613544

research-article2015
UEXXXX10.1177/0042085915613544Urban EducationDavis and Warner

Article
Urban Education
1­–22
Schools Matter: The © The Author(s) 2015
Reprints and permissions:
Positive Relationship sagepub.com/journalsPermissions.nav
DOI: 10.1177/0042085915613544
Between New York City uex.sagepub.com

High Schools’ Student

Academic Progress and
School Climate

Jonathan Ryan Davis1 and Nathan Warner2

Abstract
This article investigates the link between school climate and student
academic progress in New York City Department of Education (NYCDOE)
high schools. Using a data set compiled from 2010-2011 NYCDOE school-
level aggregated demographic, survey, and progress report achievement
data, the authors ran ordinary least squares regressions where they found
that a school’s climate significantly correlated with student academic
progress; under some conditions, the school climate effects outweighed
the effects of student background factors. Finally, the school climate
domains of safety and respect, communication, engagement, and academic
expectations all proved to be important factors that were associated with
student achievement.

Keywords
academic achievement, high school, urban education, learning environment,
school climate, school effectiveness, poverty, school size

1The College of New Jersey, Ewing, NJ, USA

2George Lucas Educational Foundation, San Rafael, CA, USA

Corresponding Author:
Jonathan Ryan Davis, Department of Educational Administration and Secondary Education,
The College of New Jersey, 2000 Pennington Road, Ewing, NJ 08628-0718, USA.
Email: davisj@tcnj.edu

Downloaded from uex.sagepub.com at University of New England on June 4, 2016

2 Urban Education 

Introduction
For decades, researchers in the social sciences have tried to understand how
various factors affect student engagement and achievement. One prominent
line of inquiry focuses on “school effects,” where researchers examine how
much concepts such as school climate, teacher quality, school leader quality,
and high-stakes testing policies actually matter for low-income students
when it comes to raising and sustaining academic achievement over time.
This focus on internal school-level variables leads to a larger question related
to public education and equity: Can schools significantly counteract the
effects of poverty on their own, or are the educational fates of far too many
students pre-determined unless the root causes of poverty are dealt with
directly first?
The “Coleman Report” (Coleman et al., 1966), a foundational work in
educational research, attempted to answer this question. Published in the
mid-sixties amid swirling debates about de-segregating public schools and
equity in school funding, its findings documented how much socio-economic
status (SES), family income levels, and other family background characteris-
tics mattered for student achievement, dwarfing the impact of school effects
such as funding. This was surprising, and after its publication, some politi-
cians concluded that policies targeted at improving schools by increasing
school funding would not generate substantial enough academic gains to jus-
tify their input costs (Guthrie, 1995; Viadero, 2006).
Recent research has questioned Coleman’s methodology and claims
regarding the relative ineffectiveness of school-related inputs (Borman &
Dowling, 2010). Large-scale quantitative studies have uncovered school
effects that, after controlling for outside-of-school factors such as family
income levels and SES, still affect student achievement (Bryk, Lee, &
Holland, 1993; Hoy, Tarter, & Woolfolk Hoy, 2006; Rivkin, Hanushek, &
Kain, 2005). These, and similar studies have documented how school-
level variables, such as trust between teachers, academic rigor, and high
academic expectations predict student achievement even after controlling
for SES. Our large-scale research project on New York City high schools—
the largest public urban1 school district in the United States (see demo-
graphics in Table 2)—will build off these works to elucidate the
relationship between school climate and student achievement. School cli-
mate is also called “school culture” and “learning environment” by some
researchers; although there is some debate about small differences
between the concepts (Anderson, 1982), many scholars are comfortable
using the terms interchangeably (Cohen, McCabe, Michelli, & Pickeral,
2009), as we will do in this article.

Downloaded from uex.sagepub.com at University of New England on June 4, 2016

Davis and Warner 3

Literature Review
Teachers, school leaders, and students do not engage in their daily work in a
vacuum. They are embedded within organizational contexts. Each school’s
organizational culture and context in some ways is unique and has the ability
to influence and affect, in its own right, the individuals who work and go to
school there (Avi-Itzhak, 1987; Bryk & Schneider, 2002). Recently, scholars
have begun to analyze the complex interplay of factors within schools that
play a role in how well students learn and how well teachers teach, above and
beyond SES. The following section outlines literature from one major within-
school factor, “school climate.”

School Climate
School climate influences students emotionally, socially, and academically in
many ways (Cohen et al., 2009; Thapa, Cohen, Higgins-D’Assandro, &
Guffey, 2013). The concept of a school’s internal context as a factor influenc-
ing students was first written more than a century ago by Perry (1908). Dewey
(2008) also wrote about the importance of a school’s internal sense of com-
munity on student development. Yet, it was not until the 1960s that scholars
began to focus systematically on the effects of school climate on students
(Halpin & Croft, 1963). Since the 1980s, school climate has come to be
understood, broadly, as the internal quality and character of school life, which
is comprised of many factors, all of which affect student and adult subjective
experiences within schools. Some of these factors are environmental, struc-
tural, safety, teaching and learning, relationships, community, morale, peer
norms, and school–home–community partnerships (Cohen et al., 2009).
Recently, the National School Climate Center (NSCC), a major organiza-
tion of scholars and practitioners studying the topic, divided school climate
into four general spheres, with school-based standards associated with each:
(a) safety and respect, (b) teaching and learning, (c) relationships, and (d)
environment. Accordingly, school climate reflects the norms, values, behav-
iors, goals, interpersonal relationships, teaching and learning practices, and
organizational structures and policies that operate within a given school con-
text (Cohen et al., 2009; Freiberg, 1999; Thapa et al., 2013). The New York
City Department of Education (NYCDOE) refers to school climate as a
school’s “learning environment.” As we will discuss below, the NYCDOE
measures school climate in each school by giving teachers, parents, and stu-
dents “learning environment” surveys each year. These surveys are broken
down into four categories that, although labeled differently in parts, align to
the NSCC’s four school climate categories above. Therefore, we will briefly
review some of the literature in each of these four domains.

Downloaded from uex.sagepub.com at University of New England on June 4, 2016

4 Urban Education 

Safety and respect.  For both the NSCC and NYCDOE, the safety component
is based on students feeling socially, emotionally, physically, and intellectu-
ally safe in school, which promotes students’ learning and development
(Center for Social and Emotional Education, 2010; Cooper, 2013). When stu-
dents feel safe, they are more apt to learn and succeed academically (Devine
& Cohen, 2007). Conversely, when students are bullied (Klein, 2012; Stock-
dale, Hangaduambo, Duys, Larson, & Sarvela, 2002; Thapa et al., 2013) or
feel imprisoned in their school (Ferguson, 2000; Noguera, 2008), they tend to
suffer academically. The NYCDOE domain of safety and respect also
includes the promotion of group cohesion and mutual respect (Gutmann,
1995; Thapa et al., 2013), which help create a positive school climate for
students (Conchas & Rodriguez, 2008; Valenzuela, 1999).

Communication.  The NYCDOE’s communication domain of school climate

focuses on the importance of relationships—among students, between stu-
dents and teachers, between students and administrators, and among the
teaching faculty itself—in creating a positive learning and teaching environ-
ment. Studies have found that when this component of school climate is suc-
cessfully addressed by a school, it leads to effective risk prevention and
health promotion (Biag, 2014; Blum, 2005; Center for Social and Emotional
Education, 2010; Wang, Haertel, & Walberg, 1993). Moreover, Roorda,
Koomen, Split, and Oort (2011) examined 99 different studies that analyzed
the relationship between student–teacher relations, student engagement, and
academic outcomes; they found that there was a consistent link between stu-
dent–teacher relationships and outcomes.
A supportive school climate also leads to students feeling more connected
to school, and this concept of school connectedness has been shown to lead to
higher academic achievement levels (Blum, 2005; McNeely, Nonnemaker, &
Blum, 2002; Whitlock, 2006). Within the field of sociology, the work on the
nature, quality, and strength of teacher–student relationships, called “social
capital,” and their links to student achievement have provided important con-
tributions in the quest to understand what lies behind student motivation and
engagement (Conchas & Rodriguez, 2008; Noguera, 2004; Stanton-Salazar,
2001; Suarez-Orozco, Suarez-Orozco, & Todorova, 2008; Valenzuela, 1999).
Other scholars have also found a strong link between teachers’ care for their
students and student engagement (Noddings, 2005). Furthermore, the work by
Bryk and Schneider (2002) on the importance of “relational trust” between
adult school-level actors has shed light on the importance of the socio-rela-
tional processes involved with how and under what conditions, trusting rela-
tionships between teachers, parents, and school leaders take hold in a school
community, and their positive effects on student achievement.

Downloaded from uex.sagepub.com at University of New England on June 4, 2016

Davis and Warner 5

Academic expectations and engagement.  Finally, the NYCDOE’s domains of

academic expectations and engagement encompass many teaching and learn-
ing elements. Recent research in New York City finds that high expectations
of students by their teachers positively affect student achievement (Dobbie &
Fryer, 2013). The Center for Social and Emotional Education (2010) finds
that when students feel encouraged to learn, their academic achievement will
increase. Hoy et al. (2006) reinforce these findings when they uncover how a
school’s “academic optimism,” including its expectations for its students, has
a positive effect on student outcomes. Finally, a more supportive and positive
school climate also has been linked to higher student-aspiration levels, and
higher aspirations are linked to higher achievement levels (Plucker, 1998).
Conversely, low aspirations have been linked to student behavior problems
and academic underachievement in school (MacLeod, 1987).
In short, an emerging base of research now exists that documents the
importance of school climate/school culture/learning environment on student
engagement and achievement. This study adds evidence for the power of
school climate compared with student background variables and the strong
relationship between school climate and student achievement.

Research Questions
The analyses presented below address the following research questions:

Research Question 1 (RQ1): Do parent, teacher, and/or student percep-

tions of a school’s climate variables explain more variance in student
academic progress than student background variables?
  Research Question 1a (RQ1a): Does the overall school climate of a
school significantly predict student academic progress?
  Research Question 1b (RQ1b): Do the students’ background charac-
teristics significantly predict student academic progress?
Research Question 2 (RQ2): Do parent, teacher, and student perceptions
about a given school’s climate predict student academic progress?

The Data Set

Introduction
Every year, the NYCDOE releases data on each of its schools to the public.
These data include demographic information; parent, student, and teacher
survey results; and a breakdown of the school-level student achievement data
in the form of what is called a school progress report.2 For this study, we

Downloaded from uex.sagepub.com at University of New England on June 4, 2016

6 Urban Education 

created a data set that combined each high school’s (ninth through 12th
grades) 2010-2011 student background characteristics; school progress report
academic data; and parent, student, and teacher school learning environment
(“school climate”) survey data. The school-level progress reports provided us
with our dependent variable—school-level student academic progress—
whereas the survey data enabled us to determine each school’s “school cli-
mate” score that contributed to the school’s final grade on the progress report.
The data set includes 263 high schools3 and a total of 1,195 student back-
ground, achievement, and “school climate” variables. The majority of the
variables in the data set are school-level disaggregated achievement and
school climate variables. For this study, we focused on aggregated variables
provided by the NYCDOE, which we used to create our own composite vari-
ables (described below.)

NYCDOE School Progress Report Methodology

The NYCDOE issues what it terms “Progress Reports” annually for each
school. These reports are intended to provide a snapshot of a school’s perfor-
mance in five main areas: (a) student progress, (b) student performance, (c)
school environment, (d) college and career readiness, and (e) closing the
achievement gap.4 Because the purpose of this article is to understand the
association between a school’s climate and its student academic progress
(growth), we will focus on those statistical measures constructed by the
NYCDOE that make up student progress and “school climate.”

The Variables
Dependent Variable
Student academic progress composite variable.  The student academic progress
composite variable is produced by the NYCDOE each year based on the stu-
dents from each school’s credit accumulation, Regents’ exam completion
rate, and Regents’ pass rate. It is broken down as follows:

Credit accumulation.  There are six components to this measurement: (a)

percentage of students earning 10+ credits in Year 1 of high school, (b) per-
centage of students earning 10+ credits in Year 2 of high school, (c) percent-
age of students earning 10+ credits in Year 3 of high school, (d) percentage
of students in the school’s lowest third earning 10+ credits in Year 1 of high
school, (e) percentage of students in the school’s lowest third earning 10+
credits in Year 2 of high school, and (f) percentage of students in the school’s

Downloaded from uex.sagepub.com at University of New England on June 4, 2016

Davis and Warner 7

lowest third earning 10+ credits in Year 3 of high school. Students are deter-
mined to be in the school’s academic lowest third by the average of their
eighth-grade New York State English Language Arts (ELA) and math exam
proficiency ratings. Six of the 10 credits earned in each year must be earned
in the core subject areas of math, ELA, social studies, and science.

Average completion rate for regents’ exams.  In New York State, high school
students must pass state exams, called Regents exams, to graduate with the
most basic type of diploma, termed a Regents diploma. Students must pass
at least one exam with a score of 65% in the subjects of ELA, math, U.S.
history, global history, and science. This component measures what percent-
age of students pass Regents exams for each cohort within a high school;
the cumulative number of exams each cohort is expected to have taken and
passed increases each year.

Weighted regents’ pass rate.  The third measure of the student academic
progress composite variable incorporates the weighted Regents pass rate for
each high school. This measure is linked to students’ eighth-grade New York
State test scores because they are highly predictive of high school achieve-
ment. If these state exam scores do not exist for a given student, demographic
factors are used to impute a predicted score. Each student is assigned a weight
that corresponds to how well they performed on their eighth-grade math and
ELA exams.5 Students who are less likely to pass the Regents exam, based on
their eighth-grade exam scores, are awarded more points in this metric when
they pass the Regents exam.
The student academic progress variable was chosen as our dependent vari-
able because it represents a school’s growth in academic achievement. This
progress variable is determined at the end of the school year and aggregated
by the NYCDOE based on the average growth over the school year between
each student in the school using the criteria provided above. Measuring
growth is important in this analysis because it will not bias the data toward
schools that, on average, achieve at very high or low levels from year to year;
instead, the progress variable enables schools from all initial achievement
levels to be compared equally rather than simply examining “pure”
achievement.

Independent Variables
School climate composite variables.  Our school climate composite variables—
student, teacher, and parent—come from survey data collected in the spring
of each year by the NYCDOE in partnership with New York University’s

Downloaded from uex.sagepub.com at University of New England on June 4, 2016

8 Urban Education 

(NYU) Steinhardt School of Education’s Research Alliance for New York

City Schools. The surveys, which are given to all teachers, students, and par-
ents in Grades 6 through 12 are called “learning environment surveys” and
are designed to measure the school climate of each school. The surveys were
first developed by researchers, experts, community members, parents, and
NYCDOE employees in 2006 and were built on the survey development
work of the University of Chicago Consortium on Chicago School Research
(CCSR) as well as on other well established school surveys. In 2010, the
NYCDOE partnered with NYU’s Steinhardt School of Education to review
and enhance the surveys every year, which they have done. In 2013, NYU
released a brief and a technical report on this survey work to date (Nathanson
et al., 2013).
The surveys use Likert-type scales in their question responses, and each
response is awarded a point value between 0 and 10. For each question, the
response option that reflects most favorably on a school’s learning environ-
ment (i.e., strongly agree) is awarded a 10, and the response option that
reflects most negatively on a school’s learning environment (i.e., strongly
disagree) is awarded a 0. The middle two options are awarded 6.7 and 3.3
points respectively. The questions fall into four broad categories, or domains,
of school climate: academic expectations, communication, engagement, and
safety and respect. Average question scores are used to determine an overall
domain score for each respondent group (parent, teacher, student). The over-
all domain scores for each school are then calculated by averaging each of the
four domain scores for each group. Below, we describe the four learning
environment domains more in depth.6

Academic expectations. Academic expectations measures the degree to

which the adults in a school encourage students to do their best and develop
rigorous academic goals. Survey questions include the following:

Parent: (a) How satisfied are you with the quality of your child’s teacher(s)
this year? (b) How satisfied are you with the education your child has
received this year?
Teacher: (a) The principal places the learning needs of children ahead of
other interests. (b) Teachers in this school set high standards for student
work in their classes.
Student: (a) My teachers encourage me to succeed. (b) My school helps
me develop challenging academic goals.

Communication.  Communication measures the degree to which adults in a

school communicate educational goals, requirements, and provide feedback

Downloaded from uex.sagepub.com at University of New England on June 4, 2016

Davis and Warner 9

to students regarding their learning outcomes. Survey questions include the

following:

Parent: How much do you agree or disagree with the following statements
about your child’s school or teacher? (a) The school clearly communicates
its expectations for my child’s learning to me and my child. (b) My child’s
teacher(s) give helpful comments on class work and tests.
Teacher: (a) School leaders communicate a clear vision for this school. (b)
School leaders let staff know what is expected of them.
Student: (a) Most of the teachers, counselors, school leaders, and other
adults I see at school every day know my name or who I am. (b) On a scale
of 1 to 4, how comfortable are you talking to teachers and other adults at
your school about: a problem you are having in class?

Engagement.  Engagement measures the degree to which the adults in a

school involve students, parents, and educators in the education of the stu-
dents. Survey questions include the following:

Parent: (a) I feel welcome in my child’s school. (b) My child’s schools

offers a wide enough variety of courses and activities to keep my child
interested in school.
Teacher: (a) Students with disabilities are included in all school activities.
(b) School leaders encourage collaboration among teachers.
Student: (a) I feel welcome in my school. (b) The adults at my school look
out for me.

Safety and respect.  Safety and respect measure the degree to which stu-
dents feel safe and respected in the school. Survey questions include the fol-
lowing:

Parent: (a) My child is safe at school. (b) Discipline is enforced fairly at

my child’s school.
Teacher: (a) Teachers in my school trust each other. (b) Crime and vio-
lence are a problem in my school.
Student: (a) Teachers in my school treat students with respect. (b) Adults
in my school treat each other with respect.

Our school climate composite variable is constructed out of these four

areas of a school’s learning environment, and it is disaggregated for teachers,
students, and parents. We accepted the NYCDOE’s tabulation of the survey
results for each school as presented on the NYCDOE’s website.

Downloaded from uex.sagepub.com at University of New England on June 4, 2016

10 Urban Education 

We began our analysis of the “school climate” variables by evaluating the

correlation between the NYCDOE-created, domain-specific environment
scales (Safety and Respect, Engagement, Academic Expectations, and
Communications) for each of the three NYC Survey groups—parent, teacher,
and student. After determining there was a strong correlation between these
scales, with a correlation above .700 for each combination of variables, we did
a factor analysis to see if we could create a composite by combining the four
domain-specific scales for each survey group into one school climate compos-
ite variable.7 In doing so, we had a Cronbach’s alpha of .983 for our “Teacher
School Climate” variable; a Cronbach’s alpha of .985 for our “Student School
Climate” variable, and a Cronbach’s alpha of .982 for our “Parent School
Climate” variable. Finally, we ran correlations between the parent, teacher,
and student “school climate” composites to ensure the uniqueness of the scales
and guard against multicollinearity, which can be seen in Table 1. Once verify-
ing their uniqueness, we had three “school climate” scale variables: (a) student
school climate, (b) parent school climate, and (c) teacher school climate.

Student background variables.  For our student background variables, we used

all of the student background variables provided to us by the NYCDOE in the
“Progress Reports,” demographic statistics for schools, and information
about school selectivity. These variables were aggregated at the school level
and were given as percentages of a particular school’s student body. They are
(a) percent receiving free and reduced-price lunch, (b) percent of students
who are Black or Hispanic, (c) percent that is male, (d) percent of students
categorized as special education (SPED), (e) percent of English language
learners (ELL), (f) student body average for eighth-grade math and ELA
combined scores on the New York State standardized exams, (g) whether a
school selects students or not, and (h) the overall school enrollment. We cen-
tered the free and reduced-price lunch variable, and we set the school enroll-
ment variable on a scale of 100 for the purposes of presenting the data.

Steps of Analysis
Descriptive Statistics
Using SPSS for all analyses, we first ran descriptive statistics for all vari-
ables, both dependent and independent, which can be found in Table 2. They
were (a) student academic progress; (b) student, parent, and teacher school
climate scale score; (c) the disaggregated school-level engagement, safety
and respect, academic expectations, and communication variables by student,
parent, and teacher; and (d) the school-level student background variables of

Downloaded from uex.sagepub.com at University of New England on June 4, 2016

 Table 1.  Correlation Matrix for Variables Used in Analysis of New York City Department of Education School Climate and
Academic Progress.
1 2 3 4 5 6 7 8 9 10 11 12 13

 1. HS academicprogress 1.00  

 2. Student schoolclimate .444** 1.00  
 3. Teacher schoolclimate .355** .350** 1.00  
  4.  Parent school climate .419** .580** .269** 1.00  
  5.  Eighth-grade math-ELA scores .275** .345** .145* .051 1.00  
  6.  % of free and reducedlunch −.044 −.159* −.111 .156* −.698** 1.00  
  7.  % of male −.128* −.079 .024 −.017 −.131* .059 1.00  
  8.  % of Englishlanguage learner .187** .278** .049 .205** −.418** .349** .090 1.00  
  9.  % of specialeducation student −.358** −.412** −.073 −.144* −.544** .326** .166** −.412** 1.00  
10.  % of Black/Hispanicstudents −.268** −.411** −.222** .040 −.761** .690** −.008 −.009 .538** 1.00  
11. Totalenrollment −.131* −.071 .104 −.410** .320** −.434** .093 −.084 −.176** −.517** 1.00  
12.  Screened HS .168** .141* .033 −.014 .181** −.137* .040 .275** −.409** −.272** .177** 1.00  
13.vTest/audition HS .107 .117 .051 −.082 .444** −.390** −.129* −.132* −.303** −.354** .304** −.222** 1.00

Downloaded from uex.sagepub.com at University of New England on June 4, 2016

Note. HS = high school; ELA = English language arts.
*p < .05. **p < .01.

11
12 Urban Education 

Table 2.  Descriptive Statistics for All NYCDOE High Schools, 2010-2011.

N Median M SD Range
Dependent variable
  HS academic progress 263 35.20 35.34 7.89 48
Independent variables
  Parent school climate 263 7.76 7.76 0.54 3.46
  Teacher school climate 263 7.08 7.04 0.87 4.75
  Student school climate 262 6.54 6.62 0.53 3.38
  % free and reduced lunch 261 78.60 73.34 17.55 82.00
  Total enrollment 261 461.00 936.43 1,017.73 4,940
  % English language learners 255 7.66 14.63 20.77 92.00
  % special education students 261 14.11 13.64 6.77 29.91
  % Black/Hispanic students 261 93.51 80.57 24.09 95.92
  % male 261 49.55 49.54 13.72 100.00
  Eighth-grade math-ELA scores 263 2.56 2.67 0.38 2.07
  % non-selective HS 263 60.50 60.50  
  % screened HS 263 28.50 28.50  
  % test/audition HS 263 11.00 11.00  
Domain specific independent variables
  Parent engagement 263 7.49 7.51 0.52 3.36
  Teacher engagement 263 7.02 6.98 0.93 5.66
  Student engagement 262 6.49 6.56 0.57 3.70
  Parent safety and respect 263 8.19 8.10 0.65 4.14
  Teacher safety and respect 263 7.18 7.13 0.85 3.98
  Student safety and respect 262 6.55 14.48 0.70 3.84
  Parent communication 263 7.58 7.57 0.56 3.38
  Teacher communication 263 6.55 6.52 1.06 6.35
  Student communication 262 5.93 6.00 0.50 3.63
  Parent academic expectations 263 7.75 7.76 0.54 3.03
  Teacher academic expectations 263 7.55 7.53 0.86 4.35
  Student academic expectations 262 7.33 7.38 0.44 2.70

Note. NYCDOE = New York City Department of Education; HS = high schools; ELA = English
Language Arts

free and reduced-price lunch, total enrollment, % ELL, % SPED, % Black

and Hispanic, % male, and eighth-grade math plus ELA scores.

Ordinary Least Squares Regression Analyses

Because of the limitations in our data set, where each variable was a Level 2
(school-level) variable, we were only able to conduct an ordinary least

Downloaded from uex.sagepub.com at University of New England on June 4, 2016

Davis and Warner 13

squares regression analysis. Hierarchical linear modeling would have pro-

vided a more sophisticated analysis of our research questions, but it was not
possible because the NYCDOE did not provide student-level (Level 1) data
for any of the variables we used in 2010-2011.

All high school, school climate scale regressions.  To answer RQ1 and RQ2, we
ran four different ordinary least squares regression models to determine the
relationship between school climate and academic progress. The first regres-
sion analyzed our parent, teacher, and student school climate scales (all cen-
tered) in relation to student academic progress in all NYCDOE high schools.
Our second regression analyzed each of the student background variables in
relation to student progress. This enabled us to compare the adjusted R2 of
each of these two models to determine which sets of variables could explain
a greater degree of variance in the dependent variable. Our third regression
was nested, combining the “school climate” and “student background” vari-
ables—models one and two—into one larger model to determine the signifi-
cance of each variable when controlling for a more comprehensive list of
variables that might predict student academic progress. The results for all
three school climate regressions are presented in Table 3.

Results
Descriptive Statistics
All NYCDOE high schools.  As seen in Table 2, for all 263 NYCDOE high
schools, the median for our dependent variable—student academic prog-
ress—was 35.20 (out of 60). When looking at the “school climate” variables
for all high schools, median scores for parents were the highest (7.76), fol-
lowed by teachers (7.08), and then by students (6.54). As well, for every
domain of the school climate scales—engagement, safety and respect, com-
munication, and academic expectations—parents retained the highest median,
followed by teachers, and then students.
The median composition for all NYCDOE high schools demonstrates
the effects of race and class-based segregation within New York City’s
school system (Kasinitz, Mollenkopf, Waters, & Holdaway, 2008).
According to the median, 78.60% of students in each school have free or
reduced-price lunches, 7.66% of students are ELL, 14.11% of students are
in SPED, 93.51% are Black and Hispanic, 49.55% are male, and the
median eighth-grade math-ELA score is 2.56 (out of 4). The median enroll-
ment for each school is 461, which accounts for some schools that are
larger than 5,000 students and others that have only a couple hundred stu-
dents. Finally, 60.50% of the schools are non-selective, 28.50% have

Downloaded from uex.sagepub.com at University of New England on June 4, 2016

 14
Table 3.  Linear Regression Models Predicting Student Academic Progress in All NYCDOE High Schools.
Model 1 Model 2 Model 3

All NYCDOE high schools B β B β B β

Intercept −1.603 (13.244) −8.733 (14.089) −5.115 (14.497)  

“School Climate” variables
  Student school climate 3.314** (1.065) .216 0.180 (1.207) .012
  Parent school climate 3.497** (1.005) .236 3.315** (1.122) .226
  Teacher school climate 2.002*** (0.524) .220 2.264*** (0.510) .249
“Student Background” variables
  Eighth-grade math-ELA scores 9.038* (3.514) .396 4.350 (3.382) .191
  % free and reduced luncha 0.096* (0.041) .202 0.077* (0.039) .163
  % Male −.040 (0.034) −.069 −0.041 (0.031) −.071
  % English language learners 0.100* (0.047) .261 0.035 (0.046) .093
  % Special education students 0.025 (0.132) .021 −0.083 (0.122) −.069
  % Black/Hispanic students −.079 (0.042) −.229 −0.071 (0.041) −.202
  Total enrollment −.243*** (0.056) −.315 −0.169** (0.057) −.220
  (Factor of 100)  
  Screened HS 1.063 (1.203) .061 1.545 (0.057) −.220
  (Ref. Un-screened HS)  
  Test/audition HS 1.995 (1.839) .074 2.553 (1.694) .095
  (Ref. Un-screened HS)  

Downloaded from uex.sagepub.com at University of New England on June 4, 2016

N 254 254 254  
R2 .343 .312 .324  

Note. Standardized errors are in parentheses. NYCDOE = New York City Department of Education; ELA = English language arts. HS = high school.
aVariable has been centered.
*p < .05. **p < .01. ***p < .001.
Davis and Warner 15

screened enrollment (usually by grade point, test scores, or neighborhood/

district), and 11% are test or audition high schools.

Regressions for All NYCDOE High Schools

As one can see in Table 3, Model 1 examines how student, teacher, and parent
“school climate” scales relate to student academic progress. This model had
an R2 of .262, and each variable was significant in a positive direction, with
the parent school climate scale being the strongest predictor, followed by
teacher, and then student.
Counter to the findings of Coleman et al. (1966), Model 2, which had only
the student background characteristics as predictors, demonstrates that
“school climate” variables explain more of the variance in student academic
progress than “student background variables,” as Model 2 only had an R2 of
.231. Our findings, and recent studies by Hoy et al. (2006), Gamoran and
Long (2006), Bryk and Schneider (2002), and Borman and Dowling (2010)
demonstrate the strong role that a school’s climate plays in student
achievement.
In Model 2 of Table 3, only enrollment was highly significant, confirming
what proponents of the small schools movement in New York City argue that
the smaller the school, the better students will progress academically (Bartlett
& Garcia, 2011). This model also found that the better the eighth-grade math-
ELA scores, the better students progressed academically; this is the strongest
predictor in Model 2 of Table 3. Moreover, the model found that the higher
percent free and reduced-price lunches and percent ELL a school had, the
greater the student academic progress. Each of these explanatory variables is
minimally significant whereas the remaining variables are not significant at
all.
When we combined the school climate and student background variables
in Model 3 of Table 3, the student school climate scale was no longer signifi-
cant and only total enrollment and free and reduced-price lunch were signifi-
cant student background variables. Also, eighth-grade math-ELA scores,
which had been the strongest predictor in Model 2 was no longer significant.
This demonstrates that when controlling for the three school climate scales,
where a student enters high school academically did not predict how he or she
would progress academically throughout high school. In addition, teacher
and parent school climate scales were stronger predictors of academic prog-
ress than the student background variables, with the teacher’s perception of
school climate being the largest predictor. This aligns with research that has
shown a positive correlation between school context, teacher satisfaction and
retention, and student achievement (Johnson, Kraft, & Papay, 2012).

Downloaded from uex.sagepub.com at University of New England on June 4, 2016

16 Urban Education 

Discussion
This study’s research questions were as follows:

RQ1: Do parent, teacher, and/or student school climate variables explain

more variance in student academic progress than student background
variables?
  RQ1a: Does the overall school climate of a school significantly predict
student academic progress?
  RQ1b: Do the students’ background characteristics significantly pre-
dict student academic progress?
RQ2: Do parent, teacher, and student perceptions about a given school’s
climate predict student academic progress?

Through our results, we can see that the school climate variables did
explain more of the variance in academic progress than student background
characteristics (including free and reduced-price lunch status) and that both
school climate and students’ background characteristics do predict academic
achievement. These findings support previous research that found there is a
direct relationship between school climate and academic achievement (Hoy,
Hannum, & Tschannen-Moran, 1998; Sherblom, Marshall, & Sherblom,
2006; Thapa et al., 2013).
What our results add to this research is (a) highlighting how the school
climate variables explain greater variance in academic growth than student
background variables and (b) when these variables are combined into one
model, the school climate variables have a greater effect on academic growth
than all student background variables (as seen in Table 3, Model 3). Such
findings challenge Coleman et al. (1966) who argued that student background
characteristics outweighed the effects of “within school” factors on student
achievement growth indicators. In our study, the only two “student back-
ground” variables that were significant in relation to student academic growth
were a school’s enrollment and the students’ status with regard to free and
reduced-price lunch; of those two variables, only a school’s total enrollment
was consistently significant across the models (as seen in Table 3, Models 2
and 3). This finding on the importance of school size supports the recent
research on the benefits of small schools in New York City (Abdulkadiroglu,
Hu, & Pathak, 2013; Bloom & Unterman, 2013). As well, our findings point
to the importance of teachers’ perceptions of school climate, which is the
most predictive of academic growth (as seen in Table 3, Model 3).
It is also important to note that both parent and teacher perceptions of
school climate were significant although, interestingly, student perceptions of
school climate were not. This lack of a significant beta for student perception

Downloaded from uex.sagepub.com at University of New England on June 4, 2016

Davis and Warner 17

can likely be attributed to the high correlation (.580) between parent and
student perceptions, as parents’ perceptions can be influenced by what their
children tell them about their experiences in school. However, we believe that
this difference between parent and student perceptions of school climate is
interesting grounds for further research.

Limitations
There were multiple limitations to our study. The first limitation to the
study was with our data set, which only had school-level data. If disaggre-
gated, student-level data were made available by the NYCDOE, more
sophisticated multi-level models could be explored. A second limitation in
the data, related to the same issue of disaggregation, is our inability to com-
pare the relationship between school climate and academic growth based on
a student’s age or grade level; for example, does school climate matter more
for ninth graders than for 12th graders? Finally, because this study focused
solely on all 263 New York City high schools, we recommend that future
research be done that replicates our analysis for higher and lower perform-
ing schools to see if the associations differ when overall school effective-
ness is taken into account.

Conclusion
Although our study only examines New York City public high schools—
which combined, educate more than 200,000 students—it points to three sig-
nificant findings that could pertain to schools nationwide: (a) A school’s
climate has a significant relationship with how well students progress aca-
demically; (b) the school climate domains of safety and respect, communica-
tion, engagement, and academic expectations together (as reported by
teachers and parents) within a school can help predict student achievement;
and (c) school climate variables explain more variance in academic growth
than student background variables. These findings combined demonstrate
that improvement can happen from within schools and their districts, particu-
larly large, urban school districts such as New York City. This improvement
can come from districts and schools implementing initiatives that promote
healthy school climates: by fostering high academic expectations, strength-
ening communication and relationships between school stakeholders (stu-
dents, parents, teachers, administrators), actively engaging students and
families inside and outside of the classroom, and promoting a safe and
respectful learning environment. We believe district and school leaders can
put measures in place to enact policies and practices that embed students
within caring and supportive learning environments.

Downloaded from uex.sagepub.com at University of New England on June 4, 2016

18 Urban Education 

Schools are organizations, each with common, yet also unique, internal

characteristics. It follows that school improvements must go beyond solely
focusing on instructional issues, standardized testing, leadership practices,
and teacher accountability, and also deal with the ways in which teachers,
parents, students, and school leaders work together to create effective cul-
tures within schools. In other words, the organizational contexts of, and the
social relations within, schools are important because they affect instruc-
tional quality and academic rigor within classrooms and how students con-
nect and identify with school as a safe, relevant, and engaging place to spend
time.
As suggested by our results, researchers and educators should continue to
examine whether findings such as ours can be generalized to schools nation-
ally. Therefore, it is important that more researchers examine the relation-
ships between student academic progress and the various components of a
school climate while also looking at how and why a given school’s climate
might change over time (Cohen et al., 2009; Thapa et al., 2013). Some
researchers believe that the ideal method to determine this relationship is by
using an experimental research design (Tornatsky, Brookover, Hathaway,
Miller, & Passalacqua, 1980). We also believe this relationship can be
explored through quantitative analysis of nationally representative data sets,
regional data sets, or local data sets where school climate surveys are given.
Qualitative studies are also needed in this area to provide a better under-
standing of the social processes and mechanisms operating to influence stu-
dent engagement and achievement within schools. As such, researchers
should consider conducting more ethnographic studies within schools, per-
haps comparative in nature, seeking to understand the impact of social rela-
tions on pedagogical practices. Breaking down some of the themes in the
school climate surveys, researchers would be well served to ask questions
such as the following: (a) How do teachers and leaders build safe, engaging,
and respectful learning communities for students? (b) How does a sense of
community and solidarity with peers and teachers positively influence stu-
dent engagement and achievement from a student point of view? (c) How
does teacher–teacher relational trust (Bryk & Schneider, 2002) and collabo-
ration positively correlate with the quality of teachers’ instruction, and what
types of school structures, policies, and practices help teachers get on and
then stay on the same page throughout the school year?
One place where such qualitative research should occur is in successful
schools serving low-income students; these studies can describe and analyze
the complex web of pedagogical and socio-relational processes operating
within these outlier schools beating the socio-economic odds. Such studies
will add greatly to our knowledge about how to build successful schools for
all students.

Downloaded from uex.sagepub.com at University of New England on June 4, 2016

Davis and Warner 19

Declaration of Conflicting Interests

The author(s) declared no potential conflicts of interest with respect to the research,
authorship, and/or publication of this article.

Funding
The author(s) received no financial support for the research, authorship, and/or publi-
cation of this article.

Notes
1. Similar to Milner (2009), we agree that there are a myriad of definitions to the
term urban; however, we conceptualize the term in New York City as “inner-
city” public schools within the five boroughs.
2. The New York City Department of Education (NYCDOE) data can be found at
http://schools.nyc.gov/Accountability/data/default.htm.
3. There were originally 371 NYCDOE high schools; however, 108 of these
schools were too new to have progress report scores necessary to be included in
our analysis because they were not full-sized ninth- to 12th-grade schools yet.
The 263 schools in our data set represent all the high schools in the NYCDOE
that were complete ninth- to 12th-grade schools during the 2010-2011 school
year.
4. A detailed description of the statistical measures used in the “2010-2011 Progress
Report” can be found at http://schools.nyc.gov/Accountability/tools/report/
default.htm#Methods.
5. The tables of weights are found by going to the NYCDOE methods section link
in Note 3.
6. The complete survey used by the NYCDOE can be found at http://schools.nyc.
gov/NR/rdonlyres/776ED7EA-82CE-4DBE-A037-08ACD250BB44/0/2011_
citywide_gened.pdf. For a detailed description of the survey methods/
tabulations, see http://schools.nyc.gov/NR/rdonlyres/1A910A05-77B6-4046-
8E67-7380E4E32EA9/0/2011GenEdSurveyScoringGuide_061011.pdf.
7. Although there was a high correlation between the domain-specific scales, we
decided to use these individualized scales to compare each domain across the
three survey groups (parent, teacher, and student). For example, we compared
the parent, teacher, and student perception of safety and respect; these three
scales were not collinear when compared with one another.

References
Abdulkadiroglu, A., Hu, W., & Pathak, P. (2013, October). Small high schools and
student achievement: Lottery-based evidence from New York City (Working
Paper No. 19576). Cambridge, MA: National Bureau of Economic Research.
Anderson, C. (1982). The search for school climate: A review of the research. Review
of Educational Research, 52, 368-420.
Avi-Itzhak, T. E. (1987). Alienation from secondary school: The effect of school
climate on Arab Israeli students. Urban Education, 21, 353-364.

Downloaded from uex.sagepub.com at University of New England on June 4, 2016

20 Urban Education 

Bartlett, L., & Garcia, O. (2011). Additive schooling in subtractive times: Bilingual
education and Dominican immigrant youth in the Heights. Nashville, TN:
Vanderbilt University Press.
Biag, M. (2014). A descriptive analysis of school connectedness: The views
of school personnel. Urban Education. Advance online publication. doi:
10.1177/0042085914539772
Bloom, H., & Unterman, R. (2013). Sustained progress: New findings about the effec-
tiveness and operation of small public high schools of choice in New York City.
New York, NY: MDRC.
Blum, R. (2005). School connectedness: Improving the lives of students. Baltimore,
MD: Johns Hopkins Bloomberg School of Public Health.
Borman, G., & Dowling, M. (2010). Schools and inequality: A multilevel analysis of
Coleman’s equality of educational opportunity data. Teachers College Record,
112, 1201-1246.
Bryk, A., Lee, V., & Holland, P. (1993). Catholic schools and the common good.
Cambridge, MA: Harvard University Press.
Bryk, A., & Schneider, B. (2002). Trust in schools: A core resource for improvement.
New York, NY: Russell Sage.
Center for Social and Emotional Education. (2010, January). School climate research
summary (School Climate Brief, Vol. 1, No.1). New York, NY: Center for Social
and Emotional Education.
Cohen, J., McCabe, E. M., Michelli, N. M., & Pickeral, T. (2009). School climate:
Research, policy, practice, and teacher education. Teachers College Record, 111,
180-213.
Coleman, J., Campbell, E., Hobson, C., McPartland, J., Mood, A., Weinfeld, F., &
York, R. (1966). Equality of educational opportunity. Washington, DC: U.S.
Government Printing Office.
Conchas, G., & Rodriguez, L. (2008). Small schools and urban youth: Using the
power of school culture to engage students. Thousand Oaks, CA: Corwin Press.
Cooper, K. S. (2013). Safe, affirming, and productive spaces: Classroom engagement
among Latina high school students. Urban Education, 48, 490-528.
Devine, J., & Cohen, J. (2007). Making your school safe: Strategies to protect chil-
dren and promote learning. New York, NY: Teachers College Press.
Dewey, J. (2008). Democracy and education. Radford, VA: Wilder Publication.
Dobbie, W., & Fryer, R. G. (2013). Getting beneath the veil of effective schools:
Evidence from New York City. American Economic Journal: Applied Economics,
5(4), 28-60.
Ferguson, A. (2000). Bad boys: Public schools and the making of black masculinity.
Ann Arbor: University of Michigan Press.
Freiberg, H. J. (1999). School climate: Measuring, imposing and sustaining healthy
learning environments. Philadelphia, PA: Falmer Press.
Gamoran, A., & Long, D. A. (2006, December). Equality of educational opportu-
nity: A 40-year retrospective (WCER Working Paper No. 2006-9). Madison:
Wisconsin Center for Education Research, University of Wisconsin–Madison.

Downloaded from uex.sagepub.com at University of New England on June 4, 2016

Davis and Warner 21

Retrieved from http://www.wcer.wisc.edu/publications/workingPapers/papers.

php
Guthrie, J. W. (1995). Implications for policy: What might happen in American edu-
cation if it were known how money actually is spent? In L. O. Picus & J. L.
Wattenberger (Eds.), Where does the money go? Resource allocation in elemen-
tary and secondary schools (pp. 253-268). Thousand Oaks, CA: Corwin Press.
Gutmann, A. (1995, March 31). Challenges of multiculturalism in democratic
education. Keynote lecture, Philosophy of Education Society and American
Philosophical Association, Pacific Division, San Francisco, CA.
Halpin, A. W., & Croft, D. B. (1963). Organizational climate of schools. Chicago, IL:
University of Chicago.
Hoy, W. K., Hannum, J., & Tschannen-Moran, M. (1998). Organizational climate and
student achievement: A parsimonious and longitudinal view. Journal of School
Leadership, 8, 336-359.
Hoy, W. K., Tarter, C. J., & Woolfolk Hoy, A. (2006). Academic optimism of schools:
A force for student achievement. American Educational Research Association,
43, 425-446.
Johnson, S., Kraft, M., & Papay, J. (2012). How context matters in high-need schools:
The effects of teachers’ working conditions on their professional satisfaction and
their students’ achievement. Teachers College Record, 114(10), 1-39.
Kasinitz, P., Mollenkopf, J., Waters, M., & Holdaway, J. (2008). Inheriting the
city: The children of immigrants come of age. New York, NY: Russell Sage
Foundation.
Klein, J. (2012). The bully society: School shootings and the crisis of bullying in
America’s schools. New York: New York University Press.
MacLeod, J. (1987). Ain’t no makin’it: Aspirations and attainment in a low-income
neighborhood. Boulder, CO: Westview Press.
McNeely, C., Nonnemaker, J., & Blum, R. (2002). Promoting student connectedness
to school: Evidence from the national longitudinal study of adolescent health.
Journal of School Health, 72, 138-146.
Milner, H. R. (2009). Preparing teachers of African American students in urban
schools. In L. Tillman (Ed.), The SAGE handbook of African American educa-
tion (pp. 123-139). Los Angeles, CA: Sage.
Nathanson, L., Segeritz, M., Cole, R., Lent, J., McCormick, M., & Kemple, J. (2013).
NYC school survey 2008-2010: Assessing the reliability and validity of progress
repot measures. New York: The Research Alliance for New York City Schools.
Retrieved from http://media.ranycs.org/2013/010
Noddings, N. (2005). The challenge to care: An alternative approach to education.
New York, NY: Teachers College, Columbia University Press.
Noguera, P. (2004). Social capital and the education of immigrant students: Categories
and generalizations. Sociology of Education, 77, 180-183.
Noguera, P. (2008). The trouble with Black boys: . . . And other reflections on race,
equity, and the future of public education. San Francisco, CA: Jossey-Bass.
Perry, A. (1908). The management of a city school. New York, NY: Macmillan.

Downloaded from uex.sagepub.com at University of New England on June 4, 2016

22 Urban Education 

Plucker, J. (1998). The relationship between school climate conditions and student
aspirations. The Journal of Educational Research, 91, 240-246.
Rivkin, S. G., Hanushek, E. A., & Kain, J. F. (2005). Teachers, schools, and academic
achievement. Econometrica, 73, 417-458.
Roorda, D. L., Koomen, H. M. Y., Split, J. L., & Oort, F. J. (2011). The influence of affec-
tive teacher—Student relationships on students’ school engagement and achieve-
ment: A meta-analytic approach. Review of Educational Research, 81, 493-529.
Sherblom, S. A., Marshall, J. C., & Sherblom, J. C. (2006). The relationship between
school climate and math and reading achievement. Journal of Research in
Character Education, 4(1&2), 19-31.
Stanton-Salazar, R. (2001). Manufacturing hope and despair: The school and kin
support networks of U.S.-Mexican youth. New York, NY: Teachers College,
Columbia University.
Stockdale, M. S., Hangaduambo, S., Duys, D., Larson, K., & Sarvela, P. (2002). Rural
elementary students’, parents’, and teachers’ perceptions of bullying. American
Journal of Health Behavior, 26, 266-277.
Suarez-Orozco, C., Suarez-Orozco, M., & Todorova, I. (2008). Learning a new land:
Immigrant students in American society. Cambridge, MA: Harvard University Press.
Thapa, A., Cohen, J., Higgins-D’Assandro, A., & Guffey, S. (2013). A review of
school climate research. Review of Educational Research, 83, 357-385.
Tornatsky, L. G., Brookover, W. B., Hathaway, D. V., Miller, S. K., & Passalacqua,
J. (1980). Changing schools climate: A case study in implementation. Urban
Education, 15, 49-64.
Valenzuela, A. (1999). Subtractive schooling: U.S.-Mexican youth and the politics of
caring. Albany: State University of New York Press.
Viadero, D. (2006, June 21). Race report’s influence felt 40 years later: Legacy of
Coleman study was new view of equity. Education Week, 23(41), 21-24.
Wang, M. C., Haertel, G. D., & Walberg, H. J. (1993). Toward a knowledge base for
school learning. Review of Educational Research, 63, 249-294.
Whitlock, J. L. (2006). Youth perceptions of life in school: Contextual correlates of
school connectedness in adolescents. Applied Developmental Science, 10, 13-20.

Author Biographies
Jonathan Ryan Davis is an Assistant Professor in the Department of Educational
Administration and Secondary Education at The College of New Jersey. He taught
high school social studies at urban public high schools in northern Kentucky and New
York City. His research focuses on school climate, teacher education, classroom man-
agement, and middle level education.
Nathan Warner is the Director of Professional Development at Lucas Education
Research, a division of the George Lucas Educational Foundation. He was a midde
and high school teacher and instructional coach for sixteen years in the New York
City public schools. He received his PhD in sociology from the City University of
New York Graduate Center in 2014, where his research focused on school climate,
teacher professional community, and teacher development.

Downloaded from uex.sagepub.com at University of New England on June 4, 2016