Changes in the Social Environment 1

Changes in the Social Environment 1


Running head: CHANGES IN THE SOCIAL ENVIRONMENT



Changes in the Social Environment and the School Success of Middle School Students:
A Longitudinal Analysis
Gary L. Bowen
Roderick A. Rose
Joelle D. Powers
The University of North Carolina at Chapel Hill
Elizabeth Glennie
Duke University





Authors’ Note: This article was prepared with grant support from the William T. Grant
Foundation, Grant Number 2520: Evaluating the Effectiveness of the School Success Profile
(SSP) Intervention Package on School-level Performance. Special appreciation is expressed to
Dr. Robert Granger, President, William T. Grant Foundation, for his encouragement and support
in completing the analysis. Appreciation is also expressed to the John S. and James L. Knight
Foundation for its ongoing support of the SSP since 1995. Dr. John Bare, formerly director of
planning and evaluation at the Knight Foundation and now vice president of strategic planning
and evaluation at the Arthur M. Blank Family Foundation, has been instrumental to the
continued development and testing of the SSP and its associated products.

---

Changes in the Social Environment 2

Abstract
In the context of a larger intervention designed to promote the functioning of 11 middle schools
as learning organizations, longitudinal data from 2,845 students are used to examine the degree
to which changes in the social environments of students are associated with changes in
adaptation outcomes associated with school success. Students completed the School Success
Profile in both fall 2004 and fall 2005, and the survey data were combined with administrative
data from the Education Research Data Center at Duke University for purposes of analysis.
Using hierarchical linear modeling to adjust for potential school-level effects, the results align
with earlier results from cross-sectional analysis to demonstrate variation in the influence from
changes in students perceptions toward social environment dimensions on changes in their
personal beliefs and well being and their school attitudes, behavior, and grades. Social
environment dimensions associated with the family domain exerted a greater number of effects
that met or exceeded the .09 threshold for interpreting standardized beta coefficients than those
dimensions associated with the neighborhood, school, or peer group. Other findings suggested
little between-school variability in the data, which calls into the question the need for a
multilevel framework for modeling the data. Implications of the finding for informing
intervention practice are discussed, as well as implications for further research.

Key words: school success, social environment, School Success Profile, middle school students,
intervention research

---

Changes in the Social Environment 3

Changes in the Social Environment and the School Success of Middle School Students:
A Longitudinal Analysis
A growing body of research suggests the importance of the social environment as a
context for school success. Students’ relationships and students’ experiences in their
neighborhood, school, family, and peer group have been shown to influence specific student-
adaptation outcomes associated with school success (Powers, Bowen, & Rose, 2005). Moreover,
each of these social environments may affect the individual differently over time (Jessor, 1993).
School success is vital to youths’ development into competent and productive members of
society. Consequently, it is important to identify those features of the social environment that can
be leveraged to promote school success.
Using hierarchical linear modeling (HLM), the present research examines how changes
over time in the perceptions of middle school students toward their social environments
influence changes in adaptation outcomes associated with school success. These outcomes range
from students’ personal beliefs and well-being to students’ school-related attitudes and behavior
and their self-reported academic performance.
All student-level data are self-reported and are derived from the administration of the
School Success Profile (SSP) in 11 middle schools participating in a larger evaluation. Informed
by the eco-interactional developmental perspective (Richman, Bowen, & Woolley, 2004), the
SSP is an ecologically oriented survey that focuses on students within the context of
neighborhood, school, family, and friends. The focus within these social environments includes
attention both to people, in the form of interpersonal relationships, and to places, in the form of
safety and satisfaction.

---

Changes in the Social Environment 4

The focus on middle school students has particular merit when examining the relationship
between the social environment and school success. Middle school is an important transition
period and determinant of the school success trajectory for adolescents (Eccles et al., 1993).
Many students find the change from primary school to secondary school particularly difficult,
and this transition can result in decreased academic performance (Gutman & Midgley, 2000).
Patterns of both academic success and emotional functioning experienced during early
adolescence have been shown to carry over into high school (Roeser, Eccles, & Freedman-Doan,
1999), which makes the middle school years a critical design and implementation period for
social interventions.
Results of the present investigation have implications for the design and implementation
of social interventions to increase the school success of middle school students. The first step in
evidence-based practice is to identify the correlates of school-related outcomes that will produce
the greatest leverage in achieving the desired outcomes. Social interventions can then be
designed or implemented to influence those correlates that are the most consequential for
preferred results.
An Earlier Analysis
The present investigation parallels, yet expands, an earlier analysis by Powers et al.
(2005). In contrast to that earlier analysis, which included only a single time observation, the
present analysis utilizes longitudinal data. Also unlike the earlier analysis, which included
students from a broad array of educational sites and programs (351 middle school and high
school sites), the present analysis sample focused on a population of students from only 11
middle schools—sites that were participating in a universal social intervention to promote the
functioning of the schools as learning organizations (Bowen, 2004).

---

Changes in the Social Environment 5

Consequently, as a prelude to the main analysis of the result of changes in the social
environments of students, we first examine whether the data exhibit sufficient between-school
variability to warrant the use of multilevel modeling methods. We also test for variability in the
social-environment-change scores of students, using baseline group-level data about the school
as a learning organization, provided by school employees. This analysis, which includes controls
for potential student demographic profile influences, is consistent with the theory of change
model that informed the design of the social intervention.
The Context
Rates and Consequences of School Failure
Schools are among the most salient contexts for youths’ social, emotional, and academic
development (Roeser, Eccles, & Sameroff, 2000), but the rates of school failure in the United
States are alarming (Richman et al., 2004). A report by the National Center for Education
Statistics (NCES) (2005) estimates that approximately 10% of 16- to 24-year-olds were school
dropouts in 2003 (i.e., were not enrolled in school and had not completed a high school
program). A report for the Bill and Melinda Gates Foundation describes a more dismal picture
and estimates that one-third of all public high school students in America do not graduate with
their peers (Bridgeland, Dilulio, & Morison, 2006). Hispanic and Black students experience
lower levels of school performance, slower progression through school, and higher dropout rates,
and also drop out earlier than either their White or Asian counterparts (Bali & Alvarez, 2004;
Education Commission of the States, 2004; Reyes, Gillock, Kobus, & Sanchez, 2000). A recent
report on the achievement gap by the Education Commission of the States (2004) confirms that
the average Black or Hispanic student performed at the same level as a White student

---

Changes in the Social Environment 6

functioning in the lowest quartile of performance for White students. This gap in standardized
achievement test scores has been demonstrated repeatedly (Bali & Alvarez, 2004).
Research has established that students who drop out are at higher risk for numerous
negative outcomes, including poverty and incarceration (Bridgeland et al., 2006; Chen & Kaplan,
2003; Janosz, Le Blanc, Boulerice, & Tremblay, 2000). The U.S. Department of Labor (2004)
reports that students who drop out of high school are likely to earn 27% less that those who
graduate, and are 72% more likely to be unemployed. Moreover, those who drop out in earlier
grades are more likely to experience adverse outcomes (Capella & Weinstein, 2001). On the
other hand, students who succeed in school are more likely to engage in responsible adult
behavior, including the care for and financial support of their own children (Simons, Beaman,
Conger, & Chao, 1993).
The School Success Profile
The SSP is a self-report survey that assesses middle and high school students’
perceptions about their neighborhood, school, family, and friends, and about their personal
beliefs and well-being, school attitudes and behavior, and academic performance (Bowen &
Richman, 2001). The 220 multiple-choice survey questions take about 30 to 40 minutes on
average to complete, and student participation is voluntary. Twenty-two dimensions are reported
on the SSP summary report: 14 social environment dimensions and 8 individual adaptation
dimensions associated with school success. These measures have demonstrated high levels of
reliability and validity (Bowen, Rose, & Bowen, 2005). Appendix A includes a summary of each
profile dimension. More details about the history, development, and scoring of the SSP can be
found elsewhere (Bowen et al., 2005).

---

Changes in the Social Environment 7

The SSP is designed to identify students’ strengths and concerns for the purpose of
informing intervention planning. In devising a logic model for informing social work practice in
schools, Bowen, Richman, Bowen, and Woolley (2002) grouped the results from the 22
dimensions into three larger areas: (a) school attitudes and behavior, and academic performance
(distal results of intervention efforts); (b) personal beliefs and well-being (intermediate results);
and (c) social environmental factors, including the neighborhood, school, family, and peers
(proximal results; that is, those most immediately affected by intervention efforts).
In the eco-interactional developmental model that informed the development of the SSP
(Richman et al., 2004), the balance of risks and assets in a student’s social environment
(proximal results) are assumed to influence that student’s level of individual adaptation
(intermediate and distal results) (see Figure 1). Although past studies using SSP data provide
support for this assumption (Bowen et al., 2005), the present study is the first to simultaneously
examine how change in each of the 14 social environment dimensions influences change in each
of the eight individual outcomes associated with school success.
[Insert Figure 1 about here]
Social Environmental Influences on Student Success
Neighborhood, school, family, and friends are systems in a student’s life that influence
individual adaptation outcomes associated with success. A great deal of research has been
conducted to identify the influential relationships between social and environmental systems and
student school success. For purposes of the present study, school success is conceptualized and
measured broadly as eight aspects of positive individual adaptation nested in two categories: (a)
personal beliefs and well-being (i.e., social support, physical health, happiness, personal

---

Changes in the Social Environment 8

adjustment, and self-esteem), and (b) school attitudes and behavior (i.e., school engagement,
trouble avoidance, and academic performance).
In the sections below, we review findings from the earlier research by Powers et al.
(2005) within the context of SSP-related and non-SSP-related research. In reviewing findings
from the Powers et al. study, which resulted from a two-step linear regression data analysis
strategy, we focus attention on the most predictive standardized beta coefficients—those greater
than or equal to .09, which approximate the minimum value necessary to account for at least 1%
of the variance in the dependent variable (Cohen, 1988). Importantly, in Powers et al.’s (2005)
analysis, each social environment dimension was examined for its unique contribution within the
context of the other social environment dimensions and demographic controls. Also, it is
important to bear in mind that the data on which these earlier analyses were conducted were
cross-sectional, and therefore the results from the present analysis will not be directly
comparable.
Neighborhood
In the earlier research by Powers et al. (2005), two of the three neighborhood dimensions
(support and safety) had effects on one or more outcome dimensions that met or exceeded the .09
threshold. Neighborhood support had a positive influence on students’ perceptions of social
support (.09), and neighborhood safety had a positive influence on students’ perceptions of both
physical health (.13) and trouble avoidance (.22). These findings are consistent with other SSP-
related research. For example, using a latent measure of neighborhood social disorganization
(lack of neighbor support, negative peer behavior, and crime and violence), Bowen, Bowen, and
Ware (2002) found middle school and high school students’ perceptions of their neighborhoods
to be directly and negatively related to their reports of positive educational behavior at school

---

Changes in the Social Environment 9

(trouble avoidance, grades, and attendance). In addition, perceptions of neighborhood social
disorganization were indirectly and negatively related to students’ reports of positive educational
behavior through students’ reports of increased parental educational support.
In other SSP-related studies, Woolley and Grogan-Kaylor (2006) found that increased
student perception of neighborhood safety was associated with improved grades and the
avoidance of problem behavior. A study by Bowen and Bowen (1999) identified a positive
relationship between students’ perceptions of neighborhood safety and neighborhood peer
culture and the degree to which they attend school regularly, avoid getting into trouble at school,
and make good grades. SSP-related research by Nash (2002) indicates a positive relationship
between students’ sense of school coherence (the extent to which students find school
meaningful, manageable, and comprehensible) and the degree to which they experience informal
social control (i.e., neighbors support youth and monitor their behavior), low crime, and a
positive peer culture in their neighborhoods.
The hypothesized link between neighborhood dimensions and both personal beliefs/well-
being and school attitudes/behavior also aligns with other non-SSP-related research. Levels of
student happiness (lack of life stress) have been positively related to neighborhood support
(Allison et al., 1999). School attitudes and behavior also have an empirical relationship with the
neighborhood context. Neighborhood support and safety have been shown to promote school
engagement (Cook, Herman, Phillips, & Settersten, 2002). Additionally, students’ perception of
neighborhood safety and neighborhood youth behavior has been tied to both trouble avoidance
(Herrenkohl et al., 2000) and academic performance (Leventhal & Brooks-Gunn, 2004; Schwartz
& Gorman, 2003).

---

Changes in the Social Environment 10

School
In Powers et al.’s (2005) research, two of the three school dimensions (satisfaction and
teacher support) had effects on one or more outcome dimensions at or beyond the .09 lower
limit. School satisfaction had a particularly strong effect on students’ level of school engagement
(.33); teacher support had a positive influence on both students’ perceived social support (.10)
and trouble avoidance (.13). These findings are consistent with other SSP-related research. For
example, supportive teachers have been found to be positively related to middle school students’
perceptions of social support (Richman, Rosenfeld, & Bowen, 1998) and to represent a key
influence in promoting the school success of middle school and high school students (Rosenfeld,
Richman, & Bowen, 2000).
Research by Brewster and Bowen (2004) indicates that teacher support exerted a strong
and significant positive influence on the trouble avoidance of a sample of Hispanic middle
school and high school students who completed the SSP. Bowen and Bowen (1999) found school
safety to be significantly predictive of students’ attendance, trouble avoidance, and grades. Work
by Woolley and Grogan-Kaylor (2006) also suggests a positive relationship between students’
perceptions of school safety and their trouble avoidance.
The hypothesized link between school dimensions and both personal beliefs/well-being
and school attitudes/behavior also aligns with non-SSP-related research. Using a sample of
middle school and high school students, Whitlock (2006) found that positive perceptions of
school safety were related to a sense of school connectedness, which related to students’
perceptions of being affirmed and valued by adults at the school. Trouble avoidance (aggressive
behavior) also has been linked to school satisfaction and teacher support (Lopez, Olaizola,
Ferrer, & Ochoa, 2006). Similarly, academic performance has been tied to school satisfaction

---