Vera- Villarroel et al. (2013)

1646 participants between 18 and 90 years, with different educational levels and sociodemographic conditions, from Chile.

Ryff’s PWBS, reduced version, with 29 items, from the version of Diaz et al. (2006) and adapted by Tomas et al. (2010).

Internal reliability (Cronbach’s alpha). Confirmatory factor analyzes to evaluate models of five and six first order factors, and six first order factors with a second order factor. Test-retest reliability after three weeks with a subsample of 180 participants. Correlations between items of the scales and their dimensions.

Four of the six dimensions do not meet the criteria of reliability, having values below .60 in some dimensions (environmental mastery and personal growth). The correlations between each item with the total scale and with their respective theoretical dimension were, in most cases, moderate (between .50 and .80). Internal consistencies were between .47 and .82; the correlations among factors were between .47 and .82; the construct validity for a model of five, one of six factors and one of six with one of the second order indicated that the best model fit was for the six-factor model. Cronbach’s alpha values of the second application were between .59 and .82; the correlations among the scales were between .46 and .79.

The best fit model was that of six factors without second order factors, such as the model proposed by Ryff, a finding that is different from other studies in which there was a better fit in the six-factor model with a second-order model with detailed analysis done by age groups. As for the correlation of the items with their corresponding theoretical domains, the dimensions have better correlations in samples between 25 and 65 years. The scale had a good test-retest reliability.

Díaz et al. (2015)

400 participants between 18 and 64 years of age, from Colombia and Spain

Ryff’s PWBS, reduced version (van Dierendonck adapted to Spanish by Díaz et al., 2006 ) with 39 items.

Exploratory factorial analysis. Descriptive analyses and analysis for four models (one factor, two oblique factors, two orthogonal factors, two factors with a second order factor called general well-being). Goodness of fit indices for the models.

The analyses indicate that the model with the best model fit was that of six factors. A different scale that measures subjective well-being was correlated with the PWBS and the correlation was adequate. The values of Cronbach’s alpha coefficients of the scales were between .68 and .86.

The hexadimensional model proposed by Ryff in different studies was replicated.

Rosa-Rodríguez et al. (2015)

768 undergraduate and graduate students between 17 and 74 years, from Puerto Rico.

Ryff’s PWBS, reduced version with 29 items (van Dierendonck et al., 2007, in Spanish) and revised for linguistic equivalence.

Descriptive analyses and internal reliability (Cronbach’s alpha).

Alpha values were between .47 and .80 (environment mastery α = .47, personal growth α = .59, self-acceptance α = .74, positive relations with others α = .71, autonomy α = .60, and purpose in life α = .80).

Findings were similar to other studies that use the 29 and 39 item scales. The low reliability scores in the personal growth and environmental mastery domains indicate the need to evaluate the factorial structure of the scales.

Valenzuela (2015)

1100 university students with a mean age of 22 years, from Mexico.

Ryff’s PWBS ( Díaz et al., 2006 ), in its 29 items version from the original English version, translated, back-translated, and with an added item to the personal growth dimension. The scale has a total of 30 items, five for each dimension.

Confirmatory factor analyses with different models: one with six factors, then different structures. For the six factors, a confirmatory strategy was used and then exploratory analyses to test other models.

The six-factor model has inadequate fit indices and Cronbach’s alpha values between .41 and .58, far from the recommended standards. The structure with the best fit is the two-factor structure, made up of half of the items, which presented a low positive correlation between them (.10), and with Cronbach’s alpha values for the scales between .81 and .86, which allows us to think about the interdependence of dimensions. The two-factor model consisted of 15 items, in two dimensions: personal growth and self-acceptance.

The six-factor model did not show good model fit indices and the explanation for this is could be a poor factorial validity and low internal reliabilities of the scales, since Ryff derived his model from positive psychology and psychotherapy. Findings in this regard indicate an overlap in four of the six dimensions (personal growth, purpose in life, self-acceptance and environmental mastery). Finally, an items was added to the 29-item version of Diaz but the model still had difficulties in the model fit.