, 10 = completely |
| R1 | To what extent do you receive help and support from coworkers when you need it? | |
| M2 | In general, to what extent do you feel that what you do at work is valuable and worthwhile? | |
| E2 | To what extent do you feel excited and interested in your work? | |
| Lon | How lonely do you feel at work? | |
| H2 | How satisfied are you with your current physical health? | |
| P2 | At work, how often do you feel positive? | 0 = never, 10 = always |
| N2 | At work, how often do you feel angry? | |
| A3 | How often are you able to handle your work-related responsibilities? | |
| N3 | At work, how often do you feel sad? | |
| E3 | At work, how often do you lose track of time while doing something you enjoy? | |
| H3 | Compared to others of your same age and sex, how is your health? | 0 = terrible, 10 = excellent |
| R2 | To what extent do you feel appreciated by your coworkers? | 0 =not at all, 10 = completely |
| M3 | To what extent do you generally feel that you have a sense of direction in your work? | |
| R3 | How satisfied are you with your professional relationships? | |
| P3 | At work, to what extent do you feel contented? | |
| Hap | Taking all things together, how happy would you say you are with your work? | 0 =not at all, 10 = completely |
However, no multi-dimensional measurements for well-being at work have been developed in Japan. In addition, the reliability and validity of the original Workplace PERMA-Profiler has not been confirmed in published papers. Although well-being at work can cover the same dimensions as overall well-being, work-related well-being might operate in different contexts and might be associated with different outcomes (e.g., productivity) than overall well-being5,6). In addition, specification and stratification of well-being will address further questions such as a spill-over effect of well-being between work and life12). In practice, because the Workplace PERMA-Profiler is easy to complete in a short time (23 items, or 15 items using only the PERMA domains), it could be useful as an indicator of positive aspects for prevention and health promotion approaches in the workplace.
The current study aimed to investigate the reliability and validity of the Japanese version of the Workplace PERMA-Profiler among Japanese workers. The internal consistency, test-retest reliability, structural validity, and convergent validity of a translated version of the measure were tested. We hypothesized that the Japanese version of the Workplace PERMA-Profiler would have good internal consistency, test-retest reliability, and five-factor structural validity. Based on correlations for the original PERMA-Profiler14), we also hypothesized that well-being measured by the Workplace PERMA-Profiler would have a moderate-to-strong positive correlation with job satisfaction (r≥ 0.50) and a moderate negative correlation with psychological distress (r≤ −0.30). We expected that well-being at work would overlap with job satisfaction and work-related factors, and would be negatively associated with adverse health outcomes. Because work engagement18) could be a similar concept with engagement (E) in the PERMA model for the workplace, we expected that this measure would have weak-to-moderate correlations with work-related psychosocial factors and work performance (r≥ 0.20)19).
Subjects and Methods
Design
This was a validation study consisting of baseline (November 2016) and one-month follow-up (December 2016) online surveys in Japan. The internal consistency, structural validity, and convergent validity of the Japanese version of the Workplace PERMA-Profiler were investigated using the cross-sectional data. Test-retest reliability was investigated using the longitudinal data one month after follow-up. Because Seligman suggests that the PERMA domains are more stable reflections of well-being13), we conducted the follow-up study after one month, expecting scores to remain fairly stable over that period. This manuscript was written according to the COnsensus-based Standards for the selection of health Measurement INstruments (COSMIN) reporting guidelines20). Each characteristic of the measure was reported according to the COSMIN checklist.
Participants
Participants were drawn from workers registered as respondents of an Internet survey company, Macromill, Inc21). Of the available respondents, 310 workers completed a web-based questionnaire in order of arrival. Macromill had access to over 2,000,000 potential participants representing all prefectures in Japan, and recruited participants based on their demographic attributes to obtain a relatively representative sample. These registered members have diverse characteristics in terms of gender and age. Participant inclusion criteria were (a) Japanese workers who lived in a prefecture of Japan and (b) age 18 or older. There were no exclusion criteria. Based on these criteria, the Internet survey company recruited workers from their potential pool of participants, until the targeted number was reached. If the eligible workers agreed with the terms and conditions of the online survey, they could access the self-report questionnaire. After one month, the company randomly sampled 100 participants from the workers who completed the baseline survey again. Participating workers were awarded approximately 100 'Macromill points' as a reward for each survey, which could be used for cashing out and shopping (one point was equivalent to 1 Japanese yen). Informed consent was obtained from all participants via instructions on the survey. The instructions assured protection of personal information and explained that any identifying information would be removed from the data. The study protocol was approved by the research ethics committee of the Graduate School of Medicine and the Faculty of Medicine, The University of Tokyo, Japan (No. 11242).
Measurements
Participants completed an online self-reporting survey that included the Workplace PERMA-Profiler and questions regarding job and life satisfaction, work engagement, psychological distress, work-related psychosocial factors (job demands, job control, and social support from supervisors and colleagues)22,23), and work performance.
The Workplace PERMA-Profiler
The Japanese version of the Workplace PERMA-Profiler was used to measure multidimensional well-being at work. The measure includes the five factors of the PERMA model (positive emotion, engagement, relationships, meaning, and accomplishment), as well as overall happiness at work, negative emotion, health, and loneliness. Each factor score of the Workplace PERMA-Profiler was calculated as an average of the item scores. An overall score of well-being at work was calculated as an average of 15 items and happiness (1 item). All items were rated on an 11-point Likert-type scale (ranging from 0 to 10).
The Japanese version of the measure was developed according to the procedure specified in the International Society of Pharmacoeconomics and Outcomes Research (ISPOR) task force guidelines24). First, we obtained permission from the developer of the original Workplace PERMA-Profiler (MLK) to translate the measure into Japanese (preparation). Forward-translation was independently conducted and was followed by reconciliation, back-translation, back-translation review, harmonization, and cognitive debriefing. The back-translation was conducted by two experts in Japanese and English affiliated with the English Language Program of the Kanazawa Institute of Technology, who did not know the purpose of the study. The original developer checked the back-translated measure and made revisions at the back-translation review stage. Cognitive debriefing sessions were conducted with nine Japanese workers who were recruited using snowball sampling, and included a company president and occupational health staff members (occupational doctor, public health nurse, clinical psychologist, and human resource management workers). They were asked to complete the harmonized measure and revise the wording if they had difficulty understanding an item, and their feedback was used for further revision. Results from the different stages were combined to create the final measure. For the full version of the Japanese Workplace PERMA-Profiler, please see Appendix 1.
Job and life satisfaction
Job and life satisfaction were measured by questions from the Brief Job Stress Questionnaire (BJSQ)25). 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.
Work engagement
The nine-item Japanese version of the Utrecht Work Engagement Scale (UWES) was used to assess work engagement26). The UWES consists of three subscales: vigor (three items, e.g., 'At my job, I feel strong and vigorous'), dedication (three items, e.g., 'I am enthusiastic about my job'), and absorption (three items, e.g., 'I am immersed in my work'). All items are rated on a seven-point Likert scale (0 = Never, 6 = Always). The reliability and unidimensional validity of the Japanese version of the UWES were confirmed in a previous study26). The scores from each of the nine items were averaged and used for analyses (Cronbach's alpha (α) = 0.96).
Psychological distress
Two scales were used to measure non-specific and specific psychological distress. Non-specific psychological distress was measured by the Japanese version of the K6 scale27). The scale consisted 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 study27). In this study, the total continuous scores on the Japanese version of the K6 were used for analyses (α = 0.91).
Specific types of psychological distress were also measured by questions from the BJSQ25): vigor (three items, e.g., 'I have been very active'; α = 0.93), irritation (three items, e.g., 'I have felt angry'; α = 0.91), fatigue (three items, e.g., 'I have felt extremely tired'; α = 0.91), anxiety (three items, e.g., 'I have felt tense'; α = 0.82), and depression (six items, e.g., 'I have felt depressed'; α = 0.93). The BJSQ has been widely used in Japan to assess responses to stress and has demonstrated satisfactory internal consistency, test-retest reliability, convergent validity, and predictive validity for the onset of depression28). All items are rated on a four-point Likert scale (1 = Almost never, 4 = Almost always).
Work-related psychosocial factors
Job demands (three items, e.g., 'I have an extremely large amount of work to do'; α = 0.83), job control (three items, e.g., 'I can work at my own pace'; α = 0.83), and social support from supervisors (three items, e.g., 'How reliable are your superiors when you are troubled?'; α = 0.84) and colleagues (three items, e.g., 'How freely can you talk with your co-workers?'; α = 0.87) were also measured by the BJSQ25). 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.
Work performance
Work performance was assessed using an item from a validated scale, the Japanese short version of the WHO Health and Work Performance Questionnaire (WHO-HPQ)29). The item rated an individual's overall job performance for the past month on a scale of 0 to 10, with 0 being the worst job performance and 10 being the best. The ratings were multiplied by 10 to calculate work performance according to the WHO-HPQ scoring guidelines.
Analysis
To test reliability, some statistical values (Cronbach's alphas, Intra-class Correlation Coefficients, the Standard Error of Measurement, and the Smallest Detectable Change) of the Japanese version of the Workplace PERMA-Profiler were calculated. Confirmatory factor analysis (CFA) and correlational analysis were conducted to test validity. We used PASW statistics version 18 (IBM SPSS software) and Mplus version 7.430) for each analysis.
Internal consistency
To assess internal consistency, Cronbach's alphas were calculated for the total score and for each factor score (i.e., positive emotion, engagement, relationships, meaning, and accomplishment) of the Japanese Workplace PERMA-Profiler. Based on previous research31), the sample size of more than 100 was considered sufficient for methodological quality for Cronbach's alpha. Because a five-factor structure of the measure was confirmed in previous studies13,14), we did not check the dimensionality of the measure but calculated Cronbach's alphas for the total score and each factor score directly.
Test-retest reliability
Intra-class Correlation Coefficients (ICCs) for the total score and each factor score were calculated to assess test-retest reliability across the 1 month period. Although the previous study reported different parameters (Pearson's r) as the standard of test-retest reliability, the sample size can be considered good to excellent when 50-100 participants are recruited in the test-retest reliability analysis31). In addition, the Standard Error of Measurement (SEM) and the Smallest Detectable Change (SDC) were calculated as the standards of measurement error32-34). The SEM describes the standard deviation of repeated measures in one participant, and the 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 error32). The SEM was calculated as (the standard deviation of all testing scores) × √ (1 - ICC)33,34), and the SDC was calculated as 1.96 × √ (2 × SEM)32).
Structural validity
To confirm the five-factor structural validity, CFA was conducted among the 15 items, using a robust maximum likelihood estimation in Mplus30). The original five-factor model (each of three items was explained by the five factors) and a one-factor model (all 15 items were explained by one factor) were assumed and tested in several model fit indices: the chi square (χ2), the Comparative Fit Index (CFI), the Tucker-Lewis Index (TLI), the Root Mean Square Error of Approximation (RMSEA), and the Standardized Root Mean Square Residual (SRMR). We considered the model a good fit if the CFI and TLI exceeded 0.95 and the RMSEA and SRMR was less than 0.0635). Based on a previous study31), the sample size required for factor analysis was at least five to seven times the number of items, with a minimum of 100. Given that the Japanese version of the Workplace PERMA-Profiler has 15 items, an adequate number of participants (N≥ 105) was recruited in the study.
Convergent validity
Pearson's correlation coefficients (r) among the PERMA factors, job and life satisfaction, work engagement, psychological distress, work-related psychosocial factors, and work performance were calculated to examine convergent validity. The minimum effect size for detection in the study was 0.20 (ρ). Based on a sample size calculation using G*Power version 3.1.9.236), the necessary sample size was estimated to be more than 255 in the case of α error probability of 0.05 and power (1 - β) of 0.90. Therefore, an adequate number of participants was recruited in the study.
Results
Characteristics of participants
A flow chart of the participants is shown in Fig. 1. Because the survey company ceased recruitment once the target number of respondents had been reached, the baseline response rate could not be determined. In the one month follow-up survey, 86 of 100 workers randomly sampled from the baseline participants responded to the questionnaire again (response rate = 86.0%). Because the Internet-based survey required the participants to answer all items, there were no missing values on any variables or items. The demographic characteristics of the participants at baseline and follow-up are shown in Table 2. In the baseline survey (N = 310, 155 men and 155 women, mean age = 44.9 ± 13.6), the majority of the participants had graduated from university (42.6%) or had some college (23.5%). Most participants were full-time (50.3%), day-time workers (90.7%) engaged in occupations such as clerical (22.3%), service (18.1%), or professional/technical jobs (18.1%). Most workers were employed by worksites that had less than 50 workers (47.1%), which covered a wide range of job categories such as services (18.7%), manufacturing (16.5%), and medical/welfare (10.6%). Characteristics of the participants in the follow-up survey (N = 86, 40 men and 46 women, mean age = 45.8 ± 13.0) did not differ from those at baseline, and no significant change was observed.
Fig. 1.
Table 2.
Demographic characteristics of the participants
| Baseline Survey(N = 310) | Follow-up Survey(N = 86) | |||
|---|---|---|---|---|
| n (%) | Mean (SD) | n (%) | Mean (SD) | |
| Gender | ||||
| Men | 155 (50.0) | 40 (46.5) | ||
| Women | 155 (50.0) | 46 (53.5) | ||
| Age | 44.9 (13.6) | 45.8 (13.0) | ||
| Educational status | ||||
| Junior high school | 4 (1.3) | 0 (0.0) | ||
| High school | 86 (27.7) | 30 (34.9) | ||
| College | 73 (23.5) | 19 (22.1) | ||
| University | 132 (42.6) | 35 (40.7) | ||
| Graduate school | 16 (5.2) | 2 (2.3) | ||
| Employment status | ||||
| Full-time | 156 (50.3) | 47 (54.7) | ||
| Part-time | 91 (29.4) | 25 (29.1) | ||
| Contract/Dispatched | 28 (9.0) | 6 (7.0) | ||
| Freelance | 28 (9.0) | 7 (8.1) | ||
| Other | 7 (2.2) | 1 (1.2) | ||
| Employment shift status | ||||
| Day shift | 281 (90.7) | 79 (91.9) | ||
| Rotation/night shift | 29 (9.4) | 7 (8.2) | ||
| Job type | ||||
| Managerial | 26 (8.4) | 8 (9.3) | ||
| Professional/Technical | 56 (18.1) | 13 (15.1) | ||
| Clerical | 69 (22.3) | 21 (24.4) | ||
| Sales | 40 (12.9) | 5 (5.8) | ||
| Services | 56 (18.1) | 16 (18.6) | ||
| Transport/Construction | 12 (3.9) | 2 (2.4) | ||
| Production/Skilled | 31 (10.0) | 13 (15.1) | ||
| Other | 20 (6.5) | 8 (9.3) | ||
| Job category | ||||
| Services | 58 (18.7) | 20 (23.3) | ||
| Manufacturing | 51 (16.5) | 16 (18.6) | ||
| Medical/Welfare | 33 (10.6) | 7 (8.1) | ||
| Retail | 31 (10.0) | 6 (7.0) | ||
| Education | 20 (6.5) | 9 (10.5) | ||
| Construction | 20 (6.5) | 3 (3.5) | ||
| Transport | 16 (5.2) | 4 (4.7) | ||
| Public service | 15 (4.8) | 3 (3.5) | ||
| Financial/Insurance | 14 (4.5) | 5 (5.8) | ||
| Information | 13 (4.2) | 3 (3.5) | ||
| Other | 39 (12.6) | 10 (11.6) | ||
| Size of worksite | ||||
| ≤49 employees | 146 (47.1) | 37 (43.0) | ||
| 50-299 employees | 64 (20.6) | 21 (24.4) | ||
| ≥300 employees | 84 (27.1) | 26 (30.2) | ||
| Unknown | 16 (5.2) | 2 (2.3) | ||
Internal consistency and test-retest reliability
Table 3 shows mean scores, Cronbach's alphas (α), ICCs, SEMs, and SDCs for the PERMA factors. Cronbach's alpha coefficients ranged from 0.75 to 0.96. ICCs ranged from 0.77 to 0.88, meaning that approximately 80% of variance in two time measurements was explained by individuals. SDCs ranged from 1.81 to 2.56.
Table 3.
Mean scores, internal consistency, and reliability of the Japanese version of the Workplace PERMA-Profiler (N = 310)
| Factors | Baseline Mean (SD) | Min-Max | Cronbach's alpha | Follow-up Mean (SD)† | Test-retest Reliability (ICC)† | SME† | SDC† |
|---|---|---|---|---|---|---|---|
| † N = 86. ICC: intra-class correlation coefficient, SME: standard error of measurement, SDC: smallest detectable change. ** p < 0.01 | |||||||
| Positive emotion | 5.46 (2.3) | 0-10 | 0.92 | 5.38 (2.4) | 0.86** | 0.90 | 2.49 |
| P1 | 5.32 (2.5) | 0-10 | 5.35 (2.5) | 0.79** | 1.16 | 3.22 | |
| P2 | 5.76 (2.4) | 0-10 | 5.49 (2.4) | 0.82** | 1.04 | 2.89 | |
| P3 | 5.29 (2.5) | 0-10 | 5.30 (2.6) | 0.77** | 1.29 | 3.57 | |
| Engagement | 5.86 (2.2) | 0-10 | 0.85 | 5.99 (2.1) | 0.83** | 0.87 | 2.42 |
| E1 | 6.05 (2.4) | 0-10 | 6.20 (2.5) | 0.83** | 1.00 | 2.77 | |
| E2 | 5.72 (2.4) | 0-10 | 5.85 (2.3) | 0.76** | 1.13 | 3.14 | |
| E3 | 5.81 (2.6) | 0-10 | 5.92 (2.5) | 0.65** | 1.51 | 4.17 | |
| Relationships | 5.59 (2.0) | 0-10 | 0.75 | 5.60 (2.0) | 0.83** | 0.82 | 2.27 |
| R1 | 6.03 (2.5) | 0-10 | 5.62 (2.5) | 0.69** | 1.34 | 3.71 | |
| R2 | 4.87 (2.3) | 0-10 | 5.28 (2.4) | 0.70** | 1.31 | 3.64 | |
| R3 | 5.88 (2.4) | 0-10 | 5.90 (2.6) | 0.77** | 1.21 | 3.34 | |
| Meaning | 6.24 (2.1) | 0-10 | 0.88 | 6.21 (1.9) | 0.77** | 0.92 | 2.56 |
| M1 | 6.85 (2.3) | 0-10 | 6.77 (2.1) | 0.63** | 1.25 | 3.47 | |
| M2 | 5.92 (2.4) | 0-10 | 5.91 (2.3) | 0.65** | 1.36 | 3.77 | |
| M3 | 5.94 (2.3) | 0-10 | 5.95 (2.2) | 0.75** | 1.11 | 3.07 | |
| Accomplishment | 6.19 (1.9) | 0-10 | 0.84 | 6.29 (2.0) | 0.77** | 0.92 | 2.56 |
| A1 | 5.60 (2.3) | 0-10 | 5.56 (2.2) | 0.68** | 1.26 | 3.49 | |
| A2 | 6.25 (2.2) | 0-10 | 6.49 (2.4) | 0.69** | 1.29 | 3.57 | |
| A3 | 6.73 (2.1) | 0-10 | 6.84 (2.3) | 0.63** | 1.34 | 3.71 | |
| Happiness | 6.01 (2.3) | 0-10 | 6.02 (2.6) | 0.83** | 1.07 | 2.97 | |
| Overall well-being (16 items) | 5.88 (1.8) | 0-10 | 0.96 | 5.90 (1.9) | 0.88** | 0.65 | 1.81 |
| Negative emotion (3 items) | 4.53 (2.1) | 0-10 | 0.78 | 4.48 (2.3) | 0.79** | 1.04 | 2.88 |
| Health (3 items) | 5.77 (2.2) | 0-10 | 0.93 | 5.57 (2.2) | 0.87** | 0.80 | 2.23 |
| Loneliness (1 item) | 4.24 (2.9) | 0-10 | 4.34 (2.8) | 0.64** | 1.71 | 4.73 | |
Structural validity
The results of CFA are shown in Table 4. Of the one-factor and five-factor models, the original five-factor hypothesized model demonstrated marginally acceptable fit (χ2 [80] = 351.30, CFI = 0.892, TLI = 0.858, RMSEA = 0.105, SRMR = 0.051). Standardized covariances among the five factors ranged from 0.73 to 0.97, indicating strong correlations. The five-factor model demonstrated the best fit between the two models compared with the one-factor model (Δχ2 [10] = 297.13, p < 0.05).
Table 4.
Factor loadings of the 15 PERMA items, factor correlations, and model fit in confirmatory factor analyses
| Items | Factor loadings | Correlation coefficients in the 5-factor model | |||||||
|---|---|---|---|---|---|---|---|---|---|
| 1-factor model | 5-factor model | ||||||||
| The robust maximum likelihood estimation method was used. *p < 0.05. | |||||||||
| P1 | 0.87* | 0.88* | F1 (P) | F2 (E) | F3 (R) | F4 (M) | F5 (A) | ||
| P2 | 0.83* | 0.88* | F1 (P) | 1.00 | |||||
| P3 | －0.27* | 0.91* | F2 (E) | 0.94* | 1.00 | ||||
| E1 | 0.84* | 0.77* | F3 (R) | 0.92* | 0.78* | 1.00 | |||
| E2 | 0.76* | 0.86* | F4 (M) | 0.89* | 0.97* | 0.73* | 1.00 | ||
| E3 | 0.45* | 0.78* | F5 (A) | 0.90* | 0.89* | 0.77* | 0.95* | 1.00 | |
| R1 | 0.75* | 0.56* | Model fit | 1-factor | 5-factor | ||||
| R2 | －0.21* | 0.71* | χ2 (df) | 648.43 (90) * | 351.30 (80) * | ||||
| R3 | 0.68* | 0.83* | CFI | 0.705 | 0.892 | ||||
| M1 | －0.07 | 0.80* | TLI | 0.656 | 0.858 | ||||
| M2 | －0.22* | 0.84* | RMSEA (95% CI) | 0.141 (0.131, 0.152) | 0.105 (0.094, 0.116) | ||||
| M3 | 0.52* | 0.87* | SRMR | 0.100 | 0.051 | ||||
| A1 | 0.86* | 0.90* | 1-factor model vs. 5-factor model: Δχ2 (df) | 297.13 (10) * | |||||
| A2 | 0.47* | 0.70* | |||||||
| A3 | 0.68* | 0.67* | |||||||
Convergent validity
Table 5 shows Pearson's correlation coefficients (r) among the PERMA factors, job and life satisfaction, work engagement, psychological distress, work-related psychosocial factors, and work performance. The overall well-being score and five PERMA factors had strong positive correlations with job satisfaction and work engagement (0.60 ≤ r ≤ 0.82). In addition, they had small to moderate positive correlations with life satisfaction (0.19 ≤ r ≤ 0.34). Moreover, the PERMA factors were moderately negatively correlated with non-specific psychological distress (−0.53 ≤ r ≤ −0.39). With regards to specific types of psychological distress, they had comparatively strong correlations with vigor (0.47 ≤ r ≤ 0.58) and depression (−0.53≤ r ≤ −0.38). Among work-related psychosocial factors, job control and social support were moderately to strongly associated with PERMA factors (0.32 ≤ r ≤ 0.60). Self-reported work performance also had moderate to strong positive associations with the PERMA factors (0.48 ≤ r ≤ 0.73). Only job demands had comparatively weak associations (0.01 ≤ r ≤ 0.20).
Table 5.
Convergent validity (r) of the Japanese version of the Workplace PERMA-profiler (N = 310)
| Variables | Mean (SD) | P | E | R | M | A | Overall |
|---|---|---|---|---|---|---|---|
| BJSQ: brief job stress questionnaire, UWES: Utrecht work engagement scale, HPQ: health performance questionnaire. *p < 0.05, **p < 0.01. | |||||||
| Workplace PERMA-Profiler | |||||||
| Positive emotion (P) | 5.46 (2.3) | 1.00 | |||||
| Engagement (E) | 5.86 (2.2) | 0.83** | 1.00 | ||||
| Relationships (R) | 5.59 (2.0) | 0.74** | 0.64** | 1.00 | |||
| Meaning (M) | 6.24 (2.1) | 0.79** | 0.81** | 0.58** | 1.00 | ||
| Accomplishment (A) | 6.19 (1.9) | 0.73** | 0.70** | 0.59** | 0.75** | 1.00 | |
| Overall well-being | 5.88 (1.8) | 0.94** | 0.91** | 0.81** | 0.89** | 0.85** | 1.00 |
| Negative emotion | 4.53 (2.1) | －0.35** | －0.21** | －0.28** | －0.17** | －0.25** | －0.30** |
| Health | 5.77 (2.2) | 0.50** | 0.38** | 0.47** | 0.41** | 0.46** | 0.51** |
| Loneliness | 4.24 (2.9) | －0.38** | －0.34** | －0.43** | －0.32** | －0.31** | －0.41** |
| Satisfaction | |||||||
| Job satisfaction (BJSQ) | 2.59 (0.9) | 0.75** | 0.70** | 0.60** | 0.64** | 0.61** | 0.76** |
| Life satisfaction (BJSQ) | 2.79 (0.9) | 0.32** | 0.21** | 0.34** | 0.19** | 0.30** | 0.32** |
| Work engagement (UWES) | 2.79 (1.3) | 0.77** | 0.79** | 0.61** | 0.72** | 0.69** | 0.82** |
| Psychological distress (K6) | 6.52 (5.4) | －0.53** | －0.39** | －0.49** | －0.42** | －0.43** | －0.52** |
| Psychological distress (BJSQ) | |||||||
| Vigor | 6.45 (2.4) | 0.58** | 0.51** | 0.48** | 0.51** | 0.47** | 0.59** |
| Irritation | 6.64 (2.5) | －0.31** | －0.22** | －0.37** | －0.15** | －0.25** | －0.31** |
| Fatigue | 6.66 (2.6) | －0.41** | －0.27** | －0.32** | －0.26** | －0.32** | －0.37** |
| Anxiety | 6.17 (2.4) | －0.35** | －0.19** | －0.28** | －0.21** | －0.29** | －0.31** |
| Depression | 11.03 (4.7) | －0.53** | －0.38** | －0.45** | －0.39** | －0.42** | －0.50** |
| Job demands (BJSQ) | 7.73 (2.3) | 0.04 | 0.18** | 0.01 | 0.20** | 0.01 | 0.10 |
| Job control (BJSQ) | 7.98 (2.4) | 0.42** | 0.36** | 0.27** | 0.35** | 0.36** | 0.40** |
| Social support from supervisors (BJSQ) | 7.46 (2.1) | 0.41** | 0.34** | 0.53** | 0.32** | 0.33** | 0.45** |
| Social support from colleagues (BJSQ) | 7.32 (2.4) | 0.45** | 0.34** | 0.60** | 0.35** | 0.32** | 0.47** |
| Work performance (HPQ) | 61.84 (19.1) | 0.57** | 0.54** | 0.48** | 0.55** | 0.73** | 0.65** |
Discussion
In this study, the Japanese version of the Workplace PERMA-Profiler demonstrated good reliability and convergent validity, with adequate structural validity. Well-being at work was associated with not only health outcomes but also work-related psychosocial factors and work performance. Indeed, the PERMA factors were more strongly related to job satisfaction than to life satisfaction, suggesting that the concepts of the original PERMA-profiler and the Workplace PERMA are critically distinct. This measure could be applicable for assessment of well-being at work among Japanese workers.
The measure demonstrated strong internal consistency, and was generally stable over a one month period. Measurement error was low. Meaningful differences in well-being at work could be detected around 2 points within the 11-point Likert scale of the scores, and may be useful for future intervention studies.
Convergent validity was also well supported. The effect sizes for health outcomes were consistent with the previous validation study14). In addition, work engagement indicated the strongest positive correlations with the measure, especially with the engagement (E) dimension. The associations with work-related psychosocial factors (job demands, job control, and social support) were also similar with those of work engagement19). Though the relationships with job demands were weak, this can be explained by the job demands-resource model37). In this model, job demands can cause deterioration of mental illness and do not strongly affect positive outcomes (i.e., work engagement). Well-being at work was not strongly correlated with job demands.
The PERMA domains were strongly related to job satisfaction and work performance. While the original PERMA-Profiler14) had strong positive correlations with life satisfaction and weak positive correlations with work performance, the Workplace PERMA-Profiler had weak correlations with life satisfaction and strong correlations with job satisfaction and work performance. Indeed, the correlations were stronger here than in prior studies19,26,38). Future studies should investigate the extent to which the measure can predict future work performance and productivity.
The CFA did not completely support the five-factor PERMA model of the measure, and the different factors are strongly correlated with one another. However, the original PERMA-profiler14) demonstrated similar values to this study (CFI = 0.894, TLI = 0.864, RMSEA = 0.107). The lack of good model fit could occur for multiple reasons. Seligman13) argues that the five PERMA domains are separate, measurable dimensions of well-being. First, the model itself could be wrong, such that while the theory distinguishes different factors, the everyday worker does not. Second, the measure itself could be wrong, such that the current items do not adequately distinguish the five factors. Future studies might further investigate the items, using qualitative interviews and other approaches to better understand how respondents understand each item, and whether the factors can be pulled apart psychometrically. Third, the PERMA model may not be the most appropriate model for workplaces in general, or for the Japanese workplace in particular. Prior workplace wellbeing models have focused on affective (positive and negative emotion), evaluative (job satisfaction), and work engagement dimensions19,26,38). The PERMA model further breaks apart these dimensions, which may not be a helpful distinction in the workplace. Still, from a practical perspective, the PERMA domains provide specific areas to intervene (e.g., the quality of one's relationships, one's sense of competence at work), which are more tangible than the broader domains (e.g., overall job satisfaction)39,40). Future studies might further investigate the multidimensional structure of workplace well-being, the extent to which PERMA versus other well-being models are most appropriate for the workplace in general and within the Japanese culture, as well as possible practical applications of the model.
Several limitations exist in this study. First, because the response rate could not be calculated, selection bias might exist. For instance, participants who were unhealthy and had low well-being may have been reluctant to participate in the survey. Second, there could be measurement errors in the assessment of the standards of convergent validity. Third, other confounders not measured in the study might distort the results of correlation analyses, such as psychological capital (e.g., self-efficacy, optimism, and intrinsic motivation). Finally, as mentioned previously, the generalizability of the results for Japanese workers could be questioned due to the use of an online survey.
In conclusion, the Japanese version of the Workplace PERMA-Profiler indicated good reliability and validity. This measure could be useful to assess well-being at work, promote well-being research among Japanese workers, and address the problem of defining well-being in further studies.
Acknowledgments: This work is supported by the Health and Labor Sciences Research Grant 2015-2017 (H27-Rodo-Ippan-004) from the Ministry of Health, Labour and Welfare, Japan. The funders had no role in study design, data collection and analysis, decision to publish, or preparation of the manuscript.
We thank Kiyomi Fujii and Brent Wright for their cooperation in the back-translation of the Japanese version of the Workplace PERMA Profiler.
Conflicts of interest: None declared.
Supplementary material: This article contains supplementary material (Apendix), which is available in the online version (doi: 10.1539/joh.2018-0050-OA).
Supplementary Material
References
- 1). Huppert FA, So TTC. Flourishing across Europe: application of a new conceptual framework for defining well-being. Soc Indic Res 2013; 110 (3): 837-861. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 2). Steptoe A, Deaton A, Stone AA. Subjective wellbeing, health, and ageing. Lancet 2015; 385 (9968): 640-648. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 3). Diener E, Chan MY. Happy people live longer: subjective well-being contributes to health and longevity. Appl Psychol: Health Well Being 2011; 3 (1): 1-43. [Google Scholar]
- 4). Howell RT, Kern ML, Lyubomirsky S. Health benefits: meta-analytically determining the impact of wellbeing on objective health outcomes. Health Psychol Rev 2007; 1 (1): 83-136. [Google Scholar]
- 5). Schulte P, Vainio H. Well-being at work-overview and perspective. Scand J Work Environ Health 2010; 36 (5): 422-429. [DOI] [PubMed] [Google Scholar]
- 6). Mills MJ, Fleck CR, Kozikowski A. Positive psychology at work: a conceptual review, state-of-practice assessment, and a look ahead. J Posit Psychol 2013; 8: 153-164. [Google Scholar]
- 7). Dodge R, Daly AP, Huyton J, et al. . The challenge of defining wellbeing. Int J Wellbeing 2012; 2 (3): 222-235. [Google Scholar]
- 8). Chida Y, Steptoe A. Positive psychological well-being and mortality: a quantitative review of prospective observational studies. Psychosom Med 2008; 70 (7): 741-756. [DOI] [PubMed] [Google Scholar]
- 9). Ryan RM, Deci EL. On happiness and human potentials: a review of research on hedonic and eudaimonic well-being. Ann Rev Psychol 2001; 52 (1): 141-166. [DOI] [PubMed] [Google Scholar]
- 10). Diener E, Suh EM, Lucas RE, et al. . Subjective well-being: three decades of progress. Pscyhol Bull 1999; 125 (2): 276-302. [Google Scholar]
- 11). Ryff CD. Psychological well-being revisited: advances in the science and practice of eudaimonia. Psychother Psychosom 2014; 83 (1): 10-28. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 12). Page KM, Vella-Brodrick DA. The 'what', 'why' and 'how' of employee well-being: a new model. Soc Indic Res 2009; 90 (3): 441-458. [Google Scholar]
- 13). Seligman MEP. Flourish: a visionary new understanding of happiness and well-being. New York, NY: Free Press; 2011. [Google Scholar]
- 14). Butler J, Kern ML. The PERMA-Profiler: a brief multidimensional measure of flourishing. Int J Wellbeing 2016; 6 (3): 1-48. [Google Scholar]
- 15). Lamers SM, Westerhof GJ, Bohlmeijer ET, et al. . Evaluating the psychometric properties of the Mental Health Continuum-Short Form (MHC-SF). J Clin Psychol 2011; 67 (1): 99-110. [DOI] [PubMed] [Google Scholar]
- 16). Ryff CD. Happiness is everything, or is it? explorations on the meaning of psychological well-being. J Personal Soc Psychol 1989; 57 (6): 1069-1081. [Google Scholar]
- 17). Kern ML. The Workplace PERMA Profiler. [Online]. 2014 Available from: URL: http://www.peggykern.org/uploads/5/6/6/7/56678211/workplace_perma_profiler_102014.pdf
- 18). Schaufeli WB, Salanova M, Gonzalez-Roma V, et al. . The measurement of engagement and burnout: a confirmative analytic approach. J Happiness Stud 2002; 3 (1): 71-92. [Google Scholar]
- 19). Halbesleben JRB. A meta-analysis of work engagement: relationships with burnout, demands, resources, and consequences. In: Bakker AB, Leiter MP, editors. Work engagement: a handbook of essential theory and research. New York, NY: Psychology Press; 2010. p. 102-117. [Google Scholar]
- 20). Mokkink LB, Terwee CB, Patrick DL, et al. . The COSMIN checklist for assessing the methodological quality of studies on measurement properties of health status measurement instruments: an international Delphi study. Qual Life Res 2010; 19 (4): 539-549. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 21). Corporate profile. Macromill, Inc. [Online]. 2017 Available from: URL: https://www.macromill.com/company/profile.html
- 22). Karasek RA. Job demands, job decision latitude, and mental strain; implications for job redesign. Adm Sci Q 1979; 24 (2): 285-308. [Google Scholar]
- 23). Johnson JV, Hall EM. Job strain, workplace social support, and cardiovascular disease: a cross-sectional study of a random sample of the Swedish working population. Am J Public Health 1988; 78 (10): 1336-1342. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 24). Wild D, Grove A, Martin M, et al. ; ISPOR Task Force for Translation and Cultural Adaptation . Principles of good practice for the translation and cultural adaptation process for patient-reported outcomes (PRO) measures: report of the ISPOR task force for translation and cultural adaptation. Value Health 2005; 8 (2): 94-104. [DOI] [PubMed] [Google Scholar]
- 25). Shimomitsu T, Haratani T, Nakamura K, et al. . Final development of the Brief Job Stress Questionnaire mainly used for assessment of the individuals. In: Kato M. The Ministry of Labor Sponsored Grant for the Prevention of Work- Related Illness. Tokyo: Tokyo Medical University; 2000. p. 126-164 (in Japanese). [Google Scholar]
- 26). Shimazu A, Schaufeli WB, Kosugi S, et al. . Work engagement in Japan: Validation of the Japanese version of the Utrecht Work Engagement Scale. Appl Psychol 2008; 57 (3): 510-523. [Google Scholar]
- 27). Furukawa TA, Kawakami N, Saitoh M, et al. . The performance of the Japanese version of the K6 and K10 in the World Mental Health Survey Japan. Int J Methods Psychiatr Res 2008; 17 (3): 152-158. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 28). Wada K, Sairenchi T, Haruyama Y, et al. . Relationship between the onset of depression and stress response measured by the Brief Job Stress Questionnaire among Japanese employees: a cohort study. PLoS One 2013; 8 (2): e56319. (doi: 10.1371/journal.pone.0056319). [DOI] [PMC free article] [PubMed] [Google Scholar]
- 29). Kessler RC, Barber C, Beck A, et al. . The World Health Organization Health and Work Performance Questionnaire (HPQ). J Occup Environ Med 2003; 45 (2): 156-174. [DOI] [PubMed] [Google Scholar]
- 30). Muthén LK, Muthén BO. Mplus User's Guide. Seventh edition. [Online]. Los Angeles, CA, 1998-2015 Available from: URL: http://www.statmodel.com/ugexcerpts.shtml
- 31). Terwee CB, Mokkink LB, Knol DL, et al. . Rating the methodological quality in systematic reviews of studies on measurement properties: a scoring system for the COSMIN checklist. Qual Life Res 2012; 21 (4): 651-657. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 32). van Kampen DA, Willems WJ, van Beers LW, et al. . Determination and comparison of the smallest detectable change (SDC) and the minimal important change (MIC) of four-shoulder patient-reported outcome measures (PROMs). J Orthop Surg Res 2013; 8: 40. (doi: 10.1186/1749-799X-8-40). [DOI] [PMC free article] [PubMed] [Google Scholar]
- 33). Weir JP. Quantifying test-retest-reliability using the intraclass correlation coefficient and the SEM. J Strength Cond Res 2005; 19 (1): 231-240. [DOI] [PubMed] [Google Scholar]
- 34). Lu WS, Wang CH, Lin JH, et al. . The minimal detectable change of the simplified stroke rehabilitation assessment of movement measure. J Rehabil Med 2008; 40 (8): 615-619. [DOI] [PubMed] [Google Scholar]
- 35). Hu L, Bentler PM. Cutoff criteria for fit indexes in covariance structure analysis: conventional criteria versus new alternatives. Struct Equ Modeling 1999; 6 (1): 1-55. [Google Scholar]
- 36). Faul F, Erdfelder E, Buchner A, et al. . Statistical power analyses using G*Power 3.1: tests for correlation and regression analyses. Behav Res Methods 2009; 41 (4): 1149-1160. [DOI] [PubMed] [Google Scholar]
- 37). Bakker AB, Demerouti E. The job demands-resources model: state of the art. J Manag Psychol 2007; 22 (3): 309-328. [Google Scholar]
- 38). Shimazu A, Schaufeli WB, Kubota K, et al. . Do workaholism and work engagement predict employee well-being and performance in opposite directions? Ind Health 2012; 50 (4): 316-321. [DOI] [PubMed] [Google Scholar]
- 39). Kern ML, Waters LE, Adler A, et al. . A multidimensional approach to measuring well-being in students: application of the PERMA framework. J Posit Psychol 2015; 10 (3): 262-271. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 40). McQuaid M, Kern P. Your wellbeing blueprint: Feeling good and doing well at work. Australia: McQuaid Ltd; 2017. [Google Scholar]
Associated Data
This section collects any data citations, data availability statements, or supplementary materials included in this article.