Testing the psychometric properties of the Swedish version of the EPOCH measure of adolescent well-being.

The EPOCH Measure of Adolescent Well-being measures five positive indicators of the well-being of adolescents: engagement, perseverance, optimism, connectedness and happiness. This five-factor structure along with other indicators of validity and reliability were supported for the original English version and the Chinese version. In this study, we tested the psychometric properties of the Swedish version of the EPOCH with a sample (n = 846) of Swedish high school adolescents aged 16-21 years (Mage = 18, SD = .85). The participants answered a questionnaire containing the EPOCH, Coping Self-Efficacy Scale, and 21-item Depression, Anxiety, and Stress Scale (DASS-21). A confirmatory factor analysis supported a the five-factor, inter-correlated model. The internal consistency was good for all the EPOCH subscales (Cronbach's α = .76-.88, McDonald's ω = .77 -.88). The criterion validity was established by replicating correlations between the five EPOCH subscales and positive (coping self-efficacy) and negative (DASS-21) aspects of well-being. This study shows that the Swedish version of the EPOCH is suitable for assessing multiple dimensions of adolescent well-being.

Introduction

Positive mental health is conceptualized not only as a lack of mental health problems such as anxiety, stress and depression, but also by the presence of positive psychological well-being indicators such as perseverance, optimism and happiness [1,2]. Measuring adolescent well-being using a multidimensional approach—that is, encompassing multiple positive indicators—is important, given the multitude of concepts relevant for young people’s well-being [3]. Furthermore, as well-being comprises both hedonic (“feeling good”) and eudaimonic (“functioning well” [4]) aspects, it is important to take both into consideration when measuring well-being. In Sweden, a valid multidimensional measure of adolescent well-being does not yet exist. This study therefore aims to validate the Swedish version of the EPOCH Measure of Adolescent Well-being (hereafter, EPOCH) [2]. The EPOCH measure [2] has the theoretical strength of being both a multidimensional measure (consisting of five factors of well-being: engagement, perseverance, optimism, connectedness and happiness) and a measure of both the hedonic and eudaimonic aspects of well-being, discussed more next. Furthermore, the EPOCH measure is designed to operationalize PERMA theory of well-being [5] for adolescents. The PERMA model (discussed in the next section) is the main model used in positive educational studies; accordingly, having a valid measure of it will be important particularly in the positive educational research context.

Well-being theories and the EPOCH measure of adolescent well-being

Traditionally, well-being has been approached from either the hedonic or the eudaimonic viewpoint [4,6], although debate about the validity of this distinction also exists [7,8].Hedonic well-being broadly refers to emotional well-being, including frequent positive emotions and infrequent negative emotions, as well as a sense of life satisfaction [4]. Student life satisfaction is commonly measured with the Students’ Life Satisfaction Scale [9], which measures students’ context-free estimation of their life satisfaction as a whole. This scale has been used in both clinical and non-clinical studies with student samples ranging in age from 8 to 18 years [10,11]; however, it is not a multidimensional measure. Separating the different dimensions of well-being can help to recognize areas in which students particularly flourish or flounder, and can suggest further efforts for different types of intervention in areas that are lacking [3]. For instance, the Multidimensional Students’ Life Satisfaction Scale [12,13] for 8- to 18-year-olds measures student life satisfaction across five important domains: school, family, friends, self and living environment.

Eudaimonic well-being refers to well-being resources and positive functioning, such as a sense of meaning in life, authenticity, personal growth, autonomy and positive relationships [4,6,14]. Although thorough consideration of the debate about the validity of distinguishing between various types of well-being is beyond the scope of this article, some researchers advocate for the distinction of hedonia from eudaimonia [e.g. 4,6,15], whereas others oppose this distinction [e.g. 7,8]. For instance, Longo et al. [16] studied the factor structure of Huppert and So’s [17] 10 well-being indicators—happiness, emotional stability, vitality, resilience (aspects of “feeling good”, or hedonia), competence, engagement, meaning, optimism, positive relationships and self-esteem (aspects of positive functioning, or eudaimonia)—in two different studies and found that instead of two distinct factors, the items loaded onto a single higher-order well-being factor. However, two factor analytic studies, Linley et al. [18] and Joshanloo [19], found evidence for distinct factors. In any case, a full account of well-being arguably requires consideration of both the hedonic and eudaimonic aspects. One theory encompassing both aspects is the PERMA theory of well-being [5], which approaches well-being as a multidimensional concept consisting of positive emotion, engagement, positive relationships, meaning and achievement.

The PERMA model is used commonly as a theoretical framework for school applications of positive psychology, called positive education [20], with an emphasis on providing a multidimensional view of well-being for school youth [20]. Positive education involves bringing the science of well-being into schools through well-being lessons or through influencing the school culture in other ways [20]. The PERMA profiler, a measure developed for the PERMA model, was validated in its original development study with an international sample recruited online [21] and in Turkish [22], Australian [23], Italian [24] and Indonesian [25] samples. Although some results were mixed or unsatisfactory [e.g. 26], generally the results have supported the five-factor model with high internal and test-retest reliability, factor structure and construct validity [21,24,25].

Since the PERMA model is the main theoretical framework used in positive educational studies [20,27], it is important to validate a measure based on it for adolescents. Kern et al. [2] wanted to translate the PERMA model to better suit adolescents and therefore operationalized a measure of optimal adolescent functioning that reflects both the attitudes and characteristics related to optimal outcomes for adolescents: the EPOCH Measure of Adolescent Well-being. The EPOCH consists of five subcomponents, including engagement, perseverance, optimism, connectedness and happiness. The subcomponents reflect the PERMA categories, albeit using different language, and relate to similar outcomes according to Kern et al. [2]. The motivation for including characteristics (i.e. personality traits such as perseverance) in a definition of well-being is practical: the EPOCH was developed to be used in particular positive educational intervention contexts, in which different aspects relevant for optimal adolescent functioning should be assessed in order to develop, improve upon, and evaluate interventions. Kern et al. [2] claim that using the EPOCH categories can help in developing more targeted interventions and show where adolescents might be lacking in particular. Although not a diagnostic tool, the EPOCH can aid in the development and assessment of interventions [2] aiming to promote other positive psychological constructs, such as optimism and perseverance, on top of happiness.

In the EPOCH, engagement refers to the ability to become absorbed in what one is doing, with its most intense form referring to a sense of ‘flow’ as defined by Csikszentmihalyi [28]– a state of complete absorption is what one is doing with the loss of a sense of time and self. Perseverance is the ability to keep pursuing one’s activities and goals even in the case of setbacks on one’s way. Similar concepts include grit, defined by Duckworth et al. [29] as ‘passion and perseverance to long-term goals’. Optimism refers to a sense of hopefulness about the future along with a tendency to have an explanatory style in which positive events are attributed to global and internal causes (i.e. one feels a sense of agency in making positive events possible), whereas negative events are attributed to specific and external events (i.e. one feels not directly accountable for negative events and these events are just temporary occurrences) [2,30]. Connectedness refers to one’s sense of being cared for as a person by others through positive interpersonal relationships. Finally, happiness refers to the common occurrence of positive emotionality, such as joy and a love of life [2].

Support was found for this five-factor structure of the EPOCH, both for the English [2] and Chinese versions [31]. Kern and colleagues [2] evaluated the factor structure of the original English EPOCH scale when testing the psychometric properties of the scale via 10 different studies with over 4,000 adolescents from the USA and Australia. The five EPOCH subscale scores were found to have high internal consistency (happiness had the highest at α = .87, and engagement the lowest at α = .76). The subscale scores were not correlated with age and gender but were weakly to strongly correlated with other well-being indicators (significant correlations ranged from r = .12 for optimism and autonomy to r = .83 for happiness and life satisfaction) and negatively correlated with mental health symptoms (ranging from r = -.29 for optimism and anxiety to r = -.53 for happiness and depression). Furthermore, Zeng and Kern [31] tested the Chinese version of the EPOCH with a sample of 17,854 adolescents from different regions in China and found that the five-factor structure was supported with good model fit. The subscale scores were internally consistent, but its test-retest reliability was low (with a range from r = .12 to r = .21). The criterion validity was good: the subscale scores were consistently positively correlated with various well-being indicators (ranging from r = .16 between happiness and growth mindset to r = .57 between happiness and coping) and negatively with mental health symptoms (ranging from zero correlation between engagement and anxiety to r = -.25 between happiness and depression). The correlations with mental health symptoms were slightly weaker than were those with the positive indicators. In general, engagement showed the weakest correlations with well-being indicators as compared to the other subscales, while optimism and happiness showed the strongest correlations.

Current study

In this study, we investigated the psychometric properties of the Swedish version of the EPOCH Measure of Adolescent Well-being. To establish its criterion validity, we evaluated its correlations with coping self-efficacy, a positive indicator of well-being [32], and anxiety, stress and depression symptoms, all of which are negative indicators [33].

In line with earlier studies [2,31], we expect the Swedish version of the EPOCH to have a five-factor structure with reliable subscales. Furthermore, we expect to find a positive association between the EPOCH and coping self-efficacy and a negative association between the EPOCH and mental health symptoms. More specifically, we expect that the engagement subscale will show the weakest association with well-being and mental health indicators (coping self-efficacy and DASS-21, respectively), while the happiness subscale will show the strongest association (thus replicating [2,31]). Finally, based on previous research [2], we also expect the EPOCH to have weak relationships with gender.

Materials and methods

Participants

The study was administered online on the school’s webpage. Interested students could click on the study link. Altogether, 1212 students opened the study link and out of those 852 (70.3%) responded. We excluded six participants because they did not correctly answer two or more control questions randomly placed in the survey to assess attentiveness, leaving a total of 846 participants. The participants came from three schools matched in terms of study programmes, sizes and reputation for having highly motivated students in three Swedish cities. Participants ranged in age from 16 to 21 years, with a mean age of 18 years (SD = .85). The majority of the participants (n = 555, 65.6%) were female, 286 (33.8%) were male, and 5 (0.6%) reported ‘other’. Most participants were born in Sweden (n = 722, 85.3%), with either one or both parents being born in Sweden (n = 575, 68%). All the students were enrolled in a Swedish school in a Swedish-speaking program. Therefore, their Swedish was considered fluent for the purposes of validating the EPOCH.

Procedure

The data were collected during the baseline assessment of a larger project assessing students’ mental health during the COVID-19 pandemic. The battery of measures was distributed through the websites of three schools in three Swedish cities in May 2020 to be completed at home, since all schools were closed due to the pandemic. The school administration agreed to participate in the study. Ethical approval for the study was obtained from the Swedish Ethical Review Authority. All students were over age 15, meaning that, according to Swedish ethical guidelines, they could give their own informed consent and no parental informed consent was required. The students were given information about the study, which emphasized that their responses would be kept anonymous and confidential. All participants provided informed consent before participating in this study by selecting a box on an online form, indicating their understanding of the nature of the study and that they agreed to participate.

Measures

EPOCH measure of adolescent well-being

All participants responded to the 20-item EPOCH, which contains subscales of engagement (e.g. ‘I get completely absorbed in what I am doing’), perseverance (e.g. ‘I finish whatever I begin’), optimism (e.g. ‘I think good things are going to happen to me’), connectedness (e.g. ‘When I have a problem, I have someone who will be there for me’) and happiness (e.g. ‘I love life’). Each item is rated on a 5-point Likert scale ranging from 1 ‘Almost never’ to 5 ‘Almost always’. A back-translation process was used to translate the scale into Swedish [34]. The scale was translated into Swedish by two native Swedish speakers, and then back-translated into English. Another native English speaker checked the back-translation for similarity of meaning with the original scale. There was high level of agreement in the similarity of meanings of all items in the scale. Kern et al. [2] and Zeng and Kern [31] previously found that all subscales had good internal consistency, with Cronbach’s α values ranging from 0.78 to 0.89 [31]. The internal consistency of the EPOCH for this study is reported in the Results section.

Coping self-efficacy scale

The Coping Self-Efficacy Scale (CSE) [32] consists of 13 items in three subscales: using problem-focused coping (e.g. ‘Think about one part of the problem at a time’), stopping unpleasant emotions and thoughts (e.g. ‘Take your mind off unpleasant thoughts’) and getting support from friends and family (e.g. ‘Get emotional support from friends and family’). The items are rated on an 11-point Likert scale ranging from 0 (‘cannot do at all’) to 10 (‘certain can do’). In this study, the internal consistency of the whole scale was high (α = 0.89), as was that of the subscales: using problem-focused coping (α = 0.85), stopping unpleasant emotions and thoughts (α = 0.88), and getting support from friends and family (α = 0.78).

Depression, anxiety and stress scale

The DASS-21 [33] consists of 21 items in three subscales: depression, which measures dysphoria, self-depreciation, lack of interest and hopelessness (e.g. ‘I felt that I had nothing to look forward to’); anxiety, which measures affective experience of anxiety, autonomic arousal and muscle effects (e.g. ‘I felt I was close to panic’); and stress, measuring chronic arousal (e.g. ‘I found it hard to wind down’). The items are rated on a 4-point Likert scale ranging from 0 (‘did not apply to me at all’) to 3 (‘applied to me very much, or most of the time’). In this study, the internal consistency of the whole scale was high (α = 0.91), while the subscales also satisfactory to high internal consistencies: depression (α = 0.89), anxiety (α = 0.73) and stress (α = 0.81).

Data analyses

The factor structure of the Swedish version of the EPOCH was tested using confirmatory factor analysis (CFA) with maximum likelihood estimation and robust standard errors. These analyses were conducted using Mplus version 8 [35]. Replicating Kern et al. [2], in addition to a five-factor model with inter-related latent variables corresponding to each EPOCH subscale, we estimated a one-factor model (i.e. all items loaded onto a single well-being factor) and a second-order model (i.e. items loading onto five factors, which in turn load onto a single overarching well-being latent construct) and compared these latter two models to the first one. The goodness of fit of the models was evaluated using the χ2 statistic, where a nonsignificant value represents an acceptable fit. Because of the chi-square test’s sensitivity to sample size [36], we also computed several approximate fit indices with conventional cutoffs: the root mean square error of approximation (RMSEA) with 95% confidence interval, Tucker-Lewis Index (TLI), comparative fit index (CFI) and standardized root mean square residual (SRMR). Acceptable fit standards are as follows: TLI and CFI ≥ 0.90, RMSEA ≤0.08 and SRMR ≤0.10 [37]. To test measurement invariance across genders, we compared increasingly constrained models to a less constrained model, focusing on the change in CFI (ΔCFI) and RMSEA (ΔRMSEA). We considered ΔCFI ≤ 0.010 and ΔRMSEA ≤ 0.015 [38,39] as indicative of the invariance assumption holding. The internal consistency was estimated with the use of Cronbach’s alpha and McDonald’s ω levels, with a criterion value of >.07 for acceptable consistency [40]. Cronbach’s alpha is a good measure of internal consistency, particularly with multidimensional scales (i.e. those that measure different latent constructs), whereas the McDonald’s ω is better suited measuring the internal consistency of unidimensional scales (i.e. those that measure the same latent construct; [41]). If items are unidimensional, Cronbach’s alpha gives a reliable measure of internal consistency only once all items have equal covariance with the true score, which is seldom the case [41]. Since we are testing both one-factor and five-factor models, we decided to report both indicators of internal consistency. The criterion validity was estimated with Pearson product moment correlations between the EPOCH total score and subscale scores and the indicators of coping self-efficacy (i.e. CSE) [32] and mental health symptoms (i.e. DASS-21) [33].

Results

Preliminary analyses

The data included some missing values (altogether < .01%), but an analysis of missing values with Little’s missing completely at random (MCAR) test yielded a non-significant result, χ2 (5595) = 5614.25, p = 0.46. Thus, the missing values were not related to any variables in our study. Full information maximum likelihood (FIML) was used to handle the missing data.

Construct validity of the EPOCH

The five-factor model showed acceptable model fit (χ2(160) = 818.72, p < .001; CFI = 0.917, TLI = 0.902, RMSEA = 0.070 [CI = 0.065, 0.075], SRMR = 0.044) and fit better than the 1-factor model [χ2(170) = 2530.24, p < .001; CFI = 0.703, TLI = 0.668, RMSEA = 0.128 CI [0.124, 0.133], SRMR = 0.091; Δχ2(10) = 1711.52, p < .001;) and second-order factor model (χ2(165) = 871.78, p < .001; CFI = 0.911, TLI = 0.898, RMSEA = 0.071 (CI = 0.067, 0.076), SRMR = 0.050; Δχ2(5) = 53.06, p < .001), even though the latter model’s fit was also acceptable. The CFI and TLI were both above the desired cut-off of .9 (TLI was about .9 for the second-order model), the RMSEA was below the critical .08 threshold, and the SRMR was below .09 [37] for both the five-factor and the second-order model. Therefore, both total scale scores and subscale scores are used in analyses. All items had loadings of >.40 onto their respective factors and were significant at p < .01 in the five-factor model. Scalar invariance was supported for gender (see Table 1), indicating that the five-factor model is sufficient and that mean values are directly comparable across the genders. The five-factor solution of the EPOCH is presented in Fig 1.

Table 1

Comparing configural, metric and scalar invariance across gender.

Invariance	df	AIC	BIC	χ2	Δχ2	p	CFI	RMSEA	ΔCFI	ΔRMSEA
Configural	320	40743.66	41406.50	986.22			.916	.070
Metric	335	40747.75	41339.58	1020.31	34.09	.003	.914	.070	.002	.000
Scalar	350	40789.69	41307.50	1089.25	68.94	< .001	.907	.071	.007	.001

Fig 1

Confirmatory factor analysis of the EPOCH item loadings onto their subcomponents, showing the standardized factor loadings and latent correlations (some items are abbreviated).

Internal consistency

The total EPOCH scale had high internal consistency (Cronbach’s α = 0.91, McDonald’s ω = 0.92). The subscales likewise all had high or acceptable consistencies: engagement (α = 0.79, McDonald’s ω = 0.80), perseverance (α = 0.76, McDonald’s ω = 0.77), optimism (α = 0.83, McDonald’s ω = 0.84), connectedness (α = 0.74, McDonald’s ω = 0.75) and happiness (α = 0.88, McDonald’s ω = 0.89).

Criterion validity

The intercorrelations were estimated with Pearson product moment correlation coefficients between the total and subscales of the EPOCH and positive and negative well-being indicators (CSE and DASS-21, respectively, see Table 2). Moreover, all item intercorrelations were estimated (see Table 3).

Table 2

Descriptive statistics for the EPOCH subscales and intercorrelations between the EPOCH subscales and overall scale (N = 846).

	Engagement	Perseverance	Optimism	Connectedness	Happiness	Overall
E	─					.64
P	.37	─				.69
O	.37	.49	─			.84
C	.27	.34	.52	─		.71
H	.38	.43	.72	.61	─	.85
M (SD)	3.08 (.084)	3.74 (0.78)	3.65 (0.89)	4.41 (0.69)	3.69 (0.91)	3.72(0.62)

E = Engagement; P = Perseverance; O = Optimism; C = Connectedness; H = Happiness.

All correlations were significant at p < .01.

Table 3

EPOCH scale item intercorrelations and item descriptives (N = 846).

EPOCH scale items
	M(SD)	1.	2.	3.	4.	5.	6.	7.	8.	9.	10.	11.	12.	13.	14.	15.	16.	17.	18.	19.	20.
1.	4.43(.92)
2.	3.93(.92)	.23
3.	3.74(1.14)	.37	.44
4.	3.62(1.02)	.41	.35	.57
5.	3.73(1.15)	.19	.15	.21	.21
6.	3.69(.97)	.47	.36	.52	.72	.31
7.	3.07(1.07)	.18	.29	.31	.30	.47	.40
8.	3.55(1.20)	.38	.34	.57	.73	.26	.65	.36
9.	3.24(1.22)	.14	.52	.31	.26	.14	.26	.26	.27
10.	4.01(1.13)	.52	.26	.37	.50	.19	.49	.20	.48	.27
11.	2.57(1.09)	.14	.17	.22	.19	.48	.30	.59	.24	.23	.16
12.	2.95(.99)	.13	.23	.23	.20	.37	.27	.44	.26	.30	.17	.50
13.	3.34(1.05)	.28	.33	.48	.49	.20	.48	.29	.52	.21	.33	.25	.30
14.	4.65(.72)	.43	.20	.31	.39	.12	.36	.17	.40	.15	.53	.09*	.12	.31
15.	3.97(.98)	.36	.42	.68	.58	.17	.54	.27	.64	.29	.43	.17	.22	.57	.44
16.	4.55(.83)	.49	.19	.26	.29	.19	.38	.15	.29	.13	.32	.17	.14	.24	.37	.31
17.	3.74(.92)	.22	.49	.31	.35	.17	.35	.27	.31	.36	.22	.21	.25	.30	.19	.36	.30
18.	3.55(1.12)	.31	.32	.47	.40	.24	.44	.30	.50	.20	.33	.22	.27	.70	.30	.58	.29	.38
19.	4.07(.93)	.16	.52	.34	.26	.17	.26	.30	.28	.50	.21	.18	.27	.21	.17	.36	.16	.44	.26
20.	3.90(1.00)	.39	.30	.49	.67	.18	.61	.29	.66	.19	.42	.19	.19	.48	.39	.58	.36	.31	.47	.29

1. = When something good happens to me, I have people who I like to share the good news with; 2. = I finish whatever I begin; 3. = I am optimistic about my future; 4. = I feel happy; 5. = When I do an activity, I enjoy it so much that I lose track of time; 6. = I have a lot of fun; 7. = I get completely absorbed in what I am doing; 8. = I love life; 9. = I keep at my schoolwork until I am done with it; 10. = When I have a problem, I have someone who will be there for me; 11. = I get so involved in activities that I forget about everything else; 12. = When I am learning something new, I lose track of how much time has passed; 13. = In uncertain times, I expect the best; 14. = There are people in my life who really care about me; 15. = I think good things are going to happen to me; 16. = I have friends that I really care about; 17. = Once I make a plan to get something done, I stick to it; 18. = I believe that things will work out, no matter how difficult they seem; 19. = I am a hard worker; 20. = I am a cheerful person.

* < .05. All other correlations are significant at p < .01.

All intercorrelations were significant at p < .001 (see Table 4). As expected, the total EPOCH score had a strong positive correlation with the CSE (r = 0.68) and a strong negative correlation with the DASS-21 (r = -0.56). The engagement subscale showed only a moderate positive correlation with the CSE (r = 0.32) and a weak negative correlation with the DASS-21 (r = -0.17). Engagement showed weak negative correlations with all the DASS-21 subscales and weak positive correlations with all the CSE subscales (see Table 2). Perseverance had a moderate positive correlation with the CSE (r = 0.46) and a moderate negative correlation with the DASS-21 (r = -0.34), as expected. Perseverance likewise showed weak to moderate negative correlations with the DASS-21 subscales, and moderate to strong positive correlations with the CSE subscales. Optimism, by contrast, had a strong positive correlation with the CSE (r = 0.66) and a strong negative correlation with the DASS-21 (r = -0.56). Optimism also had moderate to strong negative correlations with the DASS-21 subscales, and a strong positive correlation with the CSE subscales. Connectedness showed moderate positive correlations with CSE (r = 0.48) and a moderate negative correlation with the DASS-21 (r = -0.38). Connectedness was also negatively correlated with the DASS-21 subscales and had moderate positive correlations with the CSE subscales, as expected. Happiness showed strong correlations with the CSE (r = 0.61) and DASS-21 (r = -0.63). Moreover, the correlations with the subscales of the DASS-21 were moderate to strong and negative; those with the CSE subscales were moderate to strong and positive.

Table 4

Intercorrelations between the EPOCH subscales with well-being indicators (N = 842–846).

	EPOCH
	Engagement	Perseverance	Optimism	Connectedness	Happiness	Overall
Well-being indicators:
Coping self-efficacy, total	0.32	0.46	0.66	0.48	0.61	0.68
Problem-focused c.	0.28	0.51	0.55	0.33	0.43	0.56
Emotion-focused c.	0.25	0.28	0.54	0.28	0.51	0.50
Social support	0.24	0.30	0.53	0.63	0.58	0.61
DASS total	-0.17	-0.34	-0.56	-0.38	-0.63	-0.56
Depression	-0.22	-0.41	-0.59	-0.50	-0.70	-0.65
Anxiety	-0.10	-0.22	-0.41	-0.26	-0.44	-0.39
Stress	-0.12	-0.23	-0.44	-0.22	-0.46	-0.40

Coping SE = Coping self-efficacy; Problem-focused c = using problem-focused coping; Emotion-focused c = stopping unpleasant emotions and thoughts; Social support = getting support from friends and family.

Note: All the correlations with the W-B indicators were significant at p < .001, not indicated here.

Discussion and conclusion

In this study, we investigated the psychometric properties of the Swedish version of the EPOCH Measure of Adolescent Well-being among Swedish high school students (aged 16–21 years). Our results supported the scale’s reliability and validity.

The five-factor solution of the origin EPOCH was supported by a CFA, having good overall model fit. In line with previous findings, the five-factor model fit the data better than did the one-factor model or the second-order factor model, even though the latter model also showed acceptable fit. Therefore, both the overall scale scores and subscale scores were used in analyses. Our results showed that the Swedish EPOCH has good internal consistency for overall scores and all five subscales. The particular strengths of the EPOCH are that 1) it is multidimensional in that it takes into account five different dimensions of well-being (engagement, perseverance, optimism, connectedness and happiness) rather than just one, and 2) it accommodates both the hedonic and eudaimonic views of well-being. Considering multiple dimensions of well-being [2,5] as well as both the hedonic and eudaimonic aspects [4-6] gives a more comprehensive picture of one’s well-being compared to focusing on just one indicator or aspect.

The subscales also showed good criterion validity. As expected, all five EPOCH subscales had a moderate positive correlation with coping self-efficacy, with engagement showing the lowest correlation, and optimism and happiness showing the highest correlations. Zeng and Kern [31] also found that engagement showed the lowest correlation with other well-being measures. This may be due to the fact that engagement is related to flow, which is a state of complete absorption in an activity without a sense of positive or negative affect [28]. Perhaps feeling engaged is not as strongly related to affectivity as the other EPOCH subscales. Coping self-efficacy is the belief that one is capable of coping in a challenging circumstance [32]; given that it is a positive self-efficacy belief, it makes sense that optimism and happiness had the strongest relations with it. It is also noteworthy that, as expected, the EPOCH connectedness subscale had a strong correlation with the CSE support from friends and family subscale.

Also as expected, negative correlations were found between the five EPOCH subscales and three indicators of mental health problems, namely depression, anxiety and stress. The weakest negative correlations were found with engagement, and the strongest negative correlations were found with happiness. This, again, replicated the results of Zeng and Kern [31].

The results also indicated scalar invariance across genders, indicating that the five-factor model is sufficient and that mean values are directly comparable across the genders, replicating the findings of Zeng and Kern [31].

Limitations

The criterion validity was limited because we used only one other positive well-being indicator (coping self-efficacy). We might have included more positive well-being indicators in this study to obtain greater nuance with the criterion validity. However, since the data were based on a larger project, only a couple of comparison indicators were collected. Another limitation with the criterion validity is that all scales were measured in the same way using self-report. Therefore, there is a chance that the criterion validity correlations are inflated due to common method variance [42]. Future validation studies should attempt to use non-self-report measures as well. Furthermore, the data was collected during a global pandemic, which might have affected how students responded to the well-being items. However, we did not consider this a problem for the scale validation, since the levels of well-being at the time of measurement should not interfere with assessment of scale validity or reliability.

Conclusion

This study indicated that the Swedish version of the EPOCH Measure of Adolescent Well-being had good psychometric properties: the CFA indicated good model fit and the scale (both the total scale and the subscales) had high internal consistency and criterion validity. Therefore, this study provides support for the use of this multidimensional measurement of positive psychological functioning among Swedish-speaking adolescent samples.

The Swedish version of the EPOCH measure of adolescent well-being.

(DOCX)

Click here for additional data file.

19 May 2021

PONE-D-21-08819

Testing the psychometric properties of the Swedish version of the EPOCH measure of adolescent well-being

PLOS ONE

Dear Dr. Maurer,

Thank you for submitting your manuscript to PLOS ONE. After careful consideration, we feel that it has merit but does not fully meet PLOS ONE’s publication criteria as it currently stands. Therefore, we invite you to submit a revised version of the manuscript if you are able to address the points raised during the review process. A part of  the solution might be finding terms better describing what was measured.

Please submit your revised manuscript by Jul 03 2021 11:59PM. If you will need more time than this to complete your revisions, please reply to this message or contact the journal office at plosone@plos.org. When you're ready to submit your revision, log on to https://www.editorialmanager.com/pone/ and select the 'Submissions Needing Revision' folder to locate your manuscript file.

Please include the following items when submitting your revised manuscript:

A rebuttal letter that responds to each point raised by the academic editor and reviewer(s). You should upload this letter as a separate file labeled 'Response to Reviewers'.

A marked-up copy of your manuscript that highlights changes made to the original version. You should upload this as a separate file labeled 'Revised Manuscript with Track Changes'.

An unmarked version of your revised paper without tracked changes. You should upload this as a separate file labeled 'Manuscript'.

If you would like to make changes to your financial disclosure, please include your updated statement in your cover letter. Guidelines for resubmitting your figure files are available below the reviewer comments at the end of this letter.

If applicable, we recommend that you deposit your laboratory protocols in protocols.io to enhance the reproducibility of your results. Protocols.io assigns your protocol its own identifier (DOI) so that it can be cited independently in the future. For instructions see: http://journals.plos.org/plosone/s/submission-guidelines#loc-laboratory-protocols. Additionally, PLOS ONE offers an option for publishing peer-reviewed Lab Protocol articles, which describe protocols hosted on protocols.io. Read more information on sharing protocols at https://plos.org/protocols?utm_medium=editorial-email&utm_source=authorletters&utm_campaign=protocols.

We look forward to receiving your revised manuscript.

Kind regards,

Frantisek Sudzina

Academic Editor

PLOS ONE

Journal Requirements:

When submitting your revision, we need you to address these additional requirements.

1. Please ensure that your manuscript meets PLOS ONE's style requirements, including those for file naming. The PLOS ONE style templates can be found at

https://journals.plos.org/plosone/s/file?id=wjVg/PLOSOne_formatting_sample_main_body.pdf and

https://journals.plos.org/plosone/s/file?id=ba62/PLOSOne_formatting_sample_title_authors_affiliations.pdf

2. Please specify whether you obtained consent from parents or guardians of the minors who may have participated in this study.

3. Please note that there are several instances where the McDonald’s values are not formatted correctly and therefore appear as a blank box. In your revision, please ensure that these are replaced with the correct symbol.

Thank you for your attention. We look forward to hearing from you.

4. We note that you have indicated that data from this study are available upon request. PLOS only allows data to be available upon request if there are legal or ethical restrictions on sharing data publicly. For more information on unacceptable data access restrictions, please see http://journals.plos.org/plosone/s/data-availability#loc-unacceptable-data-access-restrictions.

In your revised cover letter, please address the following prompts:

a) If there are ethical or legal restrictions on sharing a de-identified data set, please explain them in detail (e.g., data contain potentially sensitive information, data are owned by a third-party organization, etc.) and who has imposed them (e.g., an ethics committee). Please also provide contact information for a data access committee, ethics committee, or other institutional body to which data requests may be sent.

b) If there are no restrictions, please upload the minimal anonymized data set necessary to replicate your study findings as either Supporting Information files or to a stable, public repository and provide us with the relevant URLs, DOIs, or accession numbers. For a list of acceptable repositories, please see http://journals.plos.org/plosone/s/data-availability#loc-recommended-repositories.

We will update your Data Availability statement on your behalf to reflect the information you provide.

[Note: HTML markup is below. Please do not edit.]

Reviewers' comments:

Reviewer's Responses to Questions

Comments to the Author

1. Is the manuscript technically sound, and do the data support the conclusions?

The manuscript must describe a technically sound piece of scientific research with data that supports the conclusions. Experiments must have been conducted rigorously, with appropriate controls, replication, and sample sizes. The conclusions must be drawn appropriately based on the data presented.

Reviewer #1: Partly

Reviewer #2: No

**********

2. Has the statistical analysis been performed appropriately and rigorously?

Reviewer #1: Yes

Reviewer #2: No

**********

3. Have the authors made all data underlying the findings in their manuscript fully available?

The PLOS Data policy requires authors to make all data underlying the findings described in their manuscript fully available without restriction, with rare exception (please refer to the Data Availability Statement in the manuscript PDF file). The data should be provided as part of the manuscript or its supporting information, or deposited to a public repository. For example, in addition to summary statistics, the data points behind means, medians and variance measures should be available. If there are restrictions on publicly sharing data—e.g. participant privacy or use of data from a third party—those must be specified.

Reviewer #1: No

Reviewer #2: No

**********

4. Is the manuscript presented in an intelligible fashion and written in standard English?

PLOS ONE does not copyedit accepted manuscripts, so the language in submitted articles must be clear, correct, and unambiguous. Any typographical or grammatical errors should be corrected at revision, so please note any specific errors here.

Reviewer #1: Yes

Reviewer #2: Yes

**********

5. Review Comments to the Author

Please use the space provided to explain your answers to the questions above. You may also include additional comments for the author, including concerns about dual publication, research ethics, or publication ethics. (Please upload your review as an attachment if it exceeds 20,000 characters)

Reviewer #1: I reviewed the manuscript entitled "Testing the psychometric properties of the Swedish version of the EPOCH measure of adolescent well-being" for PLOS ONE (PONE-D-21-08819). I read the manuscript word-for-word and have provided comments and suggestions that I hope will improve the manuscript. I am a clinical psychology post-doc at Kent State University with expertise in well-being measurement. My name is David Disabato and I can be contacted by the authors with any questions about this review (ddisab01@gmail.com).

Introduction:

1) The authors refer to the reliability of factors at various points in the introduction section. Reliability refers to observed scores and not latent factors; latent factors have (theoretically) perfect reliability. For example, the authors say on page 5 “They found that the five-factor model was reliable even across different time points, and each factor had high internal consistency”. Instead it should say something like, “The five EPOCH subscale scores were found to have high internal consistency”.

2) I am not sure what the authors mean when they say “… but its consistency over time was low with a range from r = .12 to r = .21.” Consistency can mean different things in psychometrics: How is that term being used here? Are those correlations test-retest coefficients? If yes, how much time had passed between repeat assessments?

3) I thought it was strange that the authors were referring to specific criterion questionnaires in the introduction section. Personally, I think it makes more sense to talk about constructs (e.g., adverse mental health) in the introduction section, and wait to talk about specific measures until the measures section (e.g., DASS-21). Obviously, for a psychometric paper you do talk about the measure of psychometric interest in the introduction. But I think the criterion measures should wait until the methods section.

Methods:

4) I would like to see more information about study recruitment and administration. How were the 852 students from these 3 schools recruited? Was the study encouraged by the school administration? What percentage of the total student body at each school chose to participate (I assume not everyone volunteered)? It says the surveys were distributed via email and completed online. How many students from each school were emailed the online survey – all of them or only a subset? Were they administered at school or at home? How long did it take participants on average to complete the full survey?

5) In the EPOCH part of the measures subsection, the authors should include a sentence telling the reader that the internal consistency of the EPOCH in this sample will be reported in the results section. As the reader, I am used to seeing the internal consistency reported in the measures section and I was confused by it not being there.

6) I would recommend placing the procedures subsection first in the methods section as it lays the foundation for understanding all other aspects of the study methodology. As the reader, I was left confused after the participants section wondering where the information about the procedure was going to be reported.

7) The type II error rate for the chi-square test of model fit is a problem in small samples, not large samples. I think the authors mean to say that the chi-square statistic tends to reject even great fitting models with negligible misfit in large samples.

8) The authors should provide a definition of McDonald’s omega for the reader and how it differs from Cronbach’s alpha as some readers may not be familiar with that index of internal consistency.

Results:

9) The Little MCAR test cannot determine if the data were truly missing completely at random. It can only tell you if the *measures in your dataset* are related to missingness, but that says nothing about the infinite number of other variables not measured in your study. You can certainly report the results of Little’s MCAR test, but I think the authors should replace the statement that says “Thus, the data were missing completely at random” with something like “Thus, the missing values were not related to any variables in our study”.

10) Since there is almost no missing data (<.01%), preliminary analyses about missing data don’t seem very important. Instead, I would like to see correlations, means, and standard deviations of the subscale scores as a Table 1 (as currently a Table 1 does not exist in this paper for some reason). I noticed the authors report the means and standard deviations in Table 2, so I would then move them to this new Table 1. I think it would also be worth including a supplemental table of the EPOCH items correlations, means, and standard deviations as those could be used by other researchers to replicate and extend the CFA analyses the authors have done; however, I will leave that up to the authors.

11) How the authors report the chi-square difference tests of model comparisons is confusing to me. The chi-square difference values appear to look like the alternative model’s chi-square statistic, but then slowly realized that those were the chi-square difference values. I think it would be easier for the reader if the authors first presented the chi-square statistics (and model fit indices) for each model and then follow-up that up with separate sentences about model comparison and the chi-square difference tests.

12) It is not clear what the coefficients are in Figure 1. I assume they are standardized factor loadings and latent correlations, but that needs to be explicitly stated in the caption.

13) If the authors are going to use the EPOCH total score, then they really should be discussing the higher-order factor model as an alternative, acceptably fitting factor model, as that is the factor model behind the total score. Right now, the results and discussion are written as if the higher-order factor model was rejected, which then implies a total score would not be used – only subscale scores. However, that is not what the authors do when moving to criterion validity. Therefore, the authors either need to talk about the higher-order factor model as an alternative, acceptable factor model or remove the analyses with the EPOCH total score. (It does makes sense to me to have an EPOCH total score given the internal consistency of the total score was very high.)

14) The authors don’t need to repeat the Cronbach’s alphas and McDonald’s omegas in Table 2 if they already have them in the text. I would put them in one or the other location, but no need for both.

15) There are some typos in the notes for Table 2.

Discussion:

16) Again, there is psychometric language about factors that should be about subscale scores. For example, on page 13 “The factors also showed good criterion validity”. The criterion validity analyses were done with the EPOCH observed subscale scores, not the latent factors in a full SEM, so the language in the discussion should refer to subscale scores, not latent factors.

17) I am not sure why the authors refer to optimism as a “positive emotion-based subscales” in the discussion section. Optimism is a cognitive construct by almost every definition I have read and the items in the EPOCH all refer to cognitive beliefs and not emotions. They need to either clarify what they mean or remove this statement.

18) The authors should remove the reference to “mood disorders” when talking about the DASS-21 in the discussion on page 14. Both depression and anxiety are sometimes referred to as involving problems with affect or mood, but the term “mood disorder” is reserved to depression and not anxiety in the mental health literature.

19) I find it strange that the authors are interested in showing no gender differences across the EPOCH subscales. Why is that important to the psychometrics? It would be important to show measurement invariance across gender in the factor models, but a gender difference for the subscale scores doesn’t say anything about the psychometrics of the questionnaire. Especially, since there is evidence for gender differences in well-being across measurement methods (e.g., women tend to be higher on both negative emotions and positive emotions; Nolen-Hoeksema & Rusting, 1999).

Nolen-Hoeksema, S., & Rusting, C. L. (1999). Gender Differences in Well-Being. Well-being: Foundations of hedonic psychology, Chapter 17.

20) The same argument goes for age. If the authors are concerned about the EPOCH being psychometrically biased due to gender or age, then they should test for measurement invariance. For example, Kern et al. (2016) tested for measurement invariance across gender. I noticed the authors cite Tim Brown’s CFA book – chapter 7 goes over measurement invariance across demographic groups if they are interested in doing this. Unless they authors test for measurement invariance across age, then they cannot state that “they are invariant with age among adolescents and changes in scores reflect changes in psychological functioning rather than changes in maturation” (pg. 14).

21) I think the authors need to point that that a limitation of the criterion validity tests is that all criterion were measured with the same method as the EPOCH measure (i.e., self-report). Therefore, there is a good chance that the criterion validity correlations are over-estimates of validity and inflated due to shared method variance. Future research should include non-self-report criterion.

Reviewer #2: Thank you for the opportunity to review the manuscript, “Testing the psychometric properties of the Swedish version of the EPOCH measure of adolescent well-being.” The purpose of this manuscript was to translate the EPOCH measure into Swedish. Participants were 846 Swedish adolescents who completed a self-report questionnaire in May 2020.

Strengths: I appreciate research efforts to translate English measures into other languages to increase access to different populations. Psychological science certainly has an issue of recruiting diverse samples, including studies that involve languages other than English. The sample size was sufficiently large and analyses are straightforward.

Nonetheless, this manuscript has noteworthy limitations. These are detailed below.

1) Several broad sweeping assumptions are made in the introduction. The authors describe “hedonia” and “eudaimonia” as two distinct forms of well-being. However, there are now several studies from different research groups showing that these two types of well-being—at least as we currently measure them—overlap quite a high degree, with some people suggesting they represent the same type of well-being. At a minimum, the authors should acknowledge that there is considerable debate about whether or not these represent two distinct forms of well-being and present relevant factor analytical work.

2) Hedonic well-being is described as “emotional well-being” and eudaimonic well-being is described as “well-being resources not tied to emotions.” There are three issues with this conceptualization (in addition to the issue noted above). First, the widely adopted model of hedonic well-being is Diener’s subjective well-being model (SWB), which is composed of the presence of positive emotions, lack of negative emotions, and life satisfaction. Diener and others have amassed a corpus of work demonstrating the utility of this model. (And in the introduction of the present study, the authors use Diener’s model to describe hedonia.) Yet, negative emotions and life satisfaction are surprisingly absent from the EPOCH measurement model. Second, it is incorrect to claim that eudaimonic is “not tied to emotions.” For example, Laura King has shown that meaning in life, arguably a core component of “eudaimonic well-being”—although also surprisingly absent from the EPOCH measure—is correlated with (and often influenced by) positive affect. Can a “type” of well-being really be devoid of or distinct from emotions? Third, the lack of a strong, clear, and cohesive definition of eudaimonic well-being opens up the possibility for any “positive” or desirable psychological variable — as judged, often arbitrarily, by the researcher — to be called “well-being.” For example, why was environmental mastery from Ryff’s widely adopted psychological well-being (PWB) model excluded from the EPOCH model presented here? A 2016 review identified 99 self-report measures of “well-being” containing nearly 200 different dimensions. Without concise definitions, theoretically informed models, and some consistency in measurement, the term “well-being” runs the risk of becoming meaningless.

3) A related but separate point: I am struggling to see how optimism and perseverance are components of well-being. Decades of research would suggest that these are personality traits.

4) P. 4: “However, Kern et al. argued that measuring well-being in adolescents should be based on well-being categories that are less abstract than those in the PERMA profiler (14), which was more suitable for adults.” Can the authors clarify what is meant by “abstract” and elaborate on which constructs were replaced and why? I do not entirely follow the logic.

5) The EPOCH measure is based on the PERMA collection of variables, for which there is limited empirical support. To date, few studies have examined the factor structure of PERMA, and some have found that it highly overlaps with other, more well-established models. Stronger rationale is needed for selection of this variable set.

6) The five components of the EPOCH measure are described as “flourishing” in the results section but “well-being” elsewhere in the manuscript. Please use consistent terminology to prevent confusion.

7) The race/ethnicity breakdown of participants is missing from the demographics section.

8) P. 15: “We might have included more positive well-being indicators in this study to obtain greater nuance with the criterion validity. However, since the data were based on a larger project, only a couple of comparison indicators were collected.” I appreciate the authors being forthcoming in this limitation. However, it does raise significant questions about the validity of this measure. It seems that at a minimum, an assessment of criterion validity should include a comparison with another measure of well-being.

9) I do not see Table 1, only Table 2. Based on the text, it seems that Table 1 would include the bivariate correlations?

**********

6. PLOS authors have the option to publish the peer review history of their article (what does this mean?). If published, this will include your full peer review and any attached files.

If you choose “no”, your identity will remain anonymous but your review may still be made public.

Do you want your identity to be public for this peer review? For information about this choice, including consent withdrawal, please see our Privacy Policy.

Reviewer #1: Yes: David Disabato

Reviewer #2: No

[NOTE: If reviewer comments were submitted as an attachment file, they will be attached to this email and accessible via the submission site. Please log into your account, locate the manuscript record, and check for the action link "View Attachments". If this link does not appear, there are no attachment files.]

While revising your submission, please upload your figure files to the Preflight Analysis and Conversion Engine (PACE) digital diagnostic tool, https://pacev2.apexcovantage.com/. PACE helps ensure that figures meet PLOS requirements. To use PACE, you must first register as a user. Registration is free. Then, login and navigate to the UPLOAD tab, where you will find detailed instructions on how to use the tool. If you encounter any issues or have any questions when using PACE, please email PLOS at figures@plos.org. Please note that Supporting Information files do not need this step.

19 Aug 2021

Journal Requirements:

When submitting your revision, we need you to address these additional requirements.

1. Please ensure that your manuscript meets PLOS ONE's style requirements, including those for file naming. The PLOS ONE style templates can be found at

https://journals.plos.org/plosone/s/file?id=wjVg/PLOSOne_formatting_sample_main_body.pdf and

https://journals.plos.org/plosone/s/file?id=ba62/PLOSOne_formatting_sample_title_authors_affiliations.pdf

Answer 1: We have followed the guidelines now.

2. Please specify whether you obtained consent from parents or guardians of the minors who may have participated in this study.

Answer 2: All the participants were above the age of 15, meaning that they were able to give their own informed consent without needing parental informed consent under Swedish ethical guidelines. This has now been added into the manuscript.

3. Please note that there are several instances where the McDonald’s values are not formatted correctly and therefore appear as a blank box. In your revision, please ensure that these are replaced with the correct symbol.

Answer 3: We have made sure that the symbols are in place in the manuscript.

Thank you for your attention. We look forward to hearing from you.

4. We note that you have indicated that data from this study are available upon request. PLOS only allows data to be available upon request if there are legal or ethical restrictions on sharing data publicly. For more information on unacceptable data access restrictions, please see http://journals.plos.org/plosone/s/data-availability#loc-unacceptable-data-access-restrictions.

Answer 4: The ethics committee from which we obtained ethical approval indicated that data should not be shared with outside parties.

In your revised cover letter, please address the following prompts:

a) If there are ethical or legal restrictions on sharing a de-identified data set, please explain them in detail (e.g., data contain potentially sensitive information, data are owned by a third-party organization, etc.) and who has imposed them (e.g., an ethics committee). Please also provide contact information for a data access committee, ethics committee, or other institutional body to which data requests may be sent.

Answer 5: We did not obtain ethical approval for sharing data. In our ethical approval, we were told that it is important that we do not share our data with outside parties. However, we want to be transparent and share data with individuals who are interested in it, so we opted to share only upon request. We hope this is acceptable.

b) If there are no restrictions, please upload the minimal anonymized data set necessary to replicate your study findings as either Supporting Information files or to a stable, public repository and provide us with the relevant URLs, DOIs, or accession numbers. For a list of acceptable repositories, please see http://journals.plos.org/plosone/s/data-availability#loc-recommended-repositories.

We will update your Data Availability statement on your behalf to reflect the information you provide.

5. Review Comments to the Author

Please use the space provided to explain your answers to the questions above. You may also include additional comments for the author, including concerns about dual publication, research ethics, or publication ethics. (Please upload your review as an attachment if it exceeds 20,000 characters)

Reviewer #1: I reviewed the manuscript entitled "Testing the psychometric properties of the Swedish version of the EPOCH measure of adolescent well-being" for PLOS ONE (PONE-D-21-08819). I read the manuscript word-for-word and have provided comments and suggestions that I hope will improve the manuscript. I am a clinical psychology post-doc at Kent State University with expertise in well-being measurement. My name is David Disabato and I can be contacted by the authors with any questions about this review (ddisab01@gmail.com).

Introduction:

1) The authors refer to the reliability of factors at various points in the introduction section. Reliability refers to observed scores and not latent factors; latent factors have (theoretically) perfect reliability. For example, the authors say on page 5 “They found that the five-factor model was reliable even across different time points, and each factor had high internal consistency”. Instead it should say something like, “The five EPOCH subscale scores were found to have high internal consistency”.

Answer 6: Thank you for this important observation! This issue has been corrected in the manuscript.

2) I am not sure what the authors mean when they say “… but its consistency over time was low with a range from r = .12 to r = .21.” Consistency can mean different things in psychometrics: How is that term being used here? Are those correlations test-retest coefficients? If yes, how much time had passed between repeat assessments?

Answer 7: The time between repeat assessments was actually not reported in the original article. These correlations are test-retest coefficients.

3) I thought it was strange that the authors were referring to specific criterion questionnaires in the introduction section. Personally, I think it makes more sense to talk about constructs (e.g., adverse mental health) in the introduction section, and wait to talk about specific measures until the measures section (e.g., DASS-21). Obviously, for a psychometric paper you do talk about the measure of psychometric interest in the introduction. But I think the criterion measures should wait until the methods section.

Answer 8: Thank you for this observation. We have removed the lengthy description of the criterion questionnaires from the introduction section.

Methods:

4) I would like to see more information about study recruitment and administration. How were the 852 students from these 3 schools recruited? Was the study encouraged by the school administration? What percentage of the total student body at each school chose to participate (I assume not everyone volunteered)? It says the surveys were distributed via email and completed online. How many students from each school were emailed the online survey – all of them or only a subset? Were they administered at school or at home? How long did it take participants on average to complete the full survey?

Answer 9: The study link was published on each school’s website, which is accessible to all students. However, we do not know how many of the students read the information about the study. Many students from the school did not participate, which might be due to their not taking notice of the study on the website. We have some statistics on how many of the students clicked the link we posted on the website and how many of them completed the questionnaire. The total number of students who clicked the link was 1212, while only 852 (about 70%) completed the questionnaire. The study was not explicitly encouraged by the school administration, but they did agree to participate. Since the pandemic was ongoing and schools were closed, the questionnaire battery was administered at home. The participation took approximately 15 minutes. Some of these details have now been added to the manuscript for clarity.

5) In the EPOCH part of the measures subsection, the authors should include a sentence telling the reader that the internal consistency of the EPOCH in this sample will be reported in the results section. As the reader, I am used to seeing the internal consistency reported in the measures section and I was confused by it not being there.

Answer 10: A clarification has now been added into the measures section.

6) I would recommend placing the procedures subsection first in the methods section as it lays the foundation for understanding all other aspects of the study methodology. As the reader, I was left confused after the participants section wondering where the information about the procedure was going to be reported.

Answer 11: Thank you for this remark. The placement of the Procedure section has been changed.

7) The type II error rate for the chi-square test of model fit is a problem in small samples, not large samples. I think the authors mean to say that the chi-square statistic tends to reject even great fitting models with negligible misfit in large samples.

Answer 12: Thank you for this remark. This has been corrected.

8) The authors should provide a definition of McDonald’s omega for the reader and how it differs from Cronbach’s alpha as some readers may not be familiar with that index of internal consistency.

Answer 13: This is an important point, thank you. A clarification of the indexes of internal consistency has been added.

Results:

9) The Little MCAR test cannot determine if the data were truly missing completely at random. It can only tell you if the *measures in your dataset* are related to missingness, but that says nothing about the infinite number of other variables not measured in your study. You can certainly report the results of Little’s MCAR test, but I think the authors should replace the statement that says “Thus, the data were missing completely at random” with something like “Thus, the missing values were not related to any variables in our study”.

Answer 14: Thank you for this important point! This correction has been made in the manuscript.

10) Since there is almost no missing data (<.01%), preliminary analyses about missing data don’t seem very important. Instead, I would like to see correlations, means, and standard deviations of the subscale scores as a Table 1 (as currently a Table 1 does not exist in this paper for some reason). I noticed the authors report the means and standard deviations in Table 2, so I would then move them to this new Table 1. I think it would also be worth including a supplemental table of the EPOCH items correlations, means, and standard deviations as those could be used by other researchers to replicate and extend the CFA analyses the authors have done; however, I will leave that up to the authors.

Answer 15: We have updated our tables and added a new one reporting item intercorrelations and descriptives, as suggested.

11) How the authors report the chi-square difference tests of model comparisons is confusing to me. The chi-square difference values appear to look like the alternative model’s chi-square statistic, but then slowly realized that those were the chi-square difference values. I think it would be easier for the reader if the authors first presented the chi-square statistics (and model fit indices) for each model and then follow-up that up with separate sentences about model comparison and the chi-square difference tests.

Answer 16: Thank you for this comment; this has now been done.

12) It is not clear what the coefficients are in Figure 1. I assume they are standardized factor loadings and latent correlations, but that needs to be explicitly stated in the caption.

Answer 17: These details have now been added to the caption.

13) If the authors are going to use the EPOCH total score, then they really should be discussing the higher-order factor model as an alternative, acceptably fitting factor model, as that is the factor model behind the total score. Right now, the results and discussion are written as if the higher-order factor model was rejected, which then implies a total score would not be used – only subscale scores. However, that is not what the authors do when moving to criterion validity. Therefore, the authors either need to talk about the higher-order factor model as an alternative, acceptable factor model or remove the analyses with the EPOCH total score. (It does makes sense to me to have an EPOCH total score given the internal consistency of the total score was very high.)

Answer 18: Thank you for the comment. Even though the higher-order model had somewhat worse fit compared to the five-factor model, the fit indices were in the acceptable range. We added in the text that the higher-order model was not rejected.

14) The authors don’t need to repeat the Cronbach’s alphas and McDonald’s omegas in Table 2 if they already have them in the text. I would put them in one or the other location, but no need for both.

Answer 19: These have been removed from the table.

15) There are some typos in the notes for Table 2.

Answer 20: We have had a professional editor check through the manuscript.

Discussion:

16) Again, there is psychometric language about factors that should be about subscale scores. For example, on page 13 “The factors also showed good criterion validity”. The criterion validity analyses were done with the EPOCH observed subscale scores, not the latent factors in a full SEM, so the language in the discussion should refer to subscale scores, not latent factors.

Answer 21: This has now been corrected.

17) I am not sure why the authors refer to optimism as a “positive emotion-based subscales” in the discussion section. Optimism is a cognitive construct by almost every definition I have read and the items in the EPOCH all refer to cognitive beliefs and not emotions. They need to either clarify what they mean or remove this statement.

Answer 22: This is a very important observation. We have omitted the statement.

18) The authors should remove the reference to “mood disorders” when talking about the DASS-21 in the discussion on page 14. Both depression and anxiety are sometimes referred to as involving problems with affect or mood, but the term “mood disorder” is reserved to depression and not anxiety in the mental health literature.

Answer 23: This is a very valid point; accordingly, we have omitted this from the manuscript.

19) I find it strange that the authors are interested in showing no gender differences across the EPOCH subscales. Why is that important to the psychometrics? It would be important to show measurement invariance across gender in the factor models, but a gender difference for the subscale scores doesn’t say anything about the psychometrics of the questionnaire. Especially, since there is evidence for gender differences in well-being across measurement methods (e.g., women tend to be higher on both negative emotions and positive emotions; Nolen-Hoeksema & Rusting, 1999).

Nolen-Hoeksema, S., & Rusting, C. L. (1999). Gender Differences in Well-Being. Well-being: Foundations of hedonic psychology, Chapter 17.

Answer 24: Thank you for this important comment. We tested the measurement invariance across genders focusing on the change in CFI (ΔCFI) and RMSEA (ΔRMSEA) when a more constrained model was compared to a less constrained one. We considered ΔCFI ≤ 0.010 and ΔRMSEA ≤ 0.015 (Cheung & Rensvold, 2002; Chen, 2007) as indications that the invariance assumption holds. The results are now presented in the manuscript.

20) The same argument goes for age. If the authors are concerned about the EPOCH being psychometrically biased due to gender or age, then they should test for measurement invariance. For example, Kern et al. (2016) tested for measurement invariance across gender. I noticed the authors cite Tim Brown’s CFA book – chapter 7 goes over measurement invariance across demographic groups if they are interested in doing this. Unless they authors test for measurement invariance across age, then they cannot state that “they are invariant with age among adolescents and changes in scores reflect changes in psychological functioning rather than changes in maturation” (pg. 14).

Answer 25: We have omitted a comparison of age since our sample is rather homogenous regarding age.

21) I think the authors need to point that that a limitation of the criterion validity tests is that all criterion were measured with the same method as the EPOCH measure (i.e., self-report). Therefore, there is a good chance that the criterion validity correlations are over-estimates of validity and inflated due to shared method variance. Future research should include non-self-report criterion.

Answer 26: Thank you for pointing out this important limitation, which has been added to the discussion.

Reviewer #2: Thank you for the opportunity to review the manuscript, “Testing the psychometric properties of the Swedish version of the EPOCH measure of adolescent well-being.” The purpose of this manuscript was to translate the EPOCH measure into Swedish. Participants were 846 Swedish adolescents who completed a self-report questionnaire in May 2020.

Strengths: I appreciate research efforts to translate English measures into other languages to increase access to different populations. Psychological science certainly has an issue of recruiting diverse samples, including studies that involve languages other than English. The sample size was sufficiently large and analyses are straightforward.

Answer 27: Thank you for pointing out the strengths of the manuscript!

Nonetheless, this manuscript has noteworthy limitations. These are detailed below.

1) Several broad sweeping assumptions are made in the introduction. The authors describe “hedonia” and “eudaimonia” as two distinct forms of well-being. However, there are now several studies from different research groups showing that these two types of well-being—at least as we currently measure them—overlap quite a high degree, with some people suggesting they represent the same type of well-being. At a minimum, the authors should acknowledge that there is considerable debate about whether or not these represent two distinct forms of well-being and present relevant factor analytical work.

Answer 28: Thank you for this very important observation. It is true that the distinction between hedonic and eudaimonic well-being is under debate, and this should have been brought forward clearly in the manuscript. We have addressed this point by adding more critical discussion of the distinction between these types of well-being and emphasizing the debate in the field.

2) Hedonic well-being is described as “emotional well-being” and eudaimonic well-being is described as “well-being resources not tied to emotions.” There are three issues with this conceptualization (in addition to the issue noted above). First, the widely adopted model of hedonic well-being is Diener’s subjective well-being model (SWB), which is composed of the presence of positive emotions, lack of negative emotions, and life satisfaction. Diener and others have amassed a corpus of work demonstrating the utility of this model. (And in the introduction of the present study, the authors use Diener’s model to describe hedonia.) Yet, negative emotions and life satisfaction are surprisingly absent from the EPOCH measurement model. Second, it is incorrect to claim that eudaimonic is “not tied to emotions.” For example, Laura King has shown that meaning in life, arguably a core component of “eudaimonic well-being”—although also surprisingly absent from the EPOCH measure—is correlated with (and often influenced by) positive affect. Can a “type” of well-being really be devoid of or distinct from emotions? Third, the lack of a strong, clear, and cohesive definition of eudaimonic well-being opens up the possibility for any “positive” or desirable psychological variable — as judged, often arbitrarily, by the researcher — to be called “well-being.” For example, why was environmental mastery from Ryff’s widely adopted psychological well-being (PWB) model excluded from the EPOCH model presented here? A 2016 review identified 99 self-report measures of “well-being” containing nearly 200 different dimensions. Without concise definitions, theoretically informed models, and some consistency in measurement, the term “well-being” runs the risk of becoming meaningless.

Answer 29: Thank you for these important arguments and pointing out the theoretical weaknesses of the manuscript. It is true that we made too broad a generalization concerning hedonic and eudaimonic well-being. Research has indeed shown that these types of well-being are highly interrelated, with some studies showing them to be in fact loading onto just a single overarching factor. We have thus added more nuance in the discussion of these two types of well-being in the introduction and underscored the debate (although still rather briefly). As you note, the EPOCH measure does not include many of the different aspects of other well-being models, such as environmental mastery, autonomy, personal growth, meaning etc. We believe this may be because Kern and colleagues considered those aspects to be too difficult for adolescents to grasp (this is just our guess, as it was not discussed by them) and because the EPOCH was meant to operationalize the PERMA model for adolescents by conceptually simplifying some of its subcomponents, such as meaning and achievement (which became optimism and perseverance, respectively). Our reason for using the EPOCH (which was not clearly stated in the manuscript before, but has been added now) was that the PERMA model is the theoretical framework underlying most positive educational studies. It is therefore important to be able to measure its facets, particularly in the field of positive education. Moreover, many positive educational interventions are based on the PERMA model, and evaluations of the interventions therefore require a measure that operationalizes PERMA for adolescents. This rationale has been more clearly articulated in the manuscript.

3) A related but separate point: I am struggling to see how optimism and perseverance are components of well-being. Decades of research would suggest that these are personality traits.

Answer 30: This is an apt observation, and a very important discussion point considering the way in which well-being is defined. The EPOCH, although a measure of well-being, is meant to define positive characteristics that predict adolescents’ optimal functioning. As we understand it, Kern and colleagues wanted to create a measure in which not only attitudes, but also characteristics, would be reflected to measure optimal adolescent functioning. In addition, these aspects were meant to reflect the PERMA categories of Achievement (perseverance) and Meaning (optimism) as more action-oriented aspects of functioning. Kern et al. suggest that through perseverance, adolescents are able to reach achievements, while through optimism, they are able to engage with meaningful pursuits and a sense of meaning in life. We have now included more discussion of optimal functioning in the manuscript, and the potential practical reason for the use of personality traits in the well-being definition (to evaluate interventions).

4) P. 4: “However, Kern et al. argued that measuring well-being in adolescents should be based on well-being categories that are less abstract than those in the PERMA profiler (14), which was more suitable for adults.” Can the authors clarify what is meant by “abstract” and elaborate on which constructs were replaced and why? I do not entirely follow the logic.

Answer 31: Pointing out this unclear part of the introduction was important. We decided to delete this part of the manuscript discussing the ‘’abstract’’ PERMA categories, mainly because we could not much elaborate on it – Kern et al. do not elaborate on this point themselves. In its place, we discuss your previous point in more detail – that is, why well-being is operationalized with personality traits. Hopefully, this paragraph is now clearer and explains the EPOCH better.

5) The EPOCH measure is based on the PERMA collection of variables, for which there is limited empirical support. To date, few studies have examined the factor structure of PERMA, and some have found that it highly overlaps with other, more well-established models. Stronger rationale is needed for selection of this variable set.

Answer 32: Thank you for this important observation. We have now added some literature evaluating the psychometric properties of the PERMA profiler, as well as added a clearer rationale for why we chose the PERMA framework for our study. Our rationale for choosing the PERMA framework is two-fold: 1) PERMA is a multidimensional measure of well-being (which does not necessarily set it above or aside many other models, however), and 2) PERMA is the main theoretical framework behind positive education studies, and for theoretical as well as practical purposes (such as intervention evaluations) it is important to validate a measure of PERMA for adolescents.

6) The five components of the EPOCH measure are described as “flourishing” in the results section but “well-being” elsewhere in the manuscript. Please use consistent terminology to prevent confusion.

Answer 33: Thank you for pointing this out. We have corrected this confusion now.

7) The race/ethnicity breakdown of participants is missing from the demographics section.

Answer 34: Thank you for pointing this out. We acknowledge the lack of such information in the participant description. For ethical reasons, however, it is not customary in Sweden to ask participants for race/ethnicity information unless there is a very important reason to do so. We did not have such a reason, and therefore we did not include this question in the demographic questions of the study. This is a limitation, unfortunately. However, in Sweden, it is acceptable to ask which country the students are from and which country their parents are from. Based on this information, we identify a group of participants with a foreign background. This is defined as the participant either being born abroad with at least one parent born abroad as well, or being born in Sweden with both parents being born abroad. This information was included in the Methods.

8) P. 15: “We might have included more positive well-being indicators in this study to obtain greater nuance with the criterion validity. However, since the data were based on a larger project, only a couple of comparison indicators were collected.” I appreciate the authors being forthcoming in this limitation. However, it does raise significant questions about the validity of this measure. It seems that at a minimum, an assessment of criterion validity should include a comparison with another measure of well-being.

Answer 35: We acknowledge this serious limitation in our study. It is unfortunate that we did not include another measure of well-being. We have only one measure reflecting positive psychological functioning, the coping self-efficacy scale, which we hope is sufficient for criterion validity. However, we understand that this might raise questions of the EPOCH’s validity.

9) I do not see Table 1, only Table 2. Based on the text, it seems that Table 1 would include the bivariate correlations?

Answer 36: Now we have included Table 1, and two additional tables into the manuscript.

________________________________________

Submitted filename: Response to Reviewers.docx

Click here for additional data file.

15 Oct 2021

Testing the psychometric properties of the Swedish version of the EPOCH measure of adolescent well-being

PONE-D-21-08819R1

Dear Dr. Maurer,

We’re pleased to inform you that your manuscript has been judged scientifically suitable for publication and will be formally accepted for publication once it meets all outstanding technical requirements.

Within one week, you’ll receive an e-mail detailing the required amendments. When these have been addressed, you’ll receive a formal acceptance letter and your manuscript will be scheduled for publication.

An invoice for payment will follow shortly after the formal acceptance. To ensure an efficient process, please log into Editorial Manager at http://www.editorialmanager.com/pone/, click the 'Update My Information' link at the top of the page, and double check that your user information is up-to-date. If you have any billing related questions, please contact our Author Billing department directly at authorbilling@plos.org.

If your institution or institutions have a press office, please notify them about your upcoming paper to help maximize its impact. If they’ll be preparing press materials, please inform our press team as soon as possible -- no later than 48 hours after receiving the formal acceptance. Your manuscript will remain under strict press embargo until 2 pm Eastern Time on the date of publication. For more information, please contact onepress@plos.org.

Kind regards,

Frantisek Sudzina

Academic Editor

PLOS ONE

Additional Editor Comments (optional):

Reviewers' comments:

20 Oct 2021

PONE-D-21-08819R1

Testing the psychometric properties of the Swedish version of the EPOCH measure of adolescent well-being

Dear Dr. Maurer:

I'm pleased to inform you that your manuscript has been deemed suitable for publication in PLOS ONE. Congratulations! Your manuscript is now with our production department.

If your institution or institutions have a press office, please let them know about your upcoming paper now to help maximize its impact. If they'll be preparing press materials, please inform our press team within the next 48 hours. Your manuscript will remain under strict press embargo until 2 pm Eastern Time on the date of publication. For more information please contact onepress@plos.org.

If we can help with anything else, please email us at plosone@plos.org.

Thank you for submitting your work to PLOS ONE and supporting open access.

Kind regards,

PLOS ONE Editorial Office Staff

on behalf of

Dr. Frantisek Sudzina

Academic Editor

PLOS ONE