Resiliency Across Cultures: A Validation of the Resiliency Scale for Young Adults

Participants and Procedure

The Canadian sample consisted of 289 undergraduate university students (226 females, 63 males) from a central Canadian university. Students’ ages ranged from 17 to 25 years (M_age = 17.94 years, SD = 0.81 years) and they completed a series of online questionnaires for course credit and a prize draw. A subset of the original Canadian sample (N = 123; 104 females, 19 males; M_age = 17.97 years, SD = 0.65 years) completed the same online questionnaires three months later. The Italian sample of 259 undergraduate students (173 females, 86 males; M_age = 23.27 years, SD = 2.81 years) from a large university in central Italy completed a series of questionnaires using a paper and pencil format. The RSYA was translated into Italian using standard assessment translating procedures by subject matter experts fluent in Italian and English, to ensure both content equivalence and lexical difficulty level were maintained (e.g., Hambleton & Lee, 2013). All additional study measures employed the existing Italian versions.

Measures

Data Analytic Strategy

To test for convergent validity of the RSYA, authors evaluated correlations between the RSYA and relevant personality variables using SPSS Version 22 (IBM, 2013). Internal consistency and test–retest reliability were also examined using SPSS Version 22. A confirmatory factor analysis was conducted using Mplus Version 7.4 (Muthén & Muthén, 1998-2012) to test the dimensionality of the RSYA at the facet level among both the Canadian and Italian samples. Maximum likelihood robust (MLR) estimator was used to correct standard errors for nonnormality in the data. To evaluate cultural invariance of the RSYA across Canadian and Italian samples, a series of nested models were tested using maximum likelihood and compared using chi-square, comparative fit index (CFI), and root mean square error approximation (RMSEA) difference tests.

Descriptive Statistics and Convergent Validity

Descriptive statistics and bivariate correlations between the RSYA 10 facets, three factors, trait EI, and general personality variables for the Canadian and Italian samples are presented in Tables 1 and 2, respectively.

OPEN IN VIEWER

Table 1.

Reliabilities, Descriptive Statistics, and Bivariate Correlations for Study Variables in the Canadian Sample.

	α	FemaleM (SD)	MaleM (SD)	1	2	3	4	5	6	7	8	9	10	11	12	13	14	15	16	17	18
1. Sense of Mastery	.89	2.71 (0.54)	2.86 (0.54)	1
2. Optimism	.87	2.91 (0.71)	2.91 (0.70)	.87**	1
3. Self-Efficacy	.78	2.71 (0.56)	2.90 (0.57)	.81**	.59**	1
4. Adaptability	.78	2.51 (0.65)	2.76 (0.65)	.83**	.57*	.51**	1
5. Sense of Relatedness	.90	2.74 (0.54)	2.80 (0.49)	.68**	.63**	.59**	.50**	1
6. Trust	.81	2.45 (0.74)	2.45 (0.70)	.53**	.49**	.44**	.40**	.84**	1
7. Access to Support	.81	3.05 (0.70)	3.02 (0.60)	.52**	.51**	.50**	.31**	.81**	.63**	1
8. Comfort with Others	.85	2.60 (0.72)	2.83 (0.62)	.62**	.52**	.54**	.50**	.81**	.59**	.48**	1
9. Tolerance	.67	2.87 (0.58)	2.92 (0.52)	.44**	.42**	.33**	.33**	.65**	.34**	.39**	.40**	1
10. Emotional Reactivity	.90	1.68 (0.59)	1.40 (0.70)	−.44**	−.37**	−.39**	−.34**	−.37**	−.27**	−.28**	−.31**	−.30**	1
11. Sensitivity	.79	1.90 (0.74)	1.50 (0.79)	−.41**	−.34**	−.36**	−.34**	−.35**	−.24**	−.24**	−.34**	−.26**	.89**	1
12. Recovery	.84	1.23 (0.71)	0.92 (0.74)	−.38**	−.36**	−.30**	−.30**	−.43**	−.33**	−.32**	−.30**	−.41**	.79**	.54**	1
13. Impairment	.82	1.89 (0.73)	1.78 (0.83)	−.31**	−.24**	−.32**	−.23**	−.16**	−.12*	−.14*	−.14*	−.10	.84**	.67**	.44**	1
14. Total EI	.89	4.64 (0.74)	4.86 (0.82)	.74**	.65**	.69**	.53**	.68**	.53**	.55**	.65**	.36**	−.59**	−.56**	−.45**	−.45**	1
15. Extraversion	.80	2.91 (0.99)	2.97 (0.86)	.34**	.29**	.26**	.31**	.47**	.41**	.24**	.62**	.13*	−.12*	−.16**	−.13*	−.03	.43**	1
16. Agreeableness	.73	3.81 (0.77)	3.73 (0.71)	.25**	.29**	.12*	.22**	.35**	.28**	.24**	.25**	.34**	−.10	−.05	−.14*	−.07	.32**	.26**	1
17. Conscientiousness	.70	3.42 (0.82)	3.62 (0.77)	.31**	.19**	.38**	.22**	.20**	.20**	.19**	.12*	.12*	−.35**	−.30**	−.25**	−.34**	.37**	−.00	.03	1
18. Neuroticism	.63	2.94 (0.83)	2.39 (0.84)	−.39**	−.35**	−.39**	−.25**	−.34**	−.25**	−.24**	−.34**	−.22**	.69**	.75**	.46**	.52**	−.57**	−.19**	−.01	−.23**	1
19. Openness	.73	3.49 (0.76)	3.66 (0.84)	.24**	.23**	.23**	.15**	.20**	.08	.13*	.20**	.20**	−.22**	−.20**	−.21**	−.15*	.27**	.12*	.29**	.08	−.12

Note. Personality variables derived from Mini-IPIP. EI = Emotional Intelligence; IPIP = International Personality Item Pool.

*

p < .05. **p < .01.

OPEN IN VIEWER

Table 2.

Reliabilities, Descriptive Statistics, and Bivariate Correlations for Study Variables in the Italian Sample.

	α	FemaleM (SD)	MaleM (SD)	1	2	3	4	5	6	7	8	9	10	11	12	13	14	15	16	17	18	19
1. Sense of Mastery	.89	2.67 (0.63)	2.72 (0.55)	1
2. Optimism	.82	2.59 (0.81)	2.66 (0.71)	.89**	1
3. Self-Efficacy	.67	2.70 (0.61)	2.77 (0.54)	.86**	.63**	1
4. Adaptability	.75	2.71 (0.69)	2.73 (0.71)	.89**	.67**	.68**	1
5. Sense of Relatedness	.86	2.79 (0.50)	2.61 (0.55)	.66**	.59**	.54**	.61**	1
6. Trust	.71	2.58 (0.65)	2.33 (0.79)	.56**	.48**	.49**	.51**	.80**	1
7. Access to Support	.73	3.10 (0.67)	2.83 (0.73)	.38**	.36**	.32**	.33**	.78**	.46**	1
8. Comfort with Others	.75	2.71 (0.72)	2.63 (0.71)	.60**	.51**	.47**	.58**	.80**	.59**	.43**	1
9. Tolerance	.55	2.79 (0.57)	2.67 (0.60)	.52**	.49**	.39**	.48**	.72**	.38**	.51**	.43**	1
10. Emotional Reactivity	.89	1.89 (0.77)	1.79 (0.71)	−.19**	−.22**	−.18**	−.11	−.09	.05	−.06	−.13*	−.14*	1
11. Sensitivity	.69	2.14 (0.79)	1.98 (0.76)	−.19**	−.22**	−.15*	−.13*	−.12	.01	−.04	−.21**	−.15*	.87**	1
12. Recovery	.80	1.47 (0.90)	1.47 (0.93)	−.19**	−.20**	−.18**	−.12	−.14*	.03	−.18**	−.07	−.22**	.83**	.57**	1
13. Impairment	.83	2.06 (0.98)	1.92 (0.82)	−.12	−.14**	−.13*	−.04	.02	.08	.07	−.07	−.00	.88**	.70**	.55**	1
14. Total EI	.85	4.51 (0.67)	4.62 (0.69)	.68**	.62**	.62**	.56**	.51**	.41**	.30**	.50**	.35**	−.42**	−.41**	−.35**	−.33**	1
15. Extraversion	.78	3.07 (0.49)	3.17 (0.60)	.49**	.45**	.47**	.39**	.38**	.37**	.20**	.43**	.17**	−.30**	−.30**	−.24**	−.24**	.51**	1
16. Agreeableness	.76	3.00 (0.49)	3.00 (0.41)	.14*	.20**	.05	.10	.17**	.01	.08	.21**	.26**	−.40**	−.40**	−.29**	−.34**	.11	.25**	1
17. Conscientiousness	.77	3.31 (0.54)	3.22 (0.47)	.19**	.12	.27**	.14*	.19**	.11	.13*	.16*	.18**	−.31**	−.20**	−.35**	−.24**	.25**	.21**	.25**	1
18. Emotionality	.79	3.19 (0.49)	3.01 (0.46)	−.26**	−.27**	−.18**	−.22**	−.05	−.02	.06	−.13*	−.07	.23**	.25**	.11	.23**	−.25**	−.12	−.02	−.06	1
19. Openness	.78	3.42 (0.58)	3.41 (0.64)	.09	.07	.07	.10	.12	.04	.04	.12	.18**	−.25**	−.19**	−.27**	−.19**	.19**	.17**	.19**	.31**	.11	1
20. Honesty–humility	.78	3.40 (0.58)	3.08 (0.63)	.17**	.15*	.17**	.13*	.21**	.14*	.11	.23**	.19**	−.27**	−.23**	−.28**	−.19**	.14**	.13*	.17**	.46**	.03	.31**

Note. Personality variables derived from HEXACO-60. EI = Emotional Intelligence.

*

p < .05. **p < .01.

In the Canadian sample, RSYA internal consistency values for the three factors were high (α = .89-.90). Subscale internal consistency values ranged from acceptable to good (α = .67-.87). Consistent with past research using the RSYA (Prince-Embury et al., 2017), the Tolerance subscale showed the lowest alpha value. Internal consistency values were high for trait EI and Mini-IPIP variables, with the exception of the Neuroticism subscale (α = .63). One item was removed from the original subscale (“I seldom feel blue”) and the internal consistency alpha increased from .58 to .63. Test–retest reliabilities of the RSYA factors over a 3-month period were also calculated for the Canadian sample and showed good test–retest reliability (Sense of Mastery r = .67, Sense of Relatedness r = .79, Emotional Reactivity = .76).

In the Italian sample, internal consistency values for the RSYA were high for the three factors (α = .86-.89). Like the Canadian sample, RSYA Subscale alphas ranged from adequate to good (α = .67-.83), again with the exception of a lower alpha value for the Tolerance Subscale (α = .55). Internal consistency values for trait EI and the HEXACO-60 variables were high (α = .76-.85).

Consistent with the initial hypotheses, Sense of Mastery and Sense of Relatedness factors and their corresponding facets were strongly and positively correlated with one another across both samples. Interestingly, Emotional Reactivity was significantly and negatively related to Sense of Mastery and Sense of Relatedness in the Canadian sample as expected, but was essentially uncorrelated with Sense of Relatedness in the Italian sample. Consistent with current research (Prince-Embury et al., 2017), trait EI was strongly positively correlated with Sense of Relatedness and Sense of Mastery, and negatively related to Emotional Reactivity in both samples.

In the Canadian sample, Sense of Mastery and Sense of Relatedness were positively related to extraversion, agreeableness, conscientiousness, and openness, and negatively related to neuroticism. Emotional Reactivity was strongly and positively related to neuroticism, and negatively related to conscientiousness, openness, and extraversion. No relationship, however, emerged between Emotional Reactivity and agreeableness.

By comparison, in the Italian sample, the Sense of Mastery and Sense of Relatedness factors were strongly and positively related to extraversion. Both factors also showed small positive relationships with agreeableness, conscientiousness, and honesty–humility. Sense of Mastery was negatively related to emotionality; however, there was no significant relationship between Sense of Relatedness and emotionality. Both Sense of Mastery and Sense of Relatedness were generally unrelated to openness. Similar to the Canadian sample, Emotional Reactivity was negatively related to extraversion, agreeableness, conscientiousness, openness, and honesty–humility, and positively related to emotionality.

Confirmatory Factor Analysis of RSYA in Canadian and Italian Samples

Two separate three-factor measurement models examining the dimensionality of the RSYA subscales in the Canadian and Italian samples were tested using confirmatory factor analytic procedures. For the Canadian sample, the overall model fit was acceptable: χ²₍₃₂₎ = 100.66, RMSEA = .086 (90% confidence interval [CI] = [0.067, 0.105]), CFI = .927, Tucker–Lewis index (TLI) = .897. All facets loaded strongly onto their respective factors, ranging from .52 (Tolerance) to .90 (Sensitivity). For the Italian sample, model fit was adequate: χ²₍₃₂₎ = 100.88, RMSEA = .091 (90% CI = [0.071, 0.112]), CFI = .929, TLI = .901. Like the Canadian sample, all facets loaded strongly onto their respective factors, ranging from .59 (Access to Support) to .85 (Adaptability), providing additional empirical evidence for the RSYA’s three-dimensional structure.

Cultural Invariance for the RSYA

A series of nested models were tested to assess the cultural invariance of the RSYA across the Canadian and Italian samples (see Table 3). Configural model fit was acceptable, χ²₍₆₄₎ = 299.94, RMSEA = .097 (90% CI = [0.084, 0.111]), CFI = .928, TLI = .899. According to the chi-square difference test, the metric model with constrained factor loadings across the Canadian and Italian sample was significantly different than the unconstrained model, Δχ²₍₇₎ = 16.80, p < .05. However, the CFI (ΔCFI = .004) and the RMSEA (ΔRMSEA = .002), difference tests displayed metric invariance indicating the factor loadings for the RSYA did not vary significantly across cultures. However, the scalar invariance model fit significantly worse compared with the metric invariance model, Δχ²₍₇₎ = 89.28, p < .01, ΔCFI = .035. This indicates that the endorsed levels of the underlying facets (i.e., intercepts) differed significantly across Canadian and Italian samples. Examination of these intercepts revealed that those in the Italian sample scored consistently higher on Emotional Reactivity compared with those in the Canadian sample.

OPEN IN VIEWER

Table 3.

Cultural Invariance Fit Indices.

Model	χ2(df)	CFI	TLI	RMSEA	RMSEA 90% CI	SRMR
1. No constraints	229.94** (64)	.928	.899	.097	[0.084, 0.111]	.059
2. Metric invariance	246.74** (71)	.924	.904	.095	[0.082, 0.108]	.069
3. Scalar invariance	336.02** (78)	.889	.872	.110	[0.098, 0.122]	.076

Note. CFI = comparative fit index; TLI = Tucker–Lewis index; RMSEA = root mean square error approximation; CI = confidence interval; SRMR = standardized root mean square residual.

*

p < .05. **p < .01.

Limitations and Future Directions

Limitations of the RSYA primarily include those suggested in previous research (Prince-Embury et al., 2017). The RSYA is a self-report scale, and shares the limitations of other assessments using this type of measurement format. Furthermore, it does not capture all features of personal resiliency across various situations and contexts, and the predictive validity of the RSYA has yet to be explored. With regard to the current study, similar to previous research, the two samples were limited to university students. Although the scale is intended for youth and young adults, future studies should assess the predictive utility of the RSYA in more diverse groups of young adults. Furthermore, there was a gender imbalance in the two samples with the participants consisting predominantly of females. This limitation should be considered when interpreting the results. Although findings of the personality–resiliency relationships were similar, the use of two different personality scales and the resulting correlations with the RSYA may have resulted in differential correlations across samples, and cannot be strictly compared across cultures. Finally, the test–retest reliabilities were only conducted with the Canadian sample; therefore, this has yet to be replicated and evaluated in an Italian sample. Overall, the current study provided an empirical evaluation and confirmation that the RSYA is a valid measure of the three-factor model of personal resiliency in a Canadian sample. Furthermore, it provided evidence that the RSYA is a promising measure for use with Italian young adults.