Well-being of Adolescents in
Military Families: Examining
the Intersections of Resilience
and Vulnerability
Chairs: Jay A. Mancini, Mallory Lucier-Greer,
& Amy Laura Arnold
Discussant: Angela Huebner
National Council on Family Relations
Annual Conference
San Antonio, TX
November 6, 2013
Data were drawn from a project funded by NIFA Award No.
2009-48680-06069, Jay A. Mancini, Principal Investigator.
Family Structure Effects on
Adolescent Well-Being
Amy Laura Arnold, The University of Georgia
Mallory Lucier-Greer, Florida State University
Jay A. Mancini, The University of Georgia
James L. Ford, The University of Georgia
K.A.S. Wickrama, The University of Georgia
Theoretical Underpinnings
• Levels of systems
• Process-Person-Context-Time model
• “Linked Lives”
– Parents’ relationship formations and children’s
adjustment intersect (Elder, 1998)
– Major life events seen as “turning points”
(Langenkamp & Krisco, 2008)
(Tudge, Mokrova, Hatfield, & Karnik, 2009)
Family Structure
• Parental relationship context
(Crosnoe & Cavanagh, 2010)
• Proxies resources that shape parents’
interactions with their children (Crosnoe & Cavanagh,
2010)
• Family disruptions and current family
environment influence youth well-being
(e.g., Cavanagh et al., 2006; Fomby & Sennott, 2013; Teachman et al., 2001)
Parents & Youth
• Parent-youth relationships  adolescents
adjustment (e.g., Conger, Ge, Elder, Lorenz, & Simons, 1994; Crosnoe et al., 2003;
Lamb, 2012; Melby et al., 2009; Steinberg, 2001; Whitbeck et al., 1997)
• Youth Self-efficacy: Cognitive perceptions of
self (Hughes, Galbraith, & White, 2010)
– Academic performance enhanced by greater selfefficacy (Zimmerman, Bandura, Martinez—Pons, 1992)
– Eroded sense of self fosters depressive symptoms
(Wickrama, Surjadi, Lorenz, Conger, & O’Neal, 2012)
Present Study
Few studies have:
• Made theoretical claims
(Cavanagh, 2008)
• Examined parent-youth connections as
processes linking family structure to youth
functioning
– Influence of youths’ cognitive traits
• Examined sub-population of military youth
– Little known about various family structures
within military context (Adler-Baeder, Pittman, & Taylor, 2008)
Research Hypotheses
Effects of family structure traced through two
paths:
Measures
• Family Structure
– Categorized by two biological parents
(reference category), stepfamily, single
parent, and “other family form” (Cavanagh, 2008; Mokrue
et al., 2011)
•
•
•
•
Biological mother & father (67.3%)
Stepfamily (17.3% )
Single parent (10.9%)
Other family form (4.4%)
• Parent-Youth Connections
– Latent factor indicated by four items from Adolescent
Coping Scale (ACOPE; Patterson & McCubbin, 1987; α = .73)
Measures
• Initiative
– Latent factor indicated by three items from General SelfEfficacy Scale (GSE; Bosscher & Smit, 1998; α = .71)
• Depressive symptoms:
– Total sum score (20 items) from Center for Epidemiologic
Studies Depression Scale for Children (CES-DC; Faulstich
et al., 1986; α = .90)
• Academic performance
– Self-reported grades
• Controls: gender, age, race
Methods
• Analyses
– Descriptives
– Zero-order correlations
– Structural equation modeling
•
•
•
•
Amos 20
Full-Information Maximum Likelihood (FIML)
Incremental nested model
Fit indices (Kline, 2005)
– Chi-square test statistic
– Comparative fit index
– Root Mean Square Error of Approximation
Findings
-.07*
Academic
Academic
Performance
Academic
Performance
(R22 = .14)
Performance
(R = .14)
.10)
(R2 = .04)
.19**
-.06*
-.07*
-.08*
-.09
-.10**
Stepfamily
Stepfamily
Stepfamily
-.08**
-.08**
-.13**
Stepfamily
-.16**
-.16**
-.16**
-.16**
Single
Single
-.10**
-.10**
Parent
Single
Parent
-.10**
Parent
-.08*
-.08*
-.08*
-.04
-.08*
Other
Other
Family
Family
Other
Other
Family
Family
Forms
Forms
Forms
Forms
-.02
-.02
-.11**
.19**
.24**
.23**
.23**
-.13**
-.13*
-.10**
-.10**
Parent-Youth
Parent-Youth
Parent-Youth .16**
Connections
.16**
Connections
Connections
2
(R = .04)2
(R2 = .04)
(R = .04)
-.04
.08**
Youth
Youth
Initiative
Initiative
(R2 = .07)
(R2 = .06)
-.09**
-.14***
-.21**
.07*
.12**
-.08*
-.08*
-.08*
.09**
.09**
.08**
-.27**
-.27**
-.30**
-.02
.06*
.06*
.03
-.25**
.03
.03 -.25**
.06*
Figure
5. Family
Family
Structure
Model
Figure
6. Family
Structure
Model
4 23
Figure
4.
Structure
Model
Note:
*
p
<
.05,
**
p
<
.01;
Standardized
betas
presented.
The
reference
category
two
Note:
*
p
<
.05,
**
p
<
.01;
Standardized
betas
presented.
TheThe
reference
category
is two
Figure
3. Family
Structure
Model
1
Note:
* p < .05,
** p < .01;
Standardized
betas
presented.
reference
category
isis two
biological
parents,
and
models
controlled
for
gender,
age,
and
race.
Factor
loadings
ranged
biological
models
controlled
for gender,
age,age,
andand
race.
Factor
loadings
Note:
* p <parents,
.05,
** and
p <and
.01;models
Standardized
betas
presented.
The
reference
category
is ranged
tworanged
biological
parents,
controlled
for
gender,
race.
Factor
loadings
from
.57-.75.
from
.57-.75.
biological
parents, and models controlled for gender, age, and race.
from .60-.72.
Depressive
Depressive
Depressive
Symptoms
Depressive
Symptoms
Symptoms
2
(R
.23)
2 2==
Symptoms
(R(R
=.17)
.23)
(R2 = .08)
2 =149.27
χ2χ=
262.74
150.71(df
(df==53),
61),
56),p p<<.01
.01
2
2 /df
=
388.88
(df = 69), p < .01
χ χ/df
==2.82
4.31
2.69
χ2CFI
/df==
5.64
CFI
=.95
.90
.95
CFI
=
.84
RMSEA
RMSEA==.04
.06
.04
RMSEA = .07
Model Comparisons
Table 2. Family Structure Model Comparisons
Model Fit Indexa
χ 2(df)
CFI
RMSEA
Δχ 2(df)b
Model 1 (Baseline model)
388.88*** (69)
.84
.07
Model 2 (Parent-youth connections
included)
262.74*** (61)
.90
.06
126.14 (8)
Model 3 (Theoretical model)
150.71*** (56)
.95
.04
112.03 (5)
Model 4 (Fully Recursive)
149.27*** (53)
.95
.04
1.44 (3)
Note. a Values for each model. b Critical value of χ2(3) = 7.82; χ2(5) = 11.07; χ2(8) = 15.51
(Bollen, 1989)
Summary of Findings
• Full or partial support for hypotheses
– H1: Living in step- and single family lower
academic performance and single and otherfamily higher levels of depressive symptoms.
– H1a: Parent-youth connections partially mediates
relationship between family structure and
outcomes
– H2: Greater parent-youth connections  higher
levels of youths’ self-efficacy, which in turn 
academic performance and depressive symptoms