Page 81 - Ana Kozina and Nora Wiium, eds. ▪︎ Positive Youth Development in Contexts. Ljubljana: Educational Research Institute, 2021. Digital Library, Dissertationes (Scientific Monographs), 42.
P. 81
measuring positive youth development in slovenia
additional analyses were performed to examine the scales’ reliability and
validity (Scales, 2011). Recently, Syvertsen and colleagues (2019) undertook
a theoretical and empirical examination of DA which in turn led to certain
changes to the mentioned questionnaire (i.e. they removed some items or
moved them to a different asset). In the last decade, another DA question-
naire has emerged for use in emergency settings (i.e. the 13-item Em-DAP;
Scales et al., 2015). In the European context, it is hard to find studies that re-
lied on the Developmental Assets Profile, which is used in the present study,
even though a recent study conducted in Spain revealed the good factorial
validity of the external and internal assets (Gomez-Baya et al., 2021).
Measurement invariance was considered for DA regarding grade and
gender identity in the USA (Syvertsen et al., 2019). Configural, metric and
scalar invariance were established for Positive identity, Positive values and a
newly invented asset category called Mattering and belonging across grade.
Further, configural and metric invariances were shown for Social compe-
tencies, Support, and Boundaries (a newly invented asset), and partial met-
ric invariance was established for Academic engagement (a newly invented
construct that includes Commitment to learning). Moreover, at least par-
tial scalar invariances were established for gender identity for Positive val-
ues, Support, Mattering and Belonging (a newly developed construct in this
study), Boundaries, Positive identity, and Social competencies. The metric
invariance of Constructive use of time for gender identity was not exam-
ined since the mentioned authors removed this asset as it was examined as
a collection of various extracurricular activities. Regardless, more research
is needed in this area.
Alternatively, the measurement properties of the 5Cs model have been
more thoroughly examined in the US context (e.g. Geldhof et al., 2014a).
The first measure of the 5Cs consisted of over 80 items and was understood
as a higher-order measure of PYD that comprised 5 first-order latent con-
structs (i.e. confidence, competence, character, caring, connection) (Lerner
et al., 2005). The 5Cs were found to be a robust construct throughout ear-
ly adolescence while scales differed for middle adolescents (Bowers et al.,
2010). Later, the scale was shortened to 34 items (see Geldhof et al., 2014b)
and applies to young people aged 10 to 18 years. The analyses suggested that
a model without a higher-order PYD construct fitted the data worse than
a model with 5 first-order latent constructs. Moreover, the bifactor mod-
el had the most appropriate fit, in which each item represents a lower-or-
der construct and a more general construct. All the aforementioned studies
81
additional analyses were performed to examine the scales’ reliability and
validity (Scales, 2011). Recently, Syvertsen and colleagues (2019) undertook
a theoretical and empirical examination of DA which in turn led to certain
changes to the mentioned questionnaire (i.e. they removed some items or
moved them to a different asset). In the last decade, another DA question-
naire has emerged for use in emergency settings (i.e. the 13-item Em-DAP;
Scales et al., 2015). In the European context, it is hard to find studies that re-
lied on the Developmental Assets Profile, which is used in the present study,
even though a recent study conducted in Spain revealed the good factorial
validity of the external and internal assets (Gomez-Baya et al., 2021).
Measurement invariance was considered for DA regarding grade and
gender identity in the USA (Syvertsen et al., 2019). Configural, metric and
scalar invariance were established for Positive identity, Positive values and a
newly invented asset category called Mattering and belonging across grade.
Further, configural and metric invariances were shown for Social compe-
tencies, Support, and Boundaries (a newly invented asset), and partial met-
ric invariance was established for Academic engagement (a newly invented
construct that includes Commitment to learning). Moreover, at least par-
tial scalar invariances were established for gender identity for Positive val-
ues, Support, Mattering and Belonging (a newly developed construct in this
study), Boundaries, Positive identity, and Social competencies. The metric
invariance of Constructive use of time for gender identity was not exam-
ined since the mentioned authors removed this asset as it was examined as
a collection of various extracurricular activities. Regardless, more research
is needed in this area.
Alternatively, the measurement properties of the 5Cs model have been
more thoroughly examined in the US context (e.g. Geldhof et al., 2014a).
The first measure of the 5Cs consisted of over 80 items and was understood
as a higher-order measure of PYD that comprised 5 first-order latent con-
structs (i.e. confidence, competence, character, caring, connection) (Lerner
et al., 2005). The 5Cs were found to be a robust construct throughout ear-
ly adolescence while scales differed for middle adolescents (Bowers et al.,
2010). Later, the scale was shortened to 34 items (see Geldhof et al., 2014b)
and applies to young people aged 10 to 18 years. The analyses suggested that
a model without a higher-order PYD construct fitted the data worse than
a model with 5 first-order latent constructs. Moreover, the bifactor mod-
el had the most appropriate fit, in which each item represents a lower-or-
der construct and a more general construct. All the aforementioned studies
81
