Sample and data

The Czech-language adaptation of the SWEMWBS for adolescents was based on data collected through an online questionnaire (CASI method) in October and November 2023 within the CZEPS project (https://czeps.soc.cas.cz/en) [11]. Respondents were first-year secondary school students aged 15 to 18. Only respondents who answered all SWEMWBS items were included in the analysis. The final research sample (N) comprised 22,498 students aged 15 to 18 (mean age 15.6 years), of which 10,757 were male (47.8%), 11,045 female (49.1%) and 696 other (3.1%). For further details regarding the structure of the research sample, see Table A1 in the Appendix.

MeasuresThe Short Warwick-Edinburgh Mental Well-Being Scale (SWEMWBS)Footnote 1

The SWEMWBS is a shortened version of the original 14-item WEMWBS by Stewart-Brown et al. [12] The SWEMWBS consists of seven positively-worded items covering both aspects of mental well-being – feeling good and functioning well – that respondents rated using a 5-point Likert scale (1 ‘none of the time’, 2 ‘rarely’, 3 ‘some of the time’, 4 ‘often’, 5 ‘all of the time’). The evaluation is calculated by summing the scores of each item. The total raw score ranges from 7 to 35 (a higher value means a higher level of mental well-being). However, for analyses, it is necessary to transform the raw score into a metric score [12]. The SWEMWBS has been translated into Czech through the TRAPD approach, which is an acronym for the steps of the translation process; specifically Translation, Review, Adjudication, Pretest, and Documentation [16]. Three independent experts translated the items into Czech, followed by a review and assessment, and the most appropriate wording was selected. Then, pilot testing of the Czech translation was conducted among 74 students from two secondary schools via an online questionnaire. The results of the pilot testing indicated that only minor changes were necessary, such as alterations to the word order and the use of synonyms. All stages of the translation process were documented. The original wording and the final Czech translation are shown in Table 1.

Table 1 Final Czech translation of the SWEMWBS

For assessment of the criterion-related validity of the Czech version of the SWEMWBS, the relations with other similar instruments were investigated. The construct used and how they were assessed are described below.

Overall life satisfaction

Life satisfaction was measured by a traditional one-item question: “All things considered, how satisfied are you with your life as a whole nowadays? Please answer on a scale of 0 to 10, where 0 means ‘extremely dissatisfied’ and 10 means ‘extremely satisfied’.”

Happiness

Positive affect was measured by a simple question on happiness asking respondents: “Taking all things together, how happy would you say you are?” They answered on an 11-point response scale from 0 ‘extremely unhappy’ to 10 ‘extremely happy’.

General health

One question was used to assess subjective health with the following wording: “Would you say that your health in general is… (1) poor, (2) average, (3) good, (4) very good, or (5) excellent?”.

General Anxiety Disorder-2 (GAD-2)

The GAD-2 is an ultra-short version of the original 7-item version containing two items (Feeling nervous, anxious, or on edge; Not being able to stop or control worrying) measuring anxiety [17]. Respondents were asked, “How often have they been bothered by these problems over the last two weeks?” and answered on a 4-point response scale (0 ‘Not at all’, 1 ‘Several days’, 2 ‘More than half days’, 3 ‘Almost every day’).

Patient Health Questionnaire-2 (PHQ-2)

Depression was measured by the 2-item PHQ-2 [18], which contains nine items in the original version. The wording of the question and response scale is the same as for GAD-2. The wording of the items is as follows: Little interest or pleasure in doing things and Feeling down, depressed or hopeless.

Brief Resilience Scale (BRS)

The BRS is a short 6-item scale created to assess the perceived ability to bounce back or recover from stress [19]. The scale includes three positively (items 1, 3, and 5), and three negatively (items 2, 4, and 6) worded items rate on a 5-point response scale (1 ‘strongly disagree’, 2 ‘disagree’, 3 ‘neither agree nor disagree’, 4 ‘agree’, 5 ‘strongly agree’). The total score is calculated as the sum of the individual items after recoding.

Statistical analysis plan and methods

The validation process was conducted in several steps using various methods. First, descriptive statistics were performed. Subsequently, the assumptions for using IRT, the main method for testing the psychometric properties of the Czech translation of the SWEMWBS, were tested. The scale items have five ordered categories and are polytomous, so we applied two models to them and compared their results, namely the General Partial Credit Model (GPCM) [20] and the Graded Response Model (GRM) [21]. Using the more appropriate model, we estimated one discriminant parameter (a) and four threshold parameters (b) for each item (the number of threshold parameters is always one less than the number of items on the response scale). The discriminant parameter indicates the item's relationship to the scale and its ability to differentiate between respondents with different levels of the measured concept, also known as the latent trait (theta, θ). The parameter a typically ranges from 0 to 2, but theoretically, it can range from –∞ to + ∞ [22]. The interpretation of parameter b, which typically ranges from –3 to + 3 [23], varies depending on the model used. For the GPCM, it represents the value of the latent variable required to move between two adjacent categories on the response scale, while for the GRM, it denotes the 50% probability that the respondent will select the given category or a higher category on the response scale. The evaluation was also conducted using illustrative graphs such as item characteristic curve (ICC), category characteristic curve (CCC), item information function (IIF), and test information function (TIF). The final step involved testing reliability by using the coefficients Cronbach's alpha and McDonald's omega, and various types of validity, namely construct validity through confirmatory factor analysis (CFA), convergent validity based on Average Variance Extracted (AVE) and Composite Reliability (CR), and criterion-related validity by correlation with other relevant measures.

Data preparation and basic analyses, including descriptive statistics, reliability, validity, and unidimensionality testing, were conducted using SPSS 27, CFA tested construct validity in Mplus 7.4, and IRT analysis was performed in STATA 17 and R using the mirt package.