1. Externalizing behavior, reading skills, and parental involvement: Patterns of influence in a sample of inner city elementary school children
Samantha Woo, MPH,
Deborah Gorman-Smith, PhD, Michael Schoeny, PhD,
Patrick Tolan, PhD, David Henry, PhD
The University of Illinois at Chicago
Institute for Juvenile Research

2. Background
Entrance into elementary school is an important developmental transition (Entwisle et al., 1988)
Significant challenges to child’s family structure and coping mechanisms (Entwisle et al., 1988)
Children maladjusted to new school environments are at higher risk of problem behaviors (Tolan, Gorman-Smith, & Henry, 2004)
Developmental Transition:
Decreased parental supervision
Increased interaction with authority figures other than parents
New social roles and expectations
Challenges to family structure and coping mechanisms:
Stronger negative effects of school entry on children at increased risk before entering elementary school
Children entering schools in the inner city may face a more difficult transition
School’s resource constraints
Schools may have lower learning expectations for the students
Parents may not have been particularly successful in school or had positive experiences as students

3. Externalizing Behaviors
More stable than internalizing behaviors (e.g., withdrawal, dysphoria, anxiety) (Hinshaw, 1992)
Characterized by hostile and aggressive behavior, impulsivity, hyperactivity, defiance, and concentration problems (Hinshaw, 1992; Deater-Deckard et al., 1998)
Often lead to antisocial behavior, immature social relationships, substance abuse, delinquency, criminality (Deater-Deckard et al., 1998)
Problem behaviors, specifically externalizing behaviors…
More stable:
Worse prognosis and resistance to most forms of intervention
Many different dimensions of externalizing behavior, which have in common symptoms that are disruptive to others
Leads to…
Immature and unrewarding relationships with peers and adults
Delinquency in childhood and adolescence
Criminality in adulthood

4. Reading Skills
Early experiences in the classroom can frame future academic pursuits (Tolan, Gorman-Smith, & Henry, 2004)
Illiteracy holds consequences for personal, professional, and emotional adjustment of the individual (Spreen, 1986)
Reading problems are stable and can incur high societal costs without early intervention (Smart, Sanson, & Prior, 1996)
Poor reading skills, in particular, may affect academic achievement, as a whole
Limited social opportunities
Inhibited from pursuing the most lucrative professions
Reading problems often persist into adulthood once established

5. Parental Involvement in School
Parents raising children in the inner city face increased levels of unique challenges (Tolan, Gorman-Smith, & Henry, 2004)
First grade children form expectations about school and value themselves based on evaluations by significant others (Entwisle et al., 1988)
Positive parental involvement promotes competence and the internalization of parental values (Beyers et al., 2003)
Parents face significant neighborhood problems, such as violence
Multiple jobs
Poor social support
However, a child’s entry into school remains as an important developmental transition
Positive parental involvement has a protective influence on conduct problems

6. Patterns of Influence Externalizing Reading Behavior Parent
The interaction of reading skills, parental involvement in school, and externalizing behavior suggest multiple patterns of influence
This model suggests the existence of unidirectional or bidirectional relations between each of the variables
Explanation of Model:
Poor reading skills precede (or exacerbate) the emergence of externalizing behaviors. For example, frustration, lowered self-image, demoralization, lack of school attachment
Behavioral problems precede reading problems. Interferes with proper classroom behavior and learning.
Considerable evidence that Behavior Problems predict poor reading skills
Mixed support for the hypothesis that poor reading skills predict BPs
Low positive parental involvement consistently predicts externalizing behavior
Limited discussion of how BPs affect Parental Involvement
Strong evidence of the effects of Parental Involvement on Reading Skills (but specific to parental involvement in reading, not necessarily in school)
Limited research on whether reading skills predict parental involvement.
These 6 patterns of influence are the focus of this study
Parent
Involvement

7. SAFEChildren Preventive intervention with two components
Multiple-family groups
Phonics-based reading tutoring program
22 weeks of intervention begun at start of first grade
This study uses data from a randomized controlled trial of a family-focused, preventive intervention designed to address risk factors which lead to behavioral problems
Family Groups:
*Parenting Skills
*Family Relationships
Managing challenges to the family
Increasing social support among parents
*Parental involvement in school
Tutoring:
*Developed as part of the Fast Track intervention
Basic skills of reading
Understanding letter-sound relationships
Blending sounds into words

8. Methods
Research participants: 424 families from 7 inner city Chicago schools
42.5% African American
57.5% Latino
51% male
49% female
40% in single-parent households
59% annual family income below $20,000
85% annual family income below $30,000
The sample population consisted of 424 low income African American and Latino families of children entering 1st grade.
The sample was taken from 7 Chicago public schools located in the inner city.
Inner city refers to a set of urban poor communities, characterized by extreme concentration of poverty, high rates of crime, high rates of multiple social problems such as early pregnancy and school dropout, and low rates of home ownership or business investment.

9. Methods Five waves of data collection:
Wave 1: End of kindergarten (Pretest 1)
Wave 2: Beginning of first grade (Pretest 2)
Wave 3: Midintervention (Only teacher data)*
Wave 4: End of first grade (Posttest)
Wave 5: 6 months later (Follow-up)
*Wave 3 is excluded in this analysis
Data were collected through individual interviews conducted with the primary caregiver, teachers, and children.
5 waves of data collection over a 24 month period
Due to financial constraints, data in Wave 3 were only collected from teachers. Therefore only data from Waves 1, 2, 4, and 5 were used for this analysis

10. Measures Woodcock Diagnostic Reading Battery
Measures reading achievement
Teacher/Parent Observations of Classroom Adaptation
Measures child’s externalizing behaviors
Fast Track Parent Involvement Scales (Parent and Teacher Forms)
Measures parental involvement in school
Various measures were used to assess specific constructs of school functioning, behavior, and parenting and family relationships.
4 subscales of the Woodcock Diagnostic Reading Battery
A composite of the TOCA/POCA measured three aspects of externalizing behavior: aggression, hyperactivity, and concentration
PINV Scales were administered to both teachers and parents, and measured parental involvement in school

11. Analytic Model – Simultaneous Latent Growth Curve1 1 1 1 2 3 1
Intercept A
Slope A
Our analytic model centered around the comparison of growth trajectories of specific intervention targets among participants, using four data points.
The specific intervention targets analyzed were reading skills, parental involvement in school, and externalizing behavior.
Using this model, we were able to compare relative levels at baseline, as well as changes in these levels over time, and compare these to each other.
Intercept is defined as the initial level of a variable.
Slope is defined as the change in levels over time of a variable.
Pathways from the Intercept of one variable to the Intercept of another variable signify the relation between the initial levels of the two variables.
Pathways from the Slope of one variable to the Slope of another variable signify the relation between the changes in the two variables.
Pathways from the Intercept of one variable to the Slope of that same variable signify the relation between the initial level and the change in levels of that variable
Pathways from the Intercept of one variable to the Slope of a different variable signify the relation between the initial level of the first variable and the change in the second variable.
This last pathway is the relation that we hope to see.
Intercept B
Slope B 1 1 1 1 1 2 3 B1 B2 B4 B5

12. Analytic Model – Alternative Model1 1 1 1 2 3 1
Intercept A
Slope A
Intervention
Controlled for the intervention effects in an alternative model and found that the results were not biased by these effects
Intercept B
Slope B 1 1 1 1 1 2 3 B1 B2 B4 B5

13. Analytic Model – Two Group Model
Intercept A
Slope A
Intercept B
Slope B 1 3 2 A1 A2 A4 A5 B1 B2 B4 B5
Intercept A
Slope A
Intercept B
Slope B 1 3 2 A1 A2 A4 A5 B1 B2 B4 B5
Tested a two group model, to see if there was any moderation for the treatment condition, and found limited moderation effects.
Gender: Negative relation between parental involvement in school at baseline and change in externalizing behavior over time significant in Tx group only
Ethnicity: African American group showed a negative relation between parental involvement at baseline and change in externalizing behavior over time. Latino group showed no significant relation.

14. Reading Skills and Externalizing Behavior
-0.13
Reading
Intercept
Reading
Slope
-0.13
-0.09
SIGNIFICANT:
Significant negative relation between externalizing behavior at a point in time and change in reading skills over time. High levels of externalizing behavior early on related to lower reading skills over time.
Significant negative relation between reading skills at a point in time and change in reading skills over time. Weaker readers early on naturally have more room to improve and therefore show greater changes over time
Significant negative link between reading skills and externalizing behavior at early on. Stronger reading skills linked to less externalizing behavior
Contrary to the literature, reading skills early on did not predict change in externalizing behavior over time.
Externalizing
Intercept
Externalizing
Slope

15. Parental Involvement in School and Externalizing Behavior
Parent Inv
Intercept
Parent Inv
Slope
-0.12
-0.09
SIGNIFICANT:
Significant negative relation between parental involvement at a point in time and change in externalizing behavior. High parental involvement in the beginning is related to decreases in externalizing behavior over time.
Significant negative relation between externalizing behavior at a point in time and change in parental involvement over time. High levels of externalizing behavior at any given point in time is related to decreases in parental involvement over time.
Externalizing
Intercept
Externalizing
Slope

16. Parental Involvement in School and Reading Skills
Parent Inv
Intercept
Parent Inv
Slope
No significant relations between parental involvement and reading skills.
Reading
Intercept
Reading
Slope

17. Summary
Evidence of unidirectional relation between reading skills and externalizing behaviors
Evidence of bidirectional relation between parental involvement in school and externalizing behaviors
No relation between parental involvement in school and reading skills
Unidirectional:
Early externalizing behaviors predict later change in reading skills
Bidirectional:
Early parental involvement predict later change in externalizing behaviors
Early externalizing behaviors predict later change in parental involvement

18. Conclusions
A focus on reducing problem behaviors holds implications for improving reading skills and increasing parental involvement in school
Interventions should consider the bidirectional relation between parental involvement in school and externalizing behavior
More intervention research is needed to determine clear causal pathways between variables
While parental involvement in reading may predict higher reading skills, parental involvement in school seems to have no significant effect
Results generalizable to urban minority samples, but not all populations and settings
In addition, since our intervention focused on affecting change in multiple variables, causal relations could not be determined from our data.
Despite the significant patterns of influence found in this study, more research is needed to clarify the strength and direction of pathways between reading skills, parental involvement, and externalizing behavior

19. Thank you!