Page 91 - Ana Kozina and Nora Wiium, eds. ▪︎ Positive Youth Development in Contexts. Ljubljana: Educational Research Institute, 2021. Digital Library, Dissertationes (Scientific Monographs), 42.
P. 91
measuring positive youth development in slovenia
were poor. Community assets’ target loadings were poor (the majority of
ps < .001) and non-target loadings were poor as well. The model fit and low
target loadings of context-related assets indicated that the solution with 8
factors was better.
CFA results showed an adequate fit for the 5Cs with 5 latent con-
structs (Competence, Confidence, Character, Caring, Connection): χ2(503)
= 3368.97, p <. 001, CFI = .910, RMSEA = .054, 90% CI [.052, .056], SRMR
= .063. For Competence, all factor loadings were fair (ps < .001). All fac-
tor loadings for Confidence were very good or even excellent (ps < .001).
For Character, almost all factor loadings were fair (they were above .38; ps
< .001), except for items referring to conduct behaviour. All factor load-
ings for Caring were good or very good (ps < .001). For Connection, all fac-
tor loadings were fair (ps < .001). In the CFA for the 5Cs, the 14 pairs of the
same-facet items (in Competence, Confidence, Character and Connection)
were allowed to correlate. Moreover, we applied a bifactor model that al-
lows for items to simultaneously indicate each first-order factor and a gen-
eral construct (i.e. PYD). The bifactor model of the 5Cs did not converge.
Reliability analyses
First, we examined the reliability of each scale for the whole sample. Since
gender and school level are to be compared, we examined the internal con-
sistencies for each group separately.
The Cronbach alphas of DA for the whole sample were adequate and
varied from .70 to .82, except for Constructive use of time, which was only
.43. Similarly, the reliabilities of DA except for Constructive use of time
(males: .42; females: .45) according to gender were adequate and ranged
from .69 to .82 for males and from .67 to .82 for females. The reliabilities of
DA according to school level were from .70 to .84 for lower secondary school
and from .68 to .82 for upper secondary school, except for Constructive use
of time (lower secondary school: .44; upper secondary school: .42).
The Cronbach alphas of the 5Cs for the whole sample ranged from .73
to .92. The internal consistencies for gender were from .72 to .91 for males
and from .70 to .92 for females. The reliabilities of the 5Cs according to the
school level were adequate: for lower secondary schools, they varied from
.76 to .92 and for upper secondary schools from .72 to 92.
91
were poor. Community assets’ target loadings were poor (the majority of
ps < .001) and non-target loadings were poor as well. The model fit and low
target loadings of context-related assets indicated that the solution with 8
factors was better.
CFA results showed an adequate fit for the 5Cs with 5 latent con-
structs (Competence, Confidence, Character, Caring, Connection): χ2(503)
= 3368.97, p <. 001, CFI = .910, RMSEA = .054, 90% CI [.052, .056], SRMR
= .063. For Competence, all factor loadings were fair (ps < .001). All fac-
tor loadings for Confidence were very good or even excellent (ps < .001).
For Character, almost all factor loadings were fair (they were above .38; ps
< .001), except for items referring to conduct behaviour. All factor load-
ings for Caring were good or very good (ps < .001). For Connection, all fac-
tor loadings were fair (ps < .001). In the CFA for the 5Cs, the 14 pairs of the
same-facet items (in Competence, Confidence, Character and Connection)
were allowed to correlate. Moreover, we applied a bifactor model that al-
lows for items to simultaneously indicate each first-order factor and a gen-
eral construct (i.e. PYD). The bifactor model of the 5Cs did not converge.
Reliability analyses
First, we examined the reliability of each scale for the whole sample. Since
gender and school level are to be compared, we examined the internal con-
sistencies for each group separately.
The Cronbach alphas of DA for the whole sample were adequate and
varied from .70 to .82, except for Constructive use of time, which was only
.43. Similarly, the reliabilities of DA except for Constructive use of time
(males: .42; females: .45) according to gender were adequate and ranged
from .69 to .82 for males and from .67 to .82 for females. The reliabilities of
DA according to school level were from .70 to .84 for lower secondary school
and from .68 to .82 for upper secondary school, except for Constructive use
of time (lower secondary school: .44; upper secondary school: .42).
The Cronbach alphas of the 5Cs for the whole sample ranged from .73
to .92. The internal consistencies for gender were from .72 to .91 for males
and from .70 to .92 for females. The reliabilities of the 5Cs according to the
school level were adequate: for lower secondary schools, they varied from
.76 to .92 and for upper secondary schools from .72 to 92.
91
