Predicting Youth Engagement: The Role of Initiating and Sustaining Factors Linda Rose-Krasnor1, Kelly Campbell1, Lisa Loiselle2, Mark Pancer3 , Michael Busseri1 The Centre of Excellence for Youth Engagement 1Brock University, 2University of Waterloo,

1
Predicting Youth Engagement The Role of
Initiating and Sustaining FactorsLinda
Rose-Krasnor1, Kelly Campbell1, Lisa Loiselle2,
Mark Pancer3 , Michael Busseri1 The Centre of
Excellence for Youth Engagement1Brock
University, 2University of Waterloo, 3Wilfrid
Laurier University

• Measures
• Personal values (Pratt, Pancer, Hunsberger,
  submitted).
•   Social responsibility (Pancer, Pratt,
  Hunsberger, 2000).
•   Religiosity/spirituality. (perceived
  importance for self and family).
• Sense of neighborhood (National Longitudinal
  Study of Children and Youth).
•   Social support. (Cutrona Russell, 1987).
•   Support for engagement.
•   Attitudes towards engagement. (negative
  beliefs, self-serving motivations and perceived
  barriers to involvement)
•   Youth efficacy (beliefs about youths ability
  to change society).
•   Youth engagement (Pancer, et al., submitted).
  Principle components analyses (oblique rotation)
• yielded 3 moderately correlated factors
  general/ volunteering, and local/community
  involvement.
• Perceived personal skills. Principle components
  analysis (oblique rotation) yielded 3 moderately
  correlated factors
• leadership, teamwork, communication.
• Self-esteem. (adapted from Rosenberg,1965).
• Optimism (Goodman, Knight, Durant, 1997).
• Coping with stress (Cohen, Kamarc,
  Mermelstein, 1983).
• Social interaction frequency with friends,
  parents, and other important adults (Pancer, et
  al., 2000).

Abstract Youth engagement has been associated
with a wide variety of positive outcomes. It is
important, therefore, to better understand
factors that may facilitate and maintain
involvement. With survey data from 192 youth, we
assessed latent constructs of initiating,
sustaining, and engagement factors. Personal
values and social responsibility had the largest
loadings on the initiating factor support for
involvement had the largest loading on the
sustaining factor and general engagement had the
largest loading on the engagement factor. A test
of the latent regression model indicated that
initiating and sustaining factors explained 39
of the variance in the latent engagement factor.
Initiating and sustaining factors were strongly
correlated, however, and the unique contribution
of the initiating factor to engagement was small.
Introduction Youth engagement is the meaningful
and sustained involvement of a young person in an
activity, focusing outside the self. Engagement
has been associated with a variety of positive
outcomes, including self-system qualities (e.g.,
self-esteem), skills (e.g., leadership), and
positive social interactions. Given engagements
potential benefits, it is important to increase
our understanding of factors that may help youth
become involved, as well as those that may
sustain involvement. Our investigations of youth
engagement are guided by the framework presented
in Fig. 1. In this model, youth engagement
results from initiating factors such as altruism.
Other influences, such as social support and
sense of efficacy, contribute to the maintenance
of youth involvement over time. Engagement, in
turn, is expected to lead to various positive
outcomes. In addition, our model also allows the
possibility that initiating and sustaining
factors may be related to each other, as well as
having direct links to outcomes.   We conducted a
partial test of our model using structural
equation modelling. We first assessed a
measurement model that included latent variables
for initiating, sustaining, and engagement
factors. Next, we tested a latent regression
model assessing links between engagement and both
initiating and sustaining factors.
Analyses Three latent factors were included as
structural model components (1) initiating
(measured by personal values, social
responsibility, religiosity/ spirituality, and
sense of neighborhood) (2) sustaining (measured
by social support, attitudes towards involvement,
support for involvement, and youth efficacy) and
(3) engagement (measured by general, political,
and local/community engagement scores). First,
we examined the relationships among the latent
initiating, sustaining, and engagement factors in
a structural equation model (the measurement
model) by specifying correlated paths between
each pair of factors. Second, we regressed the
latent engagement factor onto the latent
initiating and sustaining factors, while allowing
the latent predictors to be correlated (the
latent regression model).
Figure 2. Measurement Model Standardized
results are shown. Large ovals represent latent
factors. Rectangles represent measured variables.
Small ovals represent unique variance. All paths
are significant at p lt .05. (f) indicates path
fixed for model identification purposes. Numbers
in parentheses indicate proportion of variance
explained in each measure.
Figure 3. Latent Regression Model Large ovals
represent latent factors. Standardized results
are shown. Number in parentheses indicates
proportion of variance explained in the latent
engagement factor. All paths are significant at p
lt .05 except .
Initiating
-.02
Engagement (.39)
.77
.64
Sustaining
Results Relationships Among Initiating,
Sustaining, and Engagement Factors The fit for
the base model specifying correlational paths
between pairs of latent factors was ?2 (41)
99.40, ?2/df 2.42, CFI .86, RMSEA .09.
Based on modification indices, 2 correlated paths
between pairs of unique variance terms were added
sequentially, resulting in significant
improvement. The revised model provided a good
fit to the data ?2 (38) 74.93, ?2/df 1.92,
CFI .91, RMSEA .07. Latent factors were
moderately to strongly correlated (see Fig. 2).
The largest loadings on the initiating factor
were personal values and social responsibility.
For the sustaining factor, support for
involvement had the strongest factor loading.
Each of the 3 involvement scales showed strong
loadings on engagement. Predicting Engagement
from Initiating and Sustaining Factors Overall,
39 of the variance in the latent engagement was
explained by initiating and sustaining factors.
Of the 2 latent predictors, the sustaining factor
showed a strong positive relationship. In
contrast, although strongly correlated with the
sustaining factor, the unique contribution of the
initiating factor to the prediction of engagement
was small and statistically non-significant (see
Fig. 3.)
Fig. 1. Model of Youth Engagement

• Conclusions
• With survey data from 192 youth, we used
  structural equation modeling to assess latent
  constructs of initiating, sustaining, and
  engagement factors.
• Personal values and social responsibility had the
  largest loadings on the initiating factor
  support for involvement had the largest loading
  on the sustaining factor and general engagement
  had the largest loading on the engagement factor.
  
• There were few significant associations found
  between our three latent constructs and age or
  gender.
• The structural equation modeling approach allowed
  a simultaneous assessment of the relations among
  initiating, sustaining and engagement factors.
• A test of the latent regression model indicated
  that initiating and sustaining factors explained
  39 of the variance in the latent engagement
  factor.
• Initiating and sustaining factors were strongly
  correlated, however. This may partially explain
  our finding that the unique contribution of
  initiating factor to engagement was small and
  non-significant. Other potential initiating
  variables (e.g., parental modeling of engagement,
  teacher invitation to participate), which were
  not measured in this study, may show stronger
  unique relations to engagement.
• In future research, it will be important to look
  at the quality of engagement experience, in
  addition to the frequency of involvement.

Initiating
Engagement
Outcomes
Sustaining
Participants and Procedures A convenience
sample of 192 youth who applied to attend one of
three national youth conferences voluntarily
completed a 40-min. survey. The survey was
mailed to potential participants prior to the
conference and either returned by mail or
completed upon arrival. M age 16.64 yrs. (SD
1.03), ranging from 14 to 19 yrs 76 were
female.
The Centres of Excellence are a Health
Canada-funded program. The opinions expressed in
this publication do not necessarily reflect those
of Health Canada.