Analysing Comparability of Four Multi-Item Well-being Psychometric Scales Among 35 Countries Using Children’s Worlds 3rd Wave 10 and 12-year-olds Samples

• Published in: Child indicators research 2021-10, Vol.14 (5), p.1829-1861
• Main Authors: Casas, Ferran, González-Carrasco, Mònica
• Format: Article
• Language: English
• Subjects: Age groups; Child and School Psychology; Children; Comparative studies; Cultural differences; Cultural factors; Cultural Influences; Early Childhood Education; International comparisons; Multiculturalism & pluralism; Psychometrics; Quality of Life Research; Quantitative psychology; Reliability; Social Sciences; Social Work; Socioeconomic factors; Well being
• ISSN: 1874-897X; 1874-8988
• DOI: 10.1007/s12187-021-09825-0

The third Wave Children’s Worlds dataset included data obtained from four self-reported well-being multi-item psychometric scales. Three of these were new improved versions of scales used in the second wave and the fourth had not been previously tested for reliability or comparability. The four scales were administered to children in the 10-year-olds age group in 35 countries and in the 12-year-olds age group in 30 countries. Many of these countries were not involved in the second wave. Consequently, the fit for all of these scales needs to be checked and its comparability among the new set of countries analysed. Based on our analysis, there can be no doubt that the most advisable psychometric scale for use in international comparisons is the 5-item version of the context-free subjective well-being scale for both the 10 and 12yo groups, because it displays metric invariance. Scalar invariance for this scale is only supported with a semi-partial constraint, because its item number 4 ( Things in my life are excellent) seems to be very sensitive to the socio-economic level of the country, and generally tends to explain less variance than any other. Therefore, great caution must be used when comparing mean scores for this scale across countries. Items on the other three scales would appear to have different cultural meanings, depending on the country, as they do not display metric or scalar invariance across countries. However, the positive and negative affects scale deserves special attention, because the in-country correlation between its two components displays an interesting diversity between countries, probably due to cultural factors that should be further explored. Although statistics obtained using the overall index for any of these three scales cannot be meaningfully compared across countries, their fit is excellent and their use is deemed appropriate for in-country analysis. Further exploration is needed to test whether non-comparability is an issue with only some specific items, and whether they are comparable between some of the countries and not others.