EPOCH questionnaire for the study of psychological well-being of Russian adolescents: psychometric characteristics and possibilities for use

Introduction

Within the system of criteria and assessments for the quality of contemporary education, psychological well-being (PWB) occupies one of the leading positions and, in a certain sense, establishes a new discourse for Russian educational policy (Polivanova, 2020; Potanina, Morosanova, 2022). PWB is also considered one of the key criteria for the socialization of contemporary children, adolescents, and youth. PWB is linked to productive types of activity motivation (Deci, Ryan, 2008; Dekhtyarenko, Savchenko, Shlyagina, 2023), academic achievement (Potanina, Morosanova, 2022; Kern et al., 2015), low behavioral risks (Adler, Seligman, 2016), and physical health (Dray et al., 2017). Research emphasizes the importance of interpersonal communication factors for understanding adolescent well-being (Rasskazova, Sadovnichaya, 2023). A high level of well-being and the development of its core components ensure the positive functioning of the personality and contribute to a person's self-realization at different stages of ontogenesis (Isaeva et al., 2022; Volkova et al., 2022; Rudnova et al., 2023). PWB, as an experience of positive modality, can be considered a mechanism for the internalization of social behavior, thereby opening possibilities for pedagogical interventions and the management of socialization processes (Volkova, Sorokoumova, 2024).

The problem of assessing PWB is one of the current issues in contemporary psychology. Within various theoretical approaches to understanding the concept of PWB, different assessment methods exist (e.g., Belinskaya, Shaehov, 2023; Rikel et al., 2017; Shilko et al., 2018). For assessing adolescent PWB, in our view, the theoretical approach of positive psychology is appropriate. In this approach, well-being is conceptualized not so much as the absence of various disorders of personal or professional origin (e.g., low stress, low anxiety, absence of emotional burnout, apathy, amotivation), but rather as the feeling of fullness and sufficiency of personal resources for a productive life rich in positive experiences (Ryff, 1989; Seligman, 2012; Ryan, Deci, 2000).

Based on M. Seligman's theory, a group of psychologists led by M. Kern (Kern et al., 2016; Kern et al., 2015) developed a model of adolescent PWB. This model was symbolically named EPOCH (epoch) based on an acronym from the first letters of the five core components of PWB: adolescent engagement in activity (Engagement); perseverance and persistence (Perseverance); optimism, characterized by hope and confidence in the future (Optimism); connectedness with other people (Connectedness); happiness as a stable state of positive mood and life satisfaction (Happiness). In 2016, based on the EPOCH model, a specific questionnaire for assessing adolescent PWB was created (Kern et al., 2016). Results from its pilot testing on samples of Australian, American, and Chinese adolescents (Kern et al., 2019) demonstrated its good measurement capability and construct validity. The aim of this article is to assess the psychometric parameters of the EPOCH questionnaire and, based on this, determine the possibilities for its use in studying the PWB of Russian adolescents.

Materials and methods

Sample. The study involved 230 respondents aged 13 to 16 years (M = 14,65, SD = 1,04), 36,1% boys, 63,9% girls; students in grades 7-9 of general education schools in a large industrial metropolis. The data are stored in the database of the affiliated organization. The data collection procedure complied with the ethical standards of the Russian Psychological Society.

Methods. The 20-item version of the EPOCH questionnaire was chosen for the study as the most compact and possessing the best psychometric characteristics in both the original and translated versions (Kern et al., 2016; Kern et al., 2019). The translation was carried out following procedures for translating foreign-language instruments.
The 20 questionnaire items contained self-report statements for the scales "Engagement," "Perseverance," "Connectedness," "Optimism," and "Happiness." All questionnaire items were direct (positively worded). A Likert scale from 1 ("not at all like me") to 5 ("very much like me") was used for answer gradation. Furthermore, the questionnaire allowed for the calculation of a total score as an integral characteristic (index) of psychological well-being.
To test construct validity, the "Strengths and Difficulties Questionnaire" for children aged 11–17 (SDQ) (Goodman, 1997), recommended by the World Health Organization for assessing adolescent mental health and having evidence of validity and reliability, with normative data for Russian adolescents (Goodman et al., 2005), and the Basic Psychological Needs Satisfaction Questionnaire (Deci, Ryan, 2008; Ryan, Deci, 2000), the results of which correlate with human psychological well-being at different stages of onto- and professional development, including adolescents (Gagne, 2003), were used.

Data analysis. The analysis was performed using R (version 4.3.2, packages psych (Revelle, 2024; Rosseel, 2012) and Bifactor Indices Calculator based on the Excel calculator (Dueber, 2017). Descriptive statistics are presented in Appendix 1.
To assess the structure of the EPOCH questionnaire, confirmatory factor analysis (CFA) was used. Three models were compared: a five-factor model (each factor includes 4 items), a five-factor model with one general factor (each factor includes four items, and the obtained factors are combined into one general factor), and a one-factor model (one factor includes all 20 items). Robust estimation methods were used in model construction. The following criteria were used as model fit indices and for comparing models: Root Mean Square Error of Approximation (RMSEA), Standardized Root Mean Square Residual (SRMR), Tucker-Lewis Index (TLI), Comparative Fit Index (CFI). Models were considered acceptable with values of RMSEA ≤ 0.06, SRMR ≤ 0.09, TLI and CFI > 0.90 (Hu, Bentler, 1999).
Since the questionnaire responses are coded on a 5-point Likert scale, the data can be considered ordinal. Therefore, besides robust maximum likelihood (ML) estimation, diagonally weighted least squares (DWLS) estimation, calculated based on the polychoric correlation matrix between items, was also used.

In assessing internal consistency, alongside the traditional Cronbach's alpha (α), a hierarchical model was considered. This model allows estimation of what portion of the variance pertains to specific factors (well-being components) and what portion is explained by a general factor underlying them (general well-being). For this, a range of ω coefficients were used (Revelle, Zinbarg, 2009; Rodriguez et al., 2016a): ω estimates internal consistency based on all sources of variance (for the general factor – all 20 items, for each factor – its constituent items), ωH estimates it based solely on one general latent factor (differences in scores across the 5 factors are considered measurement error of one general factor), ωS estimates it based on the specified scales in the presence of a general factor, ωHS estimates it based solely on the specified scales excluding the commonality generated by the general factor. The relationship between ω and ωH allows assessment of how much response consistency is explained by one general factor; the relationship between ωH and ωHS – higher values of ωHS with lower values of ωH indicate the presence of multiple scales within the construct (Rodriguez et al., 2016).

Construct validity was assessed using Spearman's rank correlations between the questionnaire scales and the general factor with the SDQ and the Basic Psychological Needs Satisfaction Questionnaire. P-value calculations were performed using the Bonferroni correction.

Results

Fit indices for different models (one-factor, five-factor, and hierarchical) are presented in Table 1. All models except the one-factor model estimated with ML were acceptable. Overall, one-factor models fit the data somewhat worse than the five-factor and hierarchical models. The best-fitting model was the five-factor model (RMSEA = 0,04, 90% confidence interval [0,03; 0,06], TLI = 0,99, CFI = 1,00, SRMR = 0,06) estimated with DWLS. Covariances between scales were practically identical for ML and DWLS models (see Appendix 3). In the hierarchical DWLS model, the five factors loaded slightly more strongly onto the general factor than in the ML model. Individual items loaded moderately (from 0,4) to highly (from 0,6) onto their corresponding scales across all presented models. Item 11 ("I get so involved in activities that I forget about everything else") loaded somewhat lower (0,3-0,45) on the "Engagement" scale, and item 9 ("I complete my homework assignments from start to finish and don't stop until I finish them") loaded lower (0.25-0.4) on the "Perseverance" scale. Correlations between individual items are presented in Appendix.

Table 1. Fit indices for CFA Models

Internal consistency

The questionnaire scales demonstrate at least sufficient internal consistency (Cronbach's α ranging from 0,63 to 0,84). Overall, the questionnaire items show high internal consistency (α = 0,91, ω = 0,92) regarding the general measured construct. Excluding the factor-scales changes the overall internal consistency from high to acceptable (ωH = 0,62). The values of ωH and ωHS diverge only slightly (range of differences 0,03-0,09), indicating that several scales are identifiable within the general well-being construct.

Table 2. Internal consistency indices for EPOCH scales

Construct validity

Spearman correlation coefficients are presented in Table 3. As expected based on the content of the measured constructs, the questionnaire scales showed weak negative correlations with Emotional Problems (-0,15 to -0,27) and Conduct Problems (-0,18 to -0,29), except for the Engagement scale. The questionnaire scales showed weak positive correlations with Hyperactivity-Inattention (0,18 to 0,25) and Peer Problems (0,17 to 0,28). Moderate strength correlations were observed with the Prosocial Behavior scale (0,33 to 0,54) and the Basic Needs scales (0,28 to 0,67).

Table 3. Spearman correlations between EPOCH, SDQ and basic psychological needs satisfaction scales

Note: Correlation coefficient values for which p-values are greater than 0,05 are italicized (Not significant)

Discussion

The objective of this article was to test the psychometric indicators of the EPOCH questionnaire: its structure and construct validity. When examining the factor loadings of individual scales, it turned out that item 9 ("I complete my homework assignments from start to finish and don't stop until I finish them") loaded weakly onto its corresponding "Perseverance" scale (see Appendix). We assumed that the item might not have been entirely clear to respondents. Therefore, we changed the translation and propose using the text "I do my homework to the end" instead of "I complete my homework assignments from start to finish and don't stop until I finish them" in the future.

When comparing the mean scale values in the Russian and English-speaking samples, the higher value on the "Happiness" scale among Russian adolescents compared to adolescents from the USA is noteworthy. Deviations in the values of other scales, in our opinion, are less pronounced, taking into account the significant difference in sample sizes. A similar tendency was noted during the adaptation of the EPOCH questionnaire and its translation into Chinese (Kern et al., 2019): on average, responses to each of the four items on the "Happiness" scale were also significantly higher among Chinese adolescents and comparable to our results. This may reflect cultural specificity and should be taken into account when using the questionnaire in Russia. Some studies have shown that cultural characteristics can strongly influence the mean values on measured scales, with the changes becoming increasingly noticeable as sample sizes increase (Steinmetz, 2013), which will require further research.

Confirmatory factor analysis confirms the five-factor structure of the Russian-language version of the questionnaire. This aligns with the results of both the original psychometric study and the Chinese pilot testing of the questionnaire. The DWLS model shows slightly better fit indices, indicating rather the ordinal nature of the obtained response values. In future research using this questionnaire, methods intended for ordinal scales are preferable.

In their work (Kern et al., 2019), the authors suggest that the questionnaire might have a hierarchical structure where the five components relate to a more general well-being construct (a two-tier questionnaire structure). In our study, this assumption is partially confirmed. Firstly, in the confirmatory factor analysis, the model with five separate factors showed a better fit to the data than the hierarchical model. Secondly, the internal consistency analysis indicates that since the differences between ωH and ωHS for the individual scales and the general score are small, it is more appropriate to speak of five distinct factors that can be combined into one general construct. In practice, this means that when working with the questionnaire, each component can be measured separately as part of the general well-being indicator; however, it is premature to speak of hierarchical relationships within the model.

In studying construct validity, the questionnaire scales predictably showed positive correlations with the Basic Psychological Needs Satisfaction scales. This aligns with the core tenets of Self-Determination Theory, according to which satisfaction of these needs reflects well-being and flourishing (Martela, Sheldon, 2019). Negative correlations between well-being indicators and the problem scales of the SDQ questionnaire were also expected, in accordance with previous research (e.g., Dudovitz et al., 2022; Leavey et al., 2020). Nevertheless, the EPOCH questionnaire scales showed positive correlations with indicators of hyperactivity problems and peer relationship problems among adolescents in our sample. This could be explained, for example, by age-specific characteristics: hyperactivity tendencies and the significance of peer relationship problems are characteristic of adolescents; these problems might be differently related to the subjective experience of well-being/maladjustment in adolescence (Sekaran et al., 2024). Adolescents' perception of their own hyperactivity might reflect their competence and skillfulness and thus correlate with PWB. Problems in peer relationships might be perceived as evidence of having relationships, regardless of their valence. For an adolescent, the most traumatic experience is the absence of relationships and the feeling of exclusion from the group. Possibly, the link between PWB and peer problems reflects this feature.

Conclusion

The result of this study is the psychometric testing of the Russian-language version of the EPOCH questionnaire for assessing adolescent PWB. The questionnaire is structurally analogous to the original version and includes 20 self-report items that assess the expression of the five components of the PWB model – five scales: Engagement; Perseverance; Optimism; Connectedness; Happiness. The integral indicator of PWB is determined by summing the scale values.

The pilot testing of the questionnaire on a Russian sample of adolescents and the analysis of its characteristics demonstrated that the questionnaire meets the requirements for such measurement instruments. The questionnaire represents a reliable, convenient, and compact tool and can be used for research and practical activities in the fields of developmental psychology, educational psychology, as well as in studies of adolescent psychological well-being in adjacent fields such as sociology, medical sciences, and pedagogy across different social contexts.

A limitation of this study is the relatively small size of the research sample and its homogeneity in terms of age, gender, place of residence, and education level. Nevertheless, the results of the theoretical analysis of the well-being construct under study, as well as the psychometric characteristics of the Russian-language version of the questionnaire, suggest that the questionnaire will yield adequate results on larger and more diverse Russian-speaking samples. This assumption requires further empirical verification.

Promising directions for future research on the further adaptation of the questionnaire for Russian adolescents include testing other versions, particularly the 25-item version used in some international studies. Furthermore, investigating the relationships between well-being components and various respondent characteristics (gender, family composition, number of friends, social status, individual psychological characteristics, etc.) is of interest. The development of test norms depending on age, education level, and other indicators differentiating scale values also appears promising.

1. Adler, A., Seligman, M.E. (2016). Using wellbeing for public policy: Theory, measurement, and recommendations. International Journal of Wellbeing, 6(1), 1–35.
2. Belinskaya, E.P., Shaehov, Z.D. (2023). The relationship between psychological well-being and adaptation to digital world risks in youth. Moscow University Psychology Bulletin, 46(3), 239–260. (In Russ.). https://doi.org/10.11621/LPJ-23-35
3. Carlo, G., Randall, B.A. (2002). The Development of a Measure of Prosocial Behaviors for Late Adolescents. Journal of Youth and Adolescence, 31(1), 31–44.
4. Deci, E.L., Ryan, R.M. (2008). Self-determination theory: A macrotheory of human motivation, development, and health. Canadian Psychology, 49(3), 182–185.
5. Dekhtyarenko, A.A., Savchenko, N.L., Shlyagina, E.I. (2023). On the Differential Contribution of Personality Traits to Subjective Well-being Components. Moscow University Psychology Bulletin, 46(3), 120–142. (In Russ.). https://doi.org/10.11621/LPJ-23-30
6. Dray, J., Bowman, J., Campbell, E., Freund, M., Wolfenden, L., Hodder, R.K., McElwaine, K., Tremain, D., Bartlem, K., Bailey, J., Small, T. (2017). Systematic review of universal resilience-focused interventions targeting child and adolescent mental health in the school setting. Journal of the American Academy of Child and Adolescent Psychiatry, 56(10), 813–824.
7. Dudovitz, R.N., Thomas, K., Shah, M.D., Szilagyi, P.G., Vizueta, N., Vangala, S., Shetgiri, Kapteyn, A. (2022). School-Age Children's Wellbeing and School-Related Needs During the COVID-19 Pandemic. Academic Pediatrics, 22(8), 1368–1374. https://doi.org/10.1016/j.acap.2022.01.015
8. Dueber, D.M. (2017). Bifactor Indices Calculator: A Microsoft Excel-based tool to calculate various indices relevant to bifactor CFA models. https://doi.org/10.13023/edp.tool.01
9. Gagne, M. (2003). The role of autonomy support and autonomy orientation in prosocial behavior engagement. Motivation and Emotion, 27, 199–223. https://doi.org/10.1023/A:1025007614869
10. Goodman, R. (1997). The strengths and difficulties questionnaire: a research note. Journal of Child Psychology and Psychiatry, 38(5), 581–586.
11. Goodman, R., Slobodskaya, H.R., Knyazev, G.G. (2005). Russian child mental health: a cross-sectional study of prevalence and risk factors. European Child & Adolescent Psychiatry, 14, 28–33.
12. Hu, L., Bentler, P.M. (1999). Cutoff criteria for fit indexes in covariance structure analysis: Conventional criteria versus new alternatives. Structural Equation Modeling, 6, 1–55. https://doi.org/10.1080/10705519909540118
13. Isaeva, O.M., Akimova, A.Yu., Volkova, E.N. (2022). Factors of psychological well-being of Russian youth. Psychological Science and Education, 27(4), 24–35. (In Russ.). https://doi.org/10.17759/pse.2022270403
14. Kern, M.L., Benson, L., Steinberg, E.A., Steinberg, L. (2016). The EPOCH measure of adolescent well-being. Psychological Assessment, 28, 586–597. http://dx.doi.org/10.1037/pas0000201
15. Kern, M.L., Waters, L.E., Adler, A., White, M.A. (2015). A multidimensional approach to measuring wellbeing in students: Application of the PERMA framework. The Journal of Positive Psychology, 10(3), 262–271.
16. Kern, M.L., Zeng, G., Hou, H., Peng, K. (2019). The Chinese Version of the EPOCH Measure of Adolescent Well-Being: Testing Cross-Cultural Measurement Invariance. Journal of Psychoeducational Assessment, 37(6), 757–769. https://doi.org/10.1177/0734282918789561
17. Leavey, G., Rosato, M., Harding, S., Corry, D., Divin, N., Breslin, G. (2020). Adolescent mental health problems, suicidality and seeking help from general practice: A cross-sectional study (Northern Ireland Schools and Wellbeing study). Journal of Affective Disorders, 274, 535–544. https://doi.org/10.1016/j.jad.2020.05.083
18. Martela, F., Sheldon, K.M. (2019). Clarifying the Concept of Well-Being: Psychological Need Satisfaction as the Common Core Connecting Eudaimonic and Subjective Well-Being. Review of General Psychology, 23(4), 458–474. https://doi.org/10.1177/1089268019880886
19. Polivanova, K.N. (2020). New educational discourse: Schoolchildren’s well-being. Cultural-Historical Psychology, 16(4), 26–34. (In Russ.). https://doi.org/10.17759/chp.2020160403
20. Potanina, A.M., Morosanova, V.I. (2022). Individual-typical features of the relationship between conscious self-regulation, psychological well-being, and academic performance in 6th-grade students. Theoretical and Experimental Psychology, 15(1), 52–78. (In Russ.). https://doi.org/10.24412/2073-0861-2022-1-52-78
21. Rasskazova, E.I., Sadovnichaya, V.S. (2023). Subjective dissatisfaction with interpersonal communication needs and psychological distress in adolescents and young adults with varying levels of suicide risk. Moscow University Psychology Bulletin, 46(4), 112–130. (In Russ.). https://doi.org/10.11621/LPJ-23-41
22. Revelle, W. (2024). Psych: Procedures for Psychological, Psychometric and Personality Research. Northwestern University, Evanston, Illinois. R package version 2.4.6. https://CRAN.R-project.org/package=psych
23. Revelle, W., Zinbarg, R.E. (2009). Coefficients alpha, beta, omega, and the glb: Comments on Sijtsma. Psychometrika, 74(1), 145–154. https://doi.org/10.1007/s11336-008-9102-z
24. Rikel, A.M., Tuniyants, A.A., Batyrova, N. (2017). The concept of subjective well-being in hedonistic and eudaimonic approaches. Moscow University Psychology Bulletin, (2), 64–82. (In Russ.). https://doi.org/10.11621/vsp.2017.02
25. Rodriguez, A., Reise, S.P., Haviland, M.G. (2016). Applying bifactor statistical indices in the evaluation of psychological measures. Journal of Personality Assessment, 98(3), 223–237. https://doi.org/10.1080/00223891.2015.1089249
26. Rodriguez, A., Reise, S.P., Haviland, M.G. (2016). Evaluating bifactor models: Calculating and interpreting statistical indices. Psychological Methods, 21(2), 137–150. https://doi.org/10.1037/met0000045
27. Rosseel, Y. (2012). lavaan: An R Package for Structural Equation Modeling. Journal of Statistical Software, 48(2), 1–36. https://doi.org/10.18637/jss.v048.i02
28. Rudnova, N.A., Kornienko, D.S., Volkova, E.N., Isaeva, O.M. (2023). Digital Parental Mediation and Its Connection to Psychological Well-being Indicators in School-Aged Children. Television Science, 19(1), 175–198. (In Russ.). https://doi.org/10.30628/1994-9529-2023-19.1-175-198.
29. Ryan, R.M., Deci, E.L. (2000). Self-determination theory and the facilitation of intrinsic motivation, social development, and well-being. American Psychologist, 55(1), 68–78. https://doi.org/10.1037/0003-066X.55.1.68
30. Ryff, C.D. (1989). Happiness is everything, or is it? Explorations on the meaning of psychological well-being. Journal of Personality and Social Psychology, 57(6), 1069–1081.
31. Sekaran, V.C., Prabhu, V., Ashok, L., D'Souza, B., Shetty, S.D., Nair, R. (2024). Predictors of Adolescent Internalizing and Externalizing Mental Health Symptoms: A School-Based Study in Southern India. International Journal of Environmental Research and Public Health, 21(4), 393. https://doi.org/10.3390/ijerph21040393
32. Seligman, M.E. (2012). Flourish: A visionary new understanding of happiness and well-being. New York: Simon and Schuster.
33. Shilko, R.S., Dolgikh, A.G., Almazova, O.V. (2018). Measuring mental health in educational settings: Warwick-Edinburgh Mental Well-being Scale. In: L.A. Tsvetkova, E.N. Volkova, A.V. Miklyaeva (Ed.), Herzen Readings: Psychological Research in Education. In 2 parts. Part 1(pp. 285–290). Saint Petersburg: RSPU. (In Russ.).
34. Steinmetz, H. (2013). Analyzing observed composite differences across groups: Is partial measurement invariance enough? Methodology, 9(1), 1–12. https://doi.org/10.1027/1614-2241/a000049
35. Volkova, E.N., Miklyaeva, A.V., Khoroshikh, V.V. (2022). Subjective prerequisites for psychological well-being of gifted adolescents. Psychological Science and Education, 27(1), 92–103. (In Russ.). https://doi.org/10.17759/pse.2022270108
36. Volkova, E.N., Sorokoumova, G.V. (2024). Psychological criteria of well-being in modern adolescents in the context of studying digital socialization. Social Psychology and Society, 15(2), 12–27. (In Russ.). https://doi.org/10.17759/sps.20241502024