The Power of Schools to Improve Outcomes for Teens
Robert Wm. Blum, MD, MPH, PhDWilliam H. Gates Sr. Professor and Chair
Department of Population, Family and Reproductive Health
Johns Hopkins Bloomberg School of Public Health
Prepared for:
Smith Anniversary Conference
University of Arizona
21 October 2008
Improving the Odds
Young people who are connected to school not
only do better academically but are less involved with
every risk behavior studied.
Methods
n The National Longitudinal Study of Adolescent Health
s A stratified random sample of 80 high schools with primary feeder schools
s N=134 schools (127 participated in school survey)
s N=71,515 students in 7th through 12th grade
s N=127 school administrator surveys
The Sample
Add-Health Sample Add-Health Sample DesignDesign 129 Schools
(79%)90,118(75.6%)
• Saturation• High income African American• Twins• Disability
N=15,243
In-SchoolSample(1994)
In-HomeSamples(1995)
Special
CoreIn-school survey &
school rosters
Wave 1(1994-5)
TeenN=14,738
Wave 2(1996)
TeenN=12,105(79.5%)
ParentN=10,471(86.5%)
Young AdultsN=15,197
Wave 3(2001-2)
AdultsWave 4(2007-8)
PartnersN=1507
Methods
n Ethnicitys Black 15.0%s Latino 12.2%s White 72.8%
n Genders Female 51.8%s Male 49.2%
The Sample (continued)
Substance Use
-0.5
0
0.5
1
1.5
Not at All Very Little Somewhat Quite a Bit A Lot
AlcoholCigarettesMarijuana
Levels of connectedness
Students who feel connected to school are less likely to use substances
Frequency of Use:
Lev
el o
f S
ub
sta
nce
Use
(S
D U
nit
s)
Emotional Distress
-0.5
0
0.5
1
1.5
Not at All Very Little Somewhat Quite a Bit A Lot
EmotionalDistress
Suicide
Students who feel connected to school experience less emotional distress
Levels of connectedness
Lev
el o
f E
mo
tio
nal
Dis
tres
s (
SD
Un
its)
Violence or Deviant Behavior
-0.5
0
0.5
1
Not at All Very Little Somewhat Quite a Bit A Lot
DeviantBehavior
Violence
Students who feel connected to school engage In less violent or deviant behavior
Levels of connectedness
Lev
el o
f V
iole
nce
or
De
via
nt
Beh
avio
r (S
D U
nit
s)
Pregnancy
12
14
16
18
20
22
24
Not at All Very Little Somewhat Quite a Bit A Lot
Students who feel connected to school are less likely to become pregnant
Levels of connectedness
Per
cent
eve
r P
regn
ant
Research Questions
n What contributes to teens feeling connected to school?
n Why do some adolescents feel attached to school and others don’t?
n What school characteristics predict connectedness?
n What impact do friendships have on a young person feeling connectedness to school?
Methods
n Size 42 to 5422 (average=642)
n Public 82.7%n Class size 10 to 39 (average=23)
n Locations Rural 18.6%s Suburban 59.6%s Urban 21.8%
School Characteristics
Methods
Master’s degree 42%
1st year teachers 9%
Teacher Characteristics
n Out of school suspension for smoking
n Out of school suspension for cheating
n 10 item scale (alpha=.78) for disciplinary policy of school for:
ScalesDiscipline Policies
Possessing alcoholDrinking alcoholPossessing an illegal substanceUsing an illegal substance
Destroying school propertyVerbally abusing a teacherFightingInjuring a teacherInjuring another studentCarrying a weapon
n Getting along with teachersn Getting along with other studentsn Paying attention in schooln Getting homework done
ScalesClassroom Management8 items: alpha = .83
n I feel close to people at this schooln I am happy to be at this schooln I feel like I am part of this schooln The teachers at this school treat
students fairlyn I feel safe in this school
Scales
School Connectedness
n School size mattered…classroom size did not
n School type is not associated with connectedness
…public, private, parochialn Location of school is not associated
with connectedness…urban, suburban, rural
ResultsFactors Associated with School Connectedness
THE SCHOOL
n No single school policy was associated with connectedness …
n … A climate of harsh discipline is associated with lower school connectedness
n The directionality of the relationship cannot be deduced from the present study
ResultsFactors Associated with School Connectedness
SCHOOL POLICIES
n Teacher experience was not associated with connectedness.
n Having a master’s degree was not associated with connectedness.
ResultsFactors Associated with School Connectedness
TEACHERS
The single strongest association with
connectedness was school climate
ResultsFactors Associated with School Connectedness
SCHOOL CLIMATE & CLASSROOM MANAGEMENT
Initiated Cigarette Use Predicted Percent at Three Levels of
Teacher Support
16.5
14.3
12.2
5.33.8
2.80
2
4
6
8
10
12
14
16
18
None to Occasional None to Regular
- 1 s.d.Mean+ 1 s.d.
Multinomial logit models adjusted for social belonging, race/ethnicity, income, gender, family structure, emotional distress, relationship with parents, hx of peer suicide, hx of family suicide
Initiated Getting DrunkPredicted Percent Three Levels of
Teacher Support
0
2
4
6
8
10
12
14
16
18
None to Occasional None to Regular
- 1 s.d.Mean+ 1 s.d.6.2
11.3
9.5
4.53.2
13.4
Multinomial logit models adjusted for social belonging, race/ethnicity, income, gender, family structure, emotional distress, relationship with parents, hx of peer suicide, hx of family suicide
Initiated Marijuana Use Predicted Percent at Three Levels of
Teacher Support
0
2
4
6
8
10
12
14
16
18
None to Occasional None to Regular
- 1 s.d.Mean+ 1 s.d.
4.4
6.34.7
3.4 2.6
8.3
Multinomial logit models adjusted for social belonging, race/ethnicity, income, gender, family structure, emotional distress, relationship with parents, hx of peer suicide, hx of family suicide
Seriously Considered or Attempted Suicide
Predicted Percent at Three Levels of Teacher Support
0
2
4
6
8
10
12
14
16
18
None to Ideation None to Attempt
- 1 s.d.Mean+ 1 s.d.
2.2
5.0 4.5
1.7 1.3
5.5
Multinomial logit models adjusted for social belonging, race/ethnicity, income, gender, family structure, emotional distress, relationship with parents, hx of peer suicide, hx of family suicide
Initiated Violence Predicted Percent at Three Levels of
Teacher Support
0
2
4
6
8
10
12
14
16
18
- 1 s.d.Mean+ 1 s.d.7.5
6.04.7
Multinomial logit models adjusted for social belonging, race/ethnicity, income, gender, family structure, emotional distress, relationship with parents, hx of peer suicide, hx of family suicide
Family Engagement and School leaving: what are the effects over time?
CONCEPTUAL MODEL
Wave 1Early Adolescence:Ages 12-14
Wave 2Early to Middle Adolescence: Ages 13-16
Wave 3Emerging Adulthood:Ages 18-21
Family Closeness
Parental Behavioral Control
Skipping School Dropping out of School
Measures: ParentingFamily closeness in early adolescence
5-item measure of youth report that s/he has fun with family; family pays attention to youth; cares for youth; understands youth; how much youth felt like leaving home (α = .76)
Response categories: 1 = not at all; 5 = very much
Mean= 4.2 (.02)
Measures: ParentingParental behavioral control in early adolescence
Seven items measuring whether or not youth decides: what wears; who hangs out with; TV amount & programs watch; time home on weekends; time to bed during week; what eats.
Response categories: 0 = no; 1 = yes. Reverse coded.
Measures: ParentingParental behavioral control
3 Classes:
37% “High” - Parent decides many things
(reference group)
30% “Moderate” – Parent decides curfews and inside activities (eat, TV programs, time to bed)
32% “Low” – Parent decides curfews
HOW WERE FAMILY CLOSENESS AND PARENTAL BEHAVIORAL CONTROL OPERATIONALIZED
My Family Cares for Me
Parent decides what child eats, who child hangs with and what child wears
Parent decides what TV programs child watches and time child goes to bed on week-night
Parent decides amount of TV child watches and what time child comes home on weekend
My Family Pays Attention to Me
I Have Fun with My Family
My Family Understands MeFamily Closeness
Parental Behavioral
Control
.42***
.72***
.72***
.78***
.57***
.46***
.43***
.11**
Measures: Youth OutcomesSchool engagement
Time 1 & Time 2: Skipped school past year
Time 3: Highest education completed 10 years or less
Proportion youth skipping school:
14% (T1) & 18% (T2)
Proportion youth dropping out: 7%
0
0.5
1
1.5
2
2.5
3
3.5
+1 SD 0 -1 SD
Od
ds
of
Sc
ho
ol
Dro
po
ut
Family Closeness
Odds of School Dropout by Family Closeness and Skipping School during Early
Adolescence
No Early Skipping School
Early School Leaving
SCHOOL DROPOUT: MALE
Family ClosenessAges 12-14
Parental Behavioral Control
Ages 12-14
Skipped School in the Past Year
Ages 13-16
School DropoutAges 18-21
-.10*
-.01 (.05)
.16***
.01 (.003)
.06
SCHOOL DROPOUT: FEMALE
Family ClosenessAges 12-14
Parental Behavioral Control
Ages 12-14
Skipped School in the Past Year
Ages 13-16
School DropoutAges 18-21
-.19***
-.08 (-.12**)
.21***
.10 (.01)
.004
ConclusionThe present report suggests that
school climate and adult
connections as well as
family connectedness may
impact
health as well as educational outcomes.
1. Be a model of respectful, cooperative, positive behavior in your everyday interactions.
2. Participate in school events.
3. Show interest. Be involved in your child’s academic activities.
4. Maintain regular contact with your
child’s teacher.
ParentsTen Strategies that Foster Connections to School
5. Monitor your child’s homework completion and work with him or her on homework assignments that involve family participation.
6. Be present when things go wrong.
7. Meet your child’s friend, and their parents.
ParentsTen Strategies that Foster Connections to School
8. Ask school leaders what you can do to support them.
9. Volunteer at school.
10. Nominate effective school leaders for local awards.
ParentsTen Strategies that Foster Connections to School
1. Help students get to know each other’s (and your) strengths.
2. Involve students in planning, problem solving, identifying issues and assessing curriculum in the classroom.
3. Promote cooperation over competition. Post everyone’s best work. Offer opportunities for the class to work together to help everyone achieve their level of excellence.
TeachersTen Strategies that Foster Connections to School
4. Build a strong relationship with each student.
5. Convey attentiveness to students and excitement about learning through nonverbal gestures.
6. Involve all students in chores and responsibilities in the classroom.
7. Integrate concepts of discipline and respect for classmates through instruction.
TeachersTen Strategies that Foster Connections to School
8. Give students more say in what they will learn.
9. Involve students in developing the criteria by which their work will be assessed and provide guidelines so they clearly understand what’s expected of them.
10. Use first person plural (we, us, let’s) when presenting classroom activities.
TeachersTen Strategies that Foster Connections to School
1. Brainstorm with students, faculty, staff and parents simple changes that could make school a more pleasant place to be.
2. Create policies that are based on student, family and neighborhood strengths and assets.
3. Turn mistakes into learning opportunities rather than failures meriting punishment.
AdministratorsTen Strategies that Foster Connections to School
4. Acknowledge and honor accomplishments and all types of competencies (such as helpfulness, good citizenship, most improved performance, volunteerism, participation in decision making, and cessation of negative behavior).
5. Set high standards and challenge students to meet them.
6. Reinforce explicit expectations for positive behavior and academic success.
AdministratorsTen Strategies that Foster Connections to School
7. Encourage highly interactive teaching strategies.
8. Create a welcoming environment for all who come to the school.
9. Invite family and community members to take active and regular roles in the daily operation of the school.
10. Create a common vision of success and keep it visible.
AdministratorsTen Strategies that Foster Connections to School
Top Related