The Impact of City Connects

e Impact of

City Connects

Progress Report 2014

B C C 

O S 

ó ó ó

City Connects Progress Report 2014

ACKNOWLEDGEMENTS

We gratefully acknowledge the support of the Boston Public Schools: the

Superintendent, the Office of Data and Accountability, and the Office of

Individualized Student Learning. We also gratefully acknowledge the

support of the Springfield Public Schools: the Superintendent; the Office of

Information, Technology, and Accountability; and the Pupil Services staff

in the Department of Special Education. We could not have accomplished

this work without the unwavering support of the principals, teachers, staff,

and students of the participating City Connects schools. Finally, we thank

the Lynch School of Education, Boston College, and our funders for their

generous support.

CURRENT FOUNDATION SUPPORT

Barr Foundation

Better Way Foundation

Charles Hayden Foundation

GHR Foundation

Mathile Family Foundation

New Balance Foundation

The Philanthropic Initiative

Strategic Grant Partners

GOVERNMENT SUPPORT

Massachusetts Department of Elementary and Secondary Education

Boston Public Schools

Springfield Public Schools

Introduction

It has long been recognized that in high-poverty urban school districts,

children face challenges outside of school that impede academic success. In

the 1960’s, the Coleman Report concluded that students’ socioeconomic and

home background are significant factors affecting academic achievement.

Current research confirms that larger social structures and contexts

beyond the school are critical, accounting for up to two-thirds of the

variance in student achievement.

Schools cannot close the achievement

gap without a systemic approach to addressing out-of-school factors.

While the challenge of poverty may be society’s to solve, and while some

non-academic barriers to learning cannot be addressed by schools, in the

absence of a large-scale societal solution, schools can provide supports that

mitigate some of the impact of poverty.

To address these out-of-school factors that impede learning, we designed

City Connects (CCNX). The mission of CCNX is to have children engage

and learn in school by connecting each child with the tailored set of

prevention, intervention, and enrichment services he or she needs to

thrive. To accomplish this mission, CCNX relies on the rich supports

and enrichments provided by district programs and community

agencies. To link schools and community agencies, CCNX has developed

a school-based system that coordinates comprehensive supports for

learning and healthy development. The intervention identifies each

student’s strengths and needs in academic, social-emotional, physical,

and family domains and works with community agencies to deliver a

tailored set of services to every child. This system transforms existing

school structures and is aligned with conceptual consensus regarding

optimal practice. The intervention described in this report is designed

for elementary school students. The CCNX Implementation Team is

currently adapting the model for early childhood and for middle and

high school students. The Evaluation Team is following the elementary

school students once they leave the intervention and enter middle

school and then high school.

Over time, the evaluation of CCNX has responded to the expansion

of the intervention within and beyond Boston. In the academic year

2001-02, CCNX was initially implemented in six schools located in one

1 Harrington, 1962; Coleman, et al., 1966; Blow, 2011

2 Rothstein, 2010; Phillips, Brooks-Gunn, Duncan, Klebanov, & Crane, 1998

3 Bryk, Sebring, Allensworth, Luppescu, & Easton, 2010

City Connects Progress Report 2014

geographic neighborhood in Boston; CCNX was replicated in another BPS

neighborhood (five new schools) in 2007. In September of 2010-11, at the

invitation of the district, CCNX expanded to six “Turnaround” schools—

that is, schools officially designated by No Child Left Behind (NCLB)

standards as in the category of “Restructuring.” In 2012-13, CCNX was

implemented in a total of 16 elementary and K-8 schools in Boston.

Most

recently, in September of 2011-12, CCNX expanded to five Springfield, MA

elementary and K-8 schools, and added three Springfield middle schools

in September of 2012-13. Access to data for analyses has also expanded

over time, allowing us to increase the number of comparison schools

(previously seven randomly-chosen comparison schools in Boston and

now including all schools not implementing CCNX in Boston). Through

the collaboration of the Springfield, MA district, student data is also now

available for Springfield.

This report summarizes the quantitative and qualitative outcomes of the

CCNX ongoing evaluation in Boston. Previous findings demonstrated the

significant impact of the CCNX intervention, across K-5 grade levels, on

academic achievement and measures of student thriving. These positive

findings are particularly pronounced for English Language Learners

and extend beyond elementary school, after students leave the CCNX

intervention, into middle school and high school. See previous reports

at www.cityconnects.org. Our appendices for this report and past reports

present more detailed information about the CCNX intervention, its

phased rollout in Boston Public Schools (BPS), and the demographic

context of its implementation. The data sources and methodologies

employed and the full results of the quantitative and qualitative analyses

of those data are described more fully in the appendices.

In addition, for the first time, this report presents findings on the

implementation of City Connects in Springfield, MA. Information on

the context of implementation in Springfield and on such intermediate

outcomes as reviews and services delivered are presented in this report.

Because data from only one year of implementation are available to us at

this time, outcomes analyses drew on the 2011-12 data.

For Boston, in this report, we present selected new analyses drawing on

the new data available to us in 2011-12. Quantitative analyses drew on

a rich variety of sources, including report card scores, state test scores,

student and teacher surveys, and publicly available demographic data.

4 A pilot expansion of City Connects in high school was also started at Quincy Upper School.

The total number of schools for 2012-13 diers from the cumulative total because of school

closures and mergers.

As with Springfield, the most recent data from Boston Public Schools

available at this time are from 2011-12 and earlier. In order to supplement

and illuminate the quantitative data, CCNX also rigorously analyzed

qualitative data from key participants at the heart of the intervention:

teachers, principals, and community partners. Qualitative data were

gathered and analyzed in academic year 2012-13.

We begin with a short description of how urban poverty creates out-of-

school factors that impact student development and learning. Next, we

describe current approaches to student support and how they compare

with “best practices.” Then we briefly outline the CCNX intervention.

Next, we describe the expansion of CCNX to Springfield, comparing the

context of implementation there and in Boston. We summarize the reviews,

services delivered, and partnerships in both implementation sites. Then

we present selected new and previously established quantitative findings

on the impact of CCNX on academic achievement and on factors related to

thriving, school success, and life chances. Finally, we present data on the

impact of CCNX on principals, teachers, and community agencies.

The Impact of Urban Poverty on

Children’s Development and Learning

The pervasive effects of poverty on academic achievement underscore the

importance of addressing out-of-school factors in any education reform

effort.

Poverty impacts children’s achievement and growth in at least three

noteworthy ways: 1) limits investment—a family’s ability to invest money,

time, and energy in fostering children’s growth (e.g., less time to read and talk

with their children); 2) creates pervasive stress within families and their

neighborhoods—this undermines children’s sense of well-being and safety

(e.g., inconsistent parenting behavior or increased exposure to community

violence that may undermine children’s self-regulation and social-emotional

stability); and 3) contributes to chaotic life—unpredictable support systems

(e.g., less-reliable transportation, municipal services, and businesses).

For children living in poverty, the impact of out-of-school factors is clearly

evident in their ability to succeed in school. Limited resources, stress,

and the chaos of poverty result in poor attendance, high mobility, social-

emotional dysfunction, a lack of readiness for school, and limited cultural

capital to understand schools as institutions.

Rothstein describes the

impact on achievement of out-of-school factors relative to in-school factors:

5 See Walsh & Murphy, 2003; Berliner, 2009; and Rothstein, 2010.

6 Dearing, 2008

City Connects Progress Report 2014

“Decades of social science research have demonstrated that

differences in the quality of schools can explain about one-third

of the variation in student achievement. But the other two-thirds

is attributable to non-school factors” (emphasis added).

Figure 1 illustrates that academic success is predicated on children’s

readiness to engage and thrive in school. It also shows the overlapping

impact of the various domains of development on children’s readiness to

learn and thrive.

Figure 1. Academic success is predicated

on students’ readiness to engage and thrive

in school

In all areas in Figure 1, the harm inflicted by poverty on students’

readiness to engage in school is complex and dynamic, because poverty

itself is dynamic. The manifestations of childhood poverty are not only

pernicious; they also interact, influencing one another. For example, Coley

and Baker (2013) explore in depth how childhood poverty manifests itself

in family and parenting behaviors; exposure to environmental toxins,

access to adequate health insurance, food insecurity, parent employment,

and access to (and quality of) child care.

The impact of poverty in each of

these areas affects the others.

Richard Rothstein argues, “if we want to raise the achievement of

disadvantaged children substantially in our own country, we will

have to improve the collection of interacting and mutually reinforcing

characteristics” that impact children living in poverty.

To succeed in

7 Rothstein, 2010, p. 1

8 Coley and Baker, 2013. As Coley and Baker note, this list is not exhaustive.

9 Rothstein, 2013

addressing poverty’s effect on achievement and thriving, an intervention

in schools must be similarly dynamic and multi-faceted.

Current Models of Student Support

Many schools currently are unable to respond to the pressing challenges

facing students’ out-of-school lives. Student support structures are the

product of an earlier time, a different set of needs, and a less diverse

demographic. The typical approach to student support in most schools:

1) is fragmented and idiosyncratic, serving a small number of high-need

students; 2) does not address the full range of needs, focusing mainly on

risk; 3) does not collect data on the effectiveness of the supports offered

students; and 4) in practice, does not operate as a core function of the

school, and as a result, seeks minimal teacher engagement.

Best Practices in Student Support

Grounded in research on child development and the need that it be

implemented as a core function of schools, optimized student support

has six identifying characteristics. It is: 1) customized to the unique

strengths, needs, and interests of each student; 2) comprehensive, serving

the academic, social/emotional, health, and family needs of all students

from a variety of cultural and ethnic backgrounds; 3) coordinated

among families, schools, and community agencies; 4) cost-effective to

schools by leveraging the resources provided by community agencies;

5) continuously monitored for effectiveness through collecting and

analyzing data to evaluate and improve service delivery and student

outcomes; and 6) implemented in all sites with fidelity and oversight.

10 Walsh & DePaul, 2008

City Connects Progress Report 2014

The City Connects Model

Partners

Built on the best practices described above, CCNX is a partnership

delivering optimized student support. Figure 2 shows the partners - the

Boston and Springfield Public Schools, a wide range of community

agencies, and Boston College. Boston College is the nerve center of

CCNX. The Center for Optimized Student Support at Boston College

developed and delivers the CCNX intervention and is the home of the

Leadership and Implementation Teams. The Boston College Center for

the Study of Testing, Evaluation, and Education Policy is the home of the

Evaluation Team.

Figure 2. The CCNX partnership

Rationale

Figure 3 provides the rationale that

underpins the CCNX intervention.

City Connects Progress Report 2014

History of the City Connects intervention

In the academic year 2001-02, CCNX was initially implemented in six

schools located in one geographic neighborhood (BPS Cluster 5, which

includes Allston, Brighton, and Mission Hill sections of the city). An

external funder, who provided a planning grant in 1999, stipulated that

development and design of CCNX take place in Cluster 5. In 2007, the

District stipulated that expansion of CCNX occur in BPS Cluster 2 (the

North End, South End, and Lower Roxbury), adding five new schools.

At that time, seven schools from other BPS clusters were randomly

chosen to serve as comparison schools. CCNX and comparison schools

are our “legacy schools.”

By this we mean that the students from these

schools are being followed longitudinally from kindergarten through

high school to assess the long term impact of the CCNX intervention. In

September of 2010-11, at the invitation of the district, CCNX expanded to six

“Turnaround” (or “Transformation”) schools—that is, schools officially

designated by No Child Left Behind (NCLB) standards as in the category

of “Restructuring.” The comparison data set now includes all schools in

Boston not implementing CCNX. In September of 2011-12, CCNX expanded

to its first Massachusetts public school system outside Boston, and as of

September 2012-13 was implemented in five elementary and K-8 schools

and three middle schools in Springfield, MA.

Description of the City Connects intervention

CCNX connects students with the individually tailored set of prevention,

intervention, and enrichment services that they need to succeed in school.

There are six key components of the model:

School Site Coordinator. At the core of the intervention is a full-time

School Site Coordinator (SSC) in each school, trained as a school counselor

or school social worker, who connects students to a customized set of

services through collaboration with families, teachers, school staff, and

community agencies. The ratio of SSC to student population is 1:400. The

SSC follows standardized practices codified in the CCNX Practice Manual,

schematized in Figure 4 and detailed in the components below.

11 It is important to note that during the history of CCNX implementation, there have been

several school closings and mergers, which is a common fact of life in any urban school

district.

12 Springeld is the third-largest city in Massachusetts.

Figure 4. City Connects student support process

Whole Class Review. The SSC works with each classroom teacher to

review every student in the class and develop customized support plans

that addresses their individual strengths and needs. There are five aspects

of the Whole Class Review (WCR): 1) identifying the strengths and needs of

each student across four domains (academic, social/emotional/behavioral,

health, and family); 2) identifying and locating appropriate school- and/

or community-based services and enrichments; 3) establishing the

connection between these service providers

and individual children and their families;

4) documenting and tracking the delivery

of the service, and 5) following up to ensure

appropriateness of fit.

As they conduct the WCR, at the most general

level, the teacher and SSC group the students

in a class into three tiers: strengths and

minimal risk (Tier 1); strengths and mild

to moderate risk (Tier 2); or strengths and

severe risk (Tier 3). Tier 2 is divided into two

levels: 2a (mild risk) and 2b (moderate risk).

Figure 5. Tiers in the CCNX triangle

TIER 1

Strengths and

Minimal Risk

TIER 2a:

Strengths and Mild Risk

TIER 3

Strengths and

Severe Risk

TIER 2b:

Strengths and

Moderate Risk

City Connects Progress Report 2014

Individual Student Review. Students identified as having intensive needs,

at any point during the school year, receive an Individual Student Review

(ISR). This review is independent and distinct from a Special Education

referral. A wider team of professionals discuss and develop specific

measureable goals and strategies for the student. The ISR is conducted by

the student support team—an existing school structure that can include

school psychologists, teachers, principals, nurses, and occasional community

agency staff members and that is typically led by the SSC. The School Site

Coordinator communicates with the family before and after the ISR.

Community agency partnerships. A critical aspect of the role of the

SSC is developing and maintaining partnerships with community agencies

and institutions. These relationships are formalized through a CCNX

Community Resource Advisory Board, comprised of selected citywide

agency leaders, and a CCNX Resource Advisory Council, which includes

selected agency representatives working at the local neighborhood level.

Connecting students to services, tracking, and following up. During

and after the conversations with teachers, school staff and leaders, and

community agency representatives, CCNX School Site Coordinators

connect each student to the particular enrichment and service programs

that best meet his or her strengths and needs. School Site Coordinators

work closely with families as students are referred and connected to

particular enrichments and services. To aid with the process, and to

permit streamlined tracking and follow-up, CCNX has developed a

proprietary Web-based database, Student Support Information System

(SSIS). The SSIS allows for secure collection of data on student reviews,

individual student plans, service referrals, and providers (both school-

based and community agencies) who deliver services. SSIS data are used

for three purposes: 1) record-keeping at the individual and school level;

2) monitoring and evaluating the implementation of the intervention

throughout the school year; and 3) conducting research on the

effectiveness of the intervention.

The tailoring of services is accomplished through different combinations of

quantity and type of services, resulting in a unique set of services for each

student. Services may be prevention and enrichment in nature, including

before- and after-school programs, sports, summer programs, and health

and wellness classes; early intervention services such as adult mentoring,

academic support, social skills interventions, family assistance, and

tutoring; or more intensive services or crisis interventions such as mental

health counseling, health services, screening or diagnostic testing, violence

intervention, or family counseling. For any single student, regardless

of tier, the tailored set might include a combination of prevention and

enrichment, early intervention, and/or intensive services.

Providing specific services within the school. In response to specific

needs, School Site Coordinators provide the following services within the

school and classrooms: 1) leading small social skills groups on a time-

limited basis that address focused topics such as making friends, bullying,

and healthy eating; 2) crisis intervention for individual or small groups of

children; and 3) family outreach and support addressing specific family

needs that are impacting the child’s performance in school.

Implementation of City Connects in Boston

and Springeld

Expansion of City Connects to Springfield, MA

The expansion to the Springfield, MA public schools followed the CCNX

process for entering a district.

Program introduction. In a series of collaborative discussions, the CCNX

team presented the model to the district’s extended leadership team and

school principals. A key member of the Springfield leadership team, the

Program Manager, who oversees the work of the School Site Coordinators,

helped lead discussions.

Recruiting. Using guidance and materials from CCNX, the Program

Manager recruited School Site Coordinator candidates to serve in the

schools. District leaders participated in this process, and the needed School

Site Coordinators were hired.

Needs assessment. Through surveys with teachers, community partners,

principals, and families, CCNX led a process of discovery in which the

major needs of the district in the area of student support were identified.

Results were presented to the district.

Environmental scan. Local agencies, institutions, and other community

partners that might be able to serve students in the participating schools

were identified and categorized.

Launch of Professional Development. Newly hired School Site

Coordinators were inducted in a weeklong August Institute, where they

gained familiarity with the model and began building the professional

network that was further strengthened during yearlong in-service

professional development meetings.

City Connects Progress Report 2014

Launch in schools. During meetings with school principals, the results

of the needs assessment and environmental scans were discussed in light

of the school’s priorities. A signed Memorandum of Understanding was

provided by each school’s principal, who also assisted in preparing schools

for the arrival of the new School Site Coordinators (e.g., setting up office

space and technology).

Plan for evaluation. CCNX obtained district approval on the evaluation

design and reporting structures.

With the completion of these steps, the CCNX School Site Coordinators,

the Program Manager, and the district and schools were ready to begin

the work of CCNX in Springfield. Throughout the first two years of

implementation, indicators from the CCNX Fidelity Monitoring System

revealed areas of high program fidelity (such as strong implementation of

preparatory steps for Whole Class Review, the collaborative assessment

of each student’s strengths and needs).

These indicators also assisted the

Program Manager by highlighting areas of potential improvement or need

(such as the need to support teachers in filling out Whole Class Review

forms during the first year of implementation).

Drawing on feedback over the course of the year from the fidelity indicators,

the Springfield Program Manager collaborated with School Site Coordinators

to arrive at high fidelity at year’s end in the areas of student reviews and

service delivery. Highlights of program fidelity at the end of 2012-13 include:

• In 2012-13, 100% of students in Springfield CCNX schools

received a Whole Class Review and had complete data entry in

the electronic tracking system. In the experience of CCNX, this

percentage is unusually high for a district in only its second year

of implementation. It matched the percentage for Boston in 2012-13

(100%).

• Springfield School Site Coordinators were similarly successful with

their implementation of the practice of Individual Student Review,

intensive reviews by a team for students most at risk. Across the

Springfield schools, 6% of students in Springfield CCNX schools

received Individual Student Reviews. This percentage was the

same in Boston in 2012-13 (6%).

• Across the Springfield schools implementing CCNX, 91%

of Springfield CCNX students received at least one service.

13 For a description of the Fidelity Monitoring System, see the City Connects 2012 Progress Report.

This section presents results from selected indicators only.

Additionally, more than 80% of the students received three or

more district- or community-provided services. These percentages

also align with program targets and are similar to percentages in

Boston, where 98% of students received at least one service and

81% received three or more services.

• By the end of 2012-13, Springfield School Site Coordinators had

established and cultivated collaborations with 179 community

partners.

Boston and Springfield context

Characteristics of the public schools of Boston and Springfield are

important to interpreting and understanding the challenges CCNX

students face and the impact of the intervention. Previous CCNX reports

present a detailed overview of the social and economic disadvantages

faced by many Boston residents. This year, we compare the context of

implementation across the districts of Boston and Springfield.

CCNX was implemented in sixteen Boston Public Schools (totaling 6,845

students) and five Springfield Public Schools in 2012-13 (2,732 students). Six

of the Boston schools and one Springfield school were K-8 schools; three

Springfield schools were middle schools where implementation focused

on grade 6; all others were K-5. Table 1 presents a summary of elementary

school (grades K to 5) student characteristics for CCNX schools and all

schools in each district not implementing CCNX during school year 2011-

12, the most recent year for which data are available.

City Connects Progress Report 2014

Table 1. Boston and Springeld elementary school (K-5) student characteristics, 2011-12

Table 1 shows that in Boston, CCNX and comparison students were similar

across several characteristics, including gender and special education

status. However, CCNX students were more likely to be Asian than

comparison students, and there were significantly more students whose

first language is not English in CCNX schools. Significantly more students

in CCNX schools were living in poverty, as measured by eligibility for

free school lunch. There was also significantly more mobility (e.g., fewer

attended the same school as the previous year) and a higher average

number of school absences among CCNX students. Students in both CCNX

and comparison schools were less likely to be African-American than

those in charter schools, were much less likely to speak English as a first

language, and were more likely to be eligible for free school lunch.

In Springfield, CCNX and comparison students were similar in gender,

poverty status, enrollment in Special Education, and mobility. However,

CCNX students were more likely to be Hispanic/Latino, were more likely

to be English Language Learners, and had higher average number of

school absences.

Boston Springeld

Charter Schools Comparison

Schools

City Connects Comparison

Schools

City Connects

% Female 51.3 48.3 47.9 48.3 48.1

Race/Ethnicity

% Black 60.6 33.2 31.5 19.7 16.9

% White 8.2 13.8 7.2 15.8 7.8

% Asian 2.3 5.6 10.8 2.3 2.4

% Hispanic 26.2 44.8 48.0 58.7 70.2

% Multi-Race Non Hispanic/Other 2.8 2.6 2.5 3.5 2.6

% First Language Not English 18.8 40.7 49.1 21.3 33.9

% English Language Learners 7.2 14.9 18.1 16.9 29.3

Poverty: Eligible for Lunch Subsidy

% Reduced Price Lunch 14.1 4.6 3.8 3.9 3.1

% Free Lunch 61.6 67.8 76.3 84.0 88.5

% Special Education 10.8 17.7 21.0 10.4 10.1

Mobility (% Attending Same School) 96.4 78.3 69.4 87.7 89.2

Average Number of School Absences 8.3 6.8 7.4 2.9 4.4

Source: Boston Public Schools and Springeld Public Schools student data for 2011-12; Massachusetts Department of

Education enrollment data for Charter Schools column.

Table 1 reveals that students in CCNX schools differed in several key

ways across Boston and Springfield. First, in Springfield, CCNX students

were more likely to be Hispanic. Second, they were even more likely

than Boston students to be eligible for free or reduced-price school lunch,

a measure of poverty (92% in Springfield vs. 80% in Boston). Finally,

mobility and school absenteeism were lower in Springfield than in Boston.

Reviews and services delivered in Boston and Springfield

In the 2012-13 school year, 100% of students in both Boston and Springfield

received a Whole Class Review. As noted above, during the WCR, the

SSC and teacher group the students in a class into three tiers: strengths

and minimal risk (Tier 1), strengths and mild to moderate risk (Tier 2); or

strengths and severe risk (Tier 3). Tier 2 is divided into two levels: 2a (mild

risk) and 2b (moderate risk). Table 2 shows the number and percentage of

students in each tier for Boston and Springfield.

Table 2. Number of students placed in each tier, Boston and Springeld

Boston Springeld

Tier 1 2192 (32%) 915 (34%)

Tier 2a 2130 (31%) 896 (33%)

Tier 2b 1620 (24%) 575 (21%)

Tier 3 871 (13%) 317 (12%)

Total 6813 2703

Source: CCNX Student Support Information System database, 2012-13

Students placed in Tier 3 are considered for an Individual Student Review

so that a team of professionals can assess strengths and needs and develop

specific, measurable goals and strategies. (See the full description of

ISRs above.) In 2012-13, the number of ISRs was 331 in Boston and 154 in

Springfield.

In both districts, SSCs developed and maintained relationships with

community agencies that offer services to students. As noted above,

services range from prevention and enrichment to early intervention to

intensive intervention. In 2012-13, CCNX worked with 380 community

14 The total N for Table 2 is slightly smaller than the total number of students in CCNX schools

because the table does not include students whose Whole Class Review record lacked a tier

(32 students in Boston and 29 in Springeld).

City Connects Progress Report 2014

partners in Boston and 179 in Springfield. Along with school and district

service providers, these partners delivered approximately 39,000 services

in Boston and 21,000 in Springfield. Figures 6 and 7 show the numbers and

percentages of services delivered across categories.

Figure 6. Total number of services delivered to students, by service category, Boston Public

Schools

Source: CCNX Student Support Information System database, 2012-13. School Site Coordinators noted an additional

7,960 health screenings that were delivered by school nurses.

Service N Category % Total %

CATEGORY 1

(Prevention

and Enrichment)

Before School Program 209 2%

37%

Enrichment: Arts 2706 20%

Sports/Physical Activity 5221 39%

Enrichment: Youth Development 1166 9%

Enrichment: Academic 3540 26%

Violence Prevention 123 1%

New Balance Health & Wellness Curricula 496 4%

Category Total 13461 100%

CATEGORY 1.5 After School Program 1279 50%

Summer Program 812 32%

Vacation Program 469 18%

Category Total 2560 100%

CATEGORY 2

(Early Intervention)

Behavior Plan Special Observation 236 2%

29%

Classroom-based Social Skills Intervention 2992 28%

Adult Mentoring 704 7%

Psycho-social Group 316 3%

Academic Support 3113 29%

ESL 25 <1%

Classroom-based Health Intervention 3200 30%

Category Total 10586 100%

CATEGORY 2.5 Supplemental Educational Services 105 2%

18%

Tutoring 71 1%

Family Support 4238 65%

Family Assistance/Outreach 2152 33%

Category Total 6566 100%

CATEGORY 3

(Intensive / Crisis

Intervention)

Check-in with CCNX Site Coordinator 253 9%

Mental Health Counseling 211 7%

Informal Screening/Diagnostic 20 1%

Health/Medical 1227 43%

SPED Eval/Screening 91 3%

Crisis Intervention 88 3%

Attendance Support 960 33%

Family Counseling 13 <1%

Violence Intervention 17 1%

Category Total 2880 100%

Grand Total 36053

City Connects Progress Report 2014

Figure 7. Total number of services delivered to students, by service category,

Springeld Public Schools

Source: CCNX Student Support Information System database, 2012-13. School Site Coordinators noted an additional

3,890 health screenings that were delivered by school nurses.

Table 3 and Figures 8-9 present the distribution by tier of students

receiving different numbers of services for Boston and Springfield.

Service N Category % Total %

CATEGORY 1

(Prevention and

Enrichment)

Before School Program 25 < 1%

32%

Enrichment: Arts 2878 46%

Sports/Physical Activity 824 13%

Enrichment: Youth Development 1570 25%

Enrichment: Academic 639 10%

Violence Prevention 7 < 1%

Health & Wellness Curricula 330 5%

Category Total 6273 100%

CATEGORY 1.5 After School Program 280 66%

Summer Program 143 34%

Category Total 423 100%

CATEGORY 2

(Early

Intervention)

Behavior Plan Special Observation 105 2%

36%

Classroom-based Social Skills

Intervention

1537 22%

Adult Mentoring 663 10%

Psycho-social Group 196 3%

Academic Support 3424 49%

Classroom-based Health Intervention 996 14%

Category Total 6921 100%

CATEGORY 2.5 Additional Academic Services 38 1%

19%

Tutoring 131 4%

Family Support 2699 73%

Family Assistance 814 22%

Category Total 3682 100%

CATEGORY 3

(Intensive/

Crisis Intervention)

Check-in with CCNX Site Coordinator 208 10%

11%

Mental Health Counseling 129 6%

Informal Screening/Diagnostic 6 < 1%

Health/Medical 1222 56%

Crisis Intervention 37 2%

Attendance Support 579 27%

Family Counseling 1 < 1%

Category Total 2182 100%

Grand Total 19481

Table 3. Proportion of students in each tier receiving dierent numbers of services

Source: CCNX Student Support Information System database, 2012-13

Table 3 shows first that in both Boston and Springfield, the mean number

of services per student is smallest at Tier 1 and largest at Tier 3. Second, as

shown in both Table 3 and Figures 8-9, the proportion of students receiving

1-2 services is highest for Tier 1 students and lowest for Tier 3. Third, the

corresponding proportions for 5 or more services are the mirror image: the

proportion of students receiving 5 or more services is smallest for Tier 1

and largest for Tier 3.

Figure 8. Proportion of students in each tier receiving 1-2, 3-4, or 5 or more services, Boston

Source: CCNX Student Support Information System database, 2012-13

15 The total N for Table 3 is slightly smaller than the total number of students in CCNX schools

because the table does not include students whose Whole Class Review record lacked a tier

(32 students in Boston and 29 in Springeld).

Number of Services by Tier 1-2 Services 3-4 Services 5+ Services Total

Student N

Receiving

Services

Total

Student %

Receiving

Services

Student

Mean # of

Services

Standard

Deviation

Student

Row % Student

Row N % Student N Row N %

Boston

Public

Schools

Tier1 2192 5.16 3.76 459 21.2% 668 30.9% 1034 47.8% 2161 99%

Tier 2a 2130 5.56 3.88 368 17.6% 578 27.7% 1139 54.6% 2085 98%

Tier 2b 1620 5.90 4.05 211 13.4% 419 26.5% 949 60.1% 1579 97%

Tier 3 871 7.09 4.78 81 9.5% 155 18.2% 615 72.3% 851 98%

Total 6813 5.71 4.05 1119 16.8% 1820 27.3% 3737 56.0% 6676 98%

Springeld

Public

Schools

Tier1 915 6.02 6.46 107 13.4% 235 29.4% 456 57.1% 798 87%

Tier 2a 896 7.93 7.55 89 10.6% 194 23.1% 556 66.3% 839 94%

Tier 2b 575 8.83 7.71 38 6.9% 103 18.8% 407 74.3% 548 95%

Tier 3 317 9.68 7.98 22 7.1% 52 16.9% 234 76.0% 308 97%

Total 2703 7.68 7.40 256 10.3% 584 23.4% 1653 66.3% 2493 92%

0%

10%

20%

30%

40%

50%

60%

70%

80%

90%

100%

Tier1 Tier2a Tier2b Tier3

5+Services

34Services

12Services

City Connects Progress Report 2014

Figure 9. Proportion of students in each tier receiving 1-2, 3-4, or 5 or more services, Springeld

Source: CCNX Student Support Information System database, 2012-13

Impact on Students

Earlier reports have documented the beneficial effects of CCNX on student

achievement and thriving, as summarized in the preface to this report.

Here, we review two of these earlier outcomes: the beneficial effect of

CCNX on students’ achievement of proficiency on the Massachusetts

statewide test and lower rates of chronic absenteeism. It is notable that

these effects document long-term improvements in indicators of academic

achievement and life chances, showing that students enrolled in CCNX

schools benefit long after they have left the intervention itself.

Next, we present new findings from the most recent analyses. First, an

updated analysis of high school dropout that included two additional

cohorts of data confirmed earlier findings and also included a new

calculation: attendance in a CCNX elementary school from kindergarten

on leads to approximately 50% lower probability of dropout in high school.

Second, students enrolled in CCNX schools significantly outperform their

comparison-school peers on the Stanford Achievement Test in elementary

school. The positive finding is seen for all students and also, notably, for

immigrant students. Third, students previously enrolled in CCNX are

more likely to attend one of three selective public high schools (“exam

schools”) in Boston—an indicator of academic success. Finally, the first

analysis of student outcomes for Springfield, MA demonstrates a positive

effect: the gap in statewide test scores between the CCNX Transformation

schools and other Springfield schools narrowed in 2011-12, the first year of

implementation of CCNX.

0%

10%

20%

30%

40%

50%

60%

70%

80%

90%

100%

Tier1

Tier2a

Tier2b

Tier3

5+Services

34Services

12Services

Improving standardized test scores in middle school

(Massachusetts Comprehensive Assessment System,

or MCAS)

As reported in the 2010 CCNX Annual Report, the analysis of MCAS mean

scores relative to comparison schools yields encouraging results that show

positive effects of CCNX. Analysis of scores by results category provides

corroborating positive evidence. Students’ MCAS scores are classified into

four categories: advanced, proficient, needs improvement, and warning/

failing. An analysis comparing the percentage of CCNX students scoring

proficient or above in ELA and Math yields the results shown in Figures

10 and 11. Figure 12 shows the percentage of ELL students in CCNX and

comparison schools scoring proficient or above on the ELA test, relative

to overall statewide scores. Here and elsewhere in this report, the vertical

dotted line represents the point at which students leave CCNX and move

on to middle school.

Figure 10. Percentage of students scoring at procient or above, MCAS English Language Arts

Source: CCNX and comparison schools: Boston Public Schools MCAS data, 2003-2009. Boston Public Schools and state

data: Massachusetts Department of Elementary and Secondary Education.

20%

30%

40%

50%

60%

70%

80%

3 4 5 6 7 8

%Proﬁcient+

Grade

StateOverall(~31%LowIncome)

BPSOverall(~83%LowIncome)

CCNXOverall(~83%LowIncome)

City Connects Progress Report 2014

0%

10%

20%

30%

40%

50%

60%

70%

80%

90%

3 4 5 6 7 8

%Proﬁcient+

Grade

StateOverall

CityConnectsEnglish

LanguageLearners

ComparisonEnglish

LanguageLearners

Figure 11. Percentage of students scoring at procient or above, MCAS Math

Source: CCNX and comparison schools: Boston Public Schools MCAS data, 2003-2009. Boston Public Schools and state

data: Massachusetts Department of Elementary and Secondary Education.

Figure 12. Percentage scoring at procient or above, MCAS English Language Arts: CCNX ELL

students, comparison school ELL students, and all students statewide

Source: CCNX: Boston Public Schools MCAS data, 2003-2009. State data: Massachusetts Department of Elementary

and Secondary Education.

10%

20%

30%

40%

50%

60%

4 5 6 7 8

%Proﬁcient+

Grade

StateOverall(~31%LowIncome)

BPSOverall(~83%LowIncome)

CCNXOverall(~83%LowIncome)

• Figures 10 and 11 show that CCNX students outperform both

students from the comparison schools and their Boston peers in

middle school and achieve close to state proficiency levels on both

English and Math on MCAS.

• Figure 12 shows that ELL students in CCNX achieve gains that

move them close to statewide levels of proficiency in the MCAS

ELA test by grade 8. The positive impact of CCNX is seen for

students particularly at risk for literacy outcomes.

Preventing chronic absenteeism

High rates of absence from school are an important predictor of academic

risk and dropout. As reported in the 2012 CCNX Progress Report, students

who attended CCNX schools in elementary school are significantly less

likely to be chronically absent (defined as being absent for 10% or more of

the days within the school year) than students who never attended CCNX

schools. In this section, we provide details on the analysis and findings.

Students included in the analysis and analytic techniques

The analysis drew on students’ longitudinal data record (i.e., records of the

student’s absences within and across years). A given student’s data may be

represented at more than one grade level.

Descriptive analyses, including t-tests, were used to examine the present

and absent days in grades 1-12, overall and by treatment group. Next,

hierarchical linear models were estimated to examine longitudinal

changes in student absenteeism across grade levels.

Results

Figure 13 presents the longitudinal change (or estimated probabilities) in

chronic absenteeism for the CCNX and comparison groups.

City Connects Progress Report 2014

Figure 13. Proportion of students who were chronically absent, CCNX vs. comparison students.

Source: Boston Public Schools school attendance data, 2001-2009

• Although CCNX students start out with higher rates of chronic

absenteeism in grade 1, rates of chronic absenteeism were

significantly lower than comparison students in all middle and

high school grades 6-12, except for grade 10.

• Beyond chronic absenteeism, CCNX students were found to have a

significantly lower total number of days absent than students from

the comparison group in grades 4 to 12.

Preventing school dropout

As reported in the 2012 CCNX Progress Report, students who attended

CCNX schools in elementary school are significantly less likely to drop out

of school. A new analysis adding two more cohorts of data confirmed this

earlier finding. In this update, we present both average effects on dropout

in each high school grade and also a cumulative effect on dropout rate

across the years of high school.

The school-level proxies for dropout typically used in education program

evaluations have been criticized as misrepresenting true dropout rates.

These proxies may produce biased estimates of true dropout rates because

they are solely based on aggregate counts of the number of students enrolled

in a school at the beginning and end of high school, even though a number

of factors other than dropout contribute to enrollment figures. In contrast,

we directly examine student-level, longitudinal enrollment trajectories.

0%

2%

4%

6%

8%

10%

12%

14%

16%

18%

20%

1 2 3 4 5 6 7 8 9 10 11 12

Propor%onofstudentswhoarechronicallyabsent

GradeLevel

CCNX

Comparison

Through this process, we are able to account for each student’s reason

for disenrollment and produce a more precise measure of dropout than

aggregate proxies. Before presenting the findings, we briefly explain which

students are included in the analysis sample and provide some background

on how students were classified as dropouts or non-dropouts.

Students included in the analysis

For a student’s data to be used in the analysis, several conditions needed to

be met:

• The student was enrolled in Boston Public Schools (BPS) prior to

the completion of grade 5 and also at the start of high school.

• The student was not enrolled in a substantially separate Special

Education classroom at any point during high school.

• The student’s longitudinal record included data on all control

variables.

Students whose records met these conditions were included in either

the CCNX group (all students ever enrolled in a CCNX school) or the

comparison group (all who had never attended a CCNX school).

How students were classified as dropout or non-dropout

When students leave BPS, the reason for departure is recorded. The

analysis drew on this information to create a dichotomous dropout variable

at the repeated measures level for each student reflecting whether a student

did or did not drop out at a given time point in his or her longitudinal record.

Students classified as non-dropout: Students who leave BPS for reasons

other than dropout, such as graduation or transfer to another district.

If a student does not depart BPS, but his or her longitudinal record does

not reach grade 12 simply because the student is not old enough to have

completed high school, no withdrawal information appears in the record.

These students are also included in the non-dropout group.

Students classified as dropout: Students who (1) withdraw from BPS

entirely; (2) never return to BPS; and (3) have a record that clearly indicates

non-graduation (such as drop out, pregnant, expelled, or incarcerated).

16 Control variables include race, gender, ever eligible for free/reduced priced lunch, ever

classied as an English Language Learner, ever eligible for Special Education services,

total number of school transfers experienced since kindergarten, and grade level at end of

longitudinal BPS record.

17 Full description of the information in district school withdrawal records are provided in the appendices.

City Connects Progress Report 2014

Modeling dropout

The CCNX effect on dropout is modeled using discrete event history

analysis: repeated measures are nested within students using hierarchical

logistic regression, where repeated measures and student-level

characteristics serve as controls.

Results

This analysis finds that comparison students (those who never attended a

CCNX school) are more likely to drop out than students who had attended

CCNX schools in elementary school from kindergarten on; see Figure 14.

Figure 14. Proportion of students who drop out from school at each high school grade level,

comparison vs. CCNX students

Proportions adjusted for demographic student characteristics. Source: District withdrawal code data, 2004-2009.

Comparison N= 19,979; CCNX N=2,265

As shown in the line graph above, at every grade level, students who

attended CCNX elementary schools from kindergarten on are less likely to

drop out of school. The difference at grade 9 is particularly notable (almost

6% for comparison students and 3% for CCNX students).

The cumulative percentage of students who drop out across the four years

of high school is also significantly lower for students who attended an

elementary school implementing CCNX from kindergarten on than for

those who never attended a CCNX school, as shown in Figure 15.

0%

2%

4%

6%

8%

9 10 11 12

Grade

Comparison

CityConnects

Figure 15. Cumulative percentage of students who drop out from high school, comparison vs.

CCNX students

Proportions adjusted for demographic student characteristics. Source: District withdrawal code data, 2004-2009.

Comparison N= 19,979; CCNX N=2,265

• As shown in Figure 15, the adjusted dropout rate for students

who attended comparison schools was about 15%, compared with

8% for students who attended CCNX elementary schools from

kindergarten on.

• For students who started CCNX in kindergarten, the difference

between students who attended CCNX elementary schools and

comparison students translates to 50% lower odds of dropping

out between grades 9 and 12.

• If an entire district experienced dropout at a rate similar to that of

CCNX students, for a cohort of 5,000 students, approximately 358

fewer students would have dropped out of high school.

High school graduation is widely argued to yield public economic benefits,

including higher tax revenue and lower spending on the justice system,

healthcare, and public assistance programs. A conservative estimate of

the benefit is $127,000 per graduate.

Assuming this estimate, if a district

with a cohort of 5,000 had experienced dropout at a rate similar to CCNX

students, the public benefit would have exceeded $45 million.

18 Levin, Beleld, Muennig, & Rouse, 2006

0%

5%

10%

15%

20%

25%

Comparisonstudents

CityConnects

City Connects Progress Report 2014

Improving standardized test scores in elementary school

(Stanford Achievement Test)

In past years, CCNX has documented a significant positive effect on

report card scores in elementary schools. Similar positive effects were

seen on middle school, but not elementary school, scores for the high-

stakes statewide standardized achievement test, the Massachusetts

Comprehensive Assessment System (MCAS). In new analyses, we studied

the effect of CCNX on scores for a different standardized test—the Stanford

Achievement Test, version 9 (SAT-9). Prior to 2006, elementary school

students in Boston completed the SAT-9 to determine eligibility for

advanced work class. Although the test was not required of all students,

scores were available for most Boston elementary school students

(over 88% of students) for 2002-03 through 2005-06. While advanced class

placement is important for many students, performance on the SAT-9 is

not used by schools to make important decisions such as promotion for all

students, or for teacher evaluation. For “low-stakes” tests like the SAT-9,

teachers tend not to teach to the test—and thus the outcomes represent

more generalized academic skills. Nonetheless, long before high school,

scores on the SAT-9 have proven to predict high school graduation rates. In

fact, there is some evidence that fewer than half of students below the 50th

percentile on the SAT-9 during middle school (sixth, seventh and eighth

grade) later graduate from high school, while nearly 75% of those above the

50th percentile graduate.

Students included in the analysis. The analyses drew on students’

longitudinal data (i.e., records of student test scores within and across

years). For a student to be included, SAT-9 scores must have been

available for at least one year. Also, Grade 1 Fall report card scores and

student demographic characteristic data were required. CCNX students

in the models were required to be enrolled in CCNX during kindergarten

or first grade, and stayed in CCNX through grade 5. Comparison students

included in these analyses were all BPS students who never attended a

CCNX school.

Analytic methods and results. Analyses compared CCNX and

comparison student SAT-9 Reading and Mathematics performance. Grade

3, 4, and 5 unadjusted SAT-9 Reading and Math scores were significantly

higher for CCNX students than for comparison students (see Table 4).

19 http://www.attendancecounts.org/wordpress/wp-content/uploads/2010/04/LAUSD-

Study-2008.pdf

Table 4. Unadjusted SAT-9 Reading and Mathematics scores, CCNX vs. comparison students

* All within grade mean dierences are signicant, p<.001. Source: Boston Public Schools SAT-9 data, 2003-04 –

2008-09

Next, hierarchical linear models were used to estimate longitudinal

changes in test scores across grade levels. Generalized propensity scores

and covariate adjustments were used to control for baseline differences

between treatment groups.

In these models, CCNX was found to have a significantly positive effect on

SAT-9 scores in Reading and Math at every grade. Figure 16 compares CCNX

and comparison student SAT-9 Reading scores in terms of effect sizes based

on multi-level models.

Figure 17 presents the same comparison for Math

scores. Each bar represents the difference between CCNX and comparison

students at the corresponding grade in effect size units. The difference

between CCNX and comparison students was largest in 3rd grade.

Figure 16. Positive CCNX eects on SAT-9 Reading scores, CCNX vs. comparison students

**p<.01. Source: Boston Public Schools SAT-9 data, 2003-04 – 2008-09.

20 See Imbens 2000. Using propensity score weights helps to minimize possible study selection

eects by adjusting for the probability of being in CCNX given baseline observed background

variables (race, gender, eligibility for free- or reduced-price school lunch, bilingual status,

special needs status, school mobility, distance from home to school, and baseline Reading,

Math, Writing, Behavior, and Work Habits report card scores were used to calculate propensity

score weights).

21 Eect sizes were calculated as the dierence between CCNX and comparison group adjusted

mean score for cases at the average level of model covariates, divided by the within-group

standard deviation of the true scores (Raudenbusch & Liu, 2001).

SAT-9 Reading* SAT-9 Math*

CCNX (N=249) Comparison

(N=5774)

CCNX (N=256) Comparison

(N=6033)

Grade 3 589.3 579.2 571.8 561.4

Grade 4 618.6 608.3 604.8 592.1

Grade 5 645.8 634.9 634.8 626.9

0.00

0.25

0.50

0.75

Grade3 Grade4 Grade5

SAT9Reading:CityConnects

EﬀectSizeinStandardDevia@onUnits

CCNXbe.er

thanComparison

**

City Connects Progress Report 2014

Figure 17. Positive CCNX eects on SAT-9 Mathematics scores, CCNX vs. comparison students

***p<.001; **p<.01. Source: Boston Public Schools SAT-9 data, 2003-04 – 2008-09

In addition to examining all students, we conducted focused analyses on

first-generation immigrant students. Again, hierarchical linear models

were used with generalized propensity scores and covariate adjustments to

control for baseline differences between treatment groups, which we refer

to below as “Model 1.” However, because matching immigrant students on

baseline characteristics is more difficult than matching within the entire

sample of students, we also used school fixed effects methods to control

for baseline differences.

We refer to the school fixed effects analyses as

“Model 2.”

Figure 18 compares CCNX and comparison student SAT-9 Reading scores

in terms of effect sizes based on multi-level models. Each bar represents

the difference between CCNX and comparison students at fifth grade in

effect size units.

22 For some schools with large numbers of immigrant students, we were able to compare

achievement in the school prior to the introduction of the intervention to achievement in

that same school after the intervention had been introduced.

23 The immigrant analyses are among the most recent additions to the evaluation work. The rst

analyses focused on fth grade because it represents the end point of the elementary school

intervention. In future work, we will examine the eects on grades 3-4.









Grade3

Grade4

Grade5





CCNXbe.er

thanComparison

***

**

Figure 18. Positive CCNX eects on SAT-9 Reading scores for fth grade immigrant students

**p<.01. Source: Boston Public School SAT-9 data, 2003-04 – 2008-09

Using Model 1 (propensity score weights), although not statistically

significant, the difference between CCNX and comparison immigrant

students’ reading scores approached 25% of a standard deviation, favoring

those in CCNX schools. Using Model 2 (school fixed effects), the difference

was statistically significant and nearly twice as large (approaching 50% of

a standard deviation).

Figure 19 presents the same comparison for Mathematics.

Figure 19. Positive CCNX eects on SAT-9 Mathematics scores for fth grade immigrant students

**p<.01; *p<.1. Source: Boston Public School SAT-9 data, 2003-04 – 2008-09

Similar results held for math scores. Using Model 1 (propensity score

weights), the difference between CCNX and comparison immigrant

students’ scores was approximately 15% of a standard deviation. Using

Model 2 (school fixed effects), the difference was statistically significant

and was nearly 50% of a standard deviation.

0.00

0.25

0.50

0.75

PropensityWeights SchoolFixedEﬀects

SAT9Mathema-cs:City

ConnectsEﬀectSizeinStandard

Devia-onUnits

CCNXbe@er

thanComparison

*

**

0.00

0.25

0.50

0.75

PropensityWeights

SchoolFixedEﬀects

SAT9Reading:CityConnects

EﬀectSizeinStandardDevia;on

Units

CCNXbe@er

thanComparison

**

City Connects Progress Report 2014

Achievement gains on the SAT-9 were of great practical significance: even

the smallest effect sizes on the SAT-9 were equivalent to moving students

from the 44th to 51st percentile, and the largest effect sizes were equivalent

to moving immigrant students from the 34th to the 51st percentile.

Improving exam school attendance

Attending one of the three selective public secondary schools in Boston

known as “exam schools” has been viewed as an indicator of academic

success.

A new analysis examined the relationship between attendance at

a CCNX elementary school and later Boston exam school attendance.

Students included in the analysis. The analysis drew on students’

longitudinal data (i.e., records of student test scores within and across

years). Students whose longitudinal records extended to at least grade

7 were eligible for inclusion. Students with severe special needs were

excluded from the sample, although other students enrolled in Special

Education were included. A total of 1107 students attending one of 13

elementary or K-8 schools that implemented CCNX were eligible for the

treatment group. The comparison group included 6058 students in the

same grades attending non-CCNX Boston schools. In addition to CCNX

participation, the number of years in a CCNX school also was studied as a

treatment variable.

Analytic methods and results. The effect of CCNX on exam school

attendance was estimated through multi-level logistic regression models

in order to take into account the nested structure of students within

schools (Raudenbush & Bryk, 2002). The dependent variable was student

exam school attendance, e.g., whether or not a student attended one of

the three exam schools in grade 7 to 9 after elementary school. Regression

models included student characteristics (gender, race, Special Education

status, bilingual status, and free or reduced lunch status) as covariates.

Generalized propensity scores and covariate adjustments were used to

control for baseline differences between treatment groups.

24 Admission to the Boston Latin School, Boston Latin Academy, and the John D. O’Bryant

School of Mathematics and Science, typically in seventh grade, is based entirely on students’

academic performance and test scores on the Independent Schools Entrance Exam (ISEE).

25 See Imbens 2000. Using propensity score weights helps to minimize possible study selection

eects by adjusting for the probability of being in CCNX given baseline observed background

variables (race, gender, eligibility for free- or reduced-price lunch, bilingual status, special

needs status, school mobility, distance from home to school, and baseline Reading and Math

report card scores were used to calculate propensity score weights).

Attending a CCNX elementary school is associated with attending an exam

school in later grades, as shown in Figure 20.

Figure 20. Estimated probabilities of attending exam school, by number of years in a CCNX

school, CCNX vs. comparison students

Source: Boston Public Schools enrollment data, 2001-02 – 2009-10

Figure 20 shows the likelihood of getting into exam schools for CCNX

and comparison students. The estimated probability of getting into exam

schools is displayed on the vertical axis. The horizontal axis represents

number of years attending a CCNX school, so that each bar displays the

probability of attending an exam school for a given number of years’

attendance in a CCNX elementary school. The red line indicates the

probability of attending exam school for comparison students who were

never enrolled in CCNX schools.

As Figure 20 shows, students in CCNX elementary schools

are, on average, more likely to attend exam schools than

comparison students. Moreover, probability of attending an

exam school increased with each additional year in a CCNX

elementary school.

0.00

0.02

0.04

0.06

0.08

0.10

0.12

0.14

0.16

0.18

0.20

1 2 3 4 5 6

YearswithCCNX

(Dosage)

%

*TheredlineindicatestheprobabilityofaGendingexamschoolforcomparisonstudents

*

City Connects Progress Report 2014

Narrowing the gap between Transformation

(Turnaround) City Connects schools and comparison

schools in Springfield

Given that the schools where CCNX is implemented in Springfield are

underperforming “Transformation” schools—i.e., there is a gap in MCAS

performance between these schools and others—evaluation analyses were

focused on whether this gap had been reduced.

We compared the gap in achievement between the five Springfield

Transformation CCNX schools and other schools both prior to (2010-11)

and after (2011-12) the implementation of CCNX. Grades 3, 4, and 5 raw

MCAS scores for students attending the five Springfield Transformation

schools that CCNX entered in 2011-12 were compared to those for all non-

CCNX students at the same grade in the district. Multi-level regression

modeling techniques were applied to take into account school clustering

effects. Student demographic variables (race, gender, eligibility for free- or

reduced-price lunch, primary language status, special educational needs

status, and school mobility) were included in the model at the individual

level and CCNX treatment was estimated at the school level.

After one year of CCNX, the gap between CCNX Transformation schools

and other schools was significantly reduced at all grades for both ELA

and Math. In 2010-11, students in the Transformation schools performed

significantly worse than other students at all grades in both ELA and

Math. However, in 2011-12, there were no statistically significant

differences between CCNX Transformation and comparison students for

MCAS ELA in grade 5 and MCAS Math grades 3, 4, and 5.

• For English Language Arts, the gap in MCAS scores between CCNX

Transformation and comparison students was reduced by 30% in

grade 3, by more than 40% in grade 4, and by more than 80% in

grade 5.

• For math, the gap in MCAS scores between CCNX Transformation

and comparison students was reduced by more than 10% in grade

3 and by almost 60% in grades 4-5. By the end of 2011-12, there

was no significant difference between CCNX Transformation and

comparison students for MCAS Math in grade 3, 4, or 5.

These results provide early evidence that the Transformation schools

really are transforming, with the help of CCNX.

Promoting health and wellness knowledge

Our evaluation examines the impact of specific services, or combinations

of services, on student outcomes. An important service offered to students

in grades 2-5 in three Boston Public Schools was the New Balance Health

and Wellness Curriculum. The curriculum was developed over the course

of several years and was newly revised in 2011-12. The curriculum was

delivered on a weekly basis in the classroom in each participating school

and included four units: Nutrition, Physical Activity, Bullying Prevention,

and Healthy Choices.

To assess the effectiveness of the curriculum, CCNX asked participating

students to complete pre- and post-test surveys for each of the four units.

Survey items measured a range of health and social competence outcome

variables, which included knowledge, attitude, and behavior related to the

content of each of the four units.

For all four units, students exposed to the curriculum demonstrated

significant pre/post gains in health knowledge.

Nutrition

The Nutrition unit included lessons on the nutritional content of different

foods, how to read nutrition labels, and how the body uses nutrients.

In addition to this knowledge content, the unit explored attitudes and

behaviors related to healthy eating.

For all grades, there was significant improvement in scores on nutrition

knowledge items from pre- to post-test. Score improvements were largest

at second and third grades; see Figure 21.

26 In all four units, both content and survey items were tailored to the developmental level of

the students. Surveys included 17-18 items for grades 2-3 and 25-29 items for grades 4-5.

City Connects Progress Report 2014

Figure 21. Nutrition knowledge, pre-test vs. post-test

***p<.001; ** p<.01

One of the nutrition behavior items asked students, “Did you eat breakfast

this morning?” Significant pre-post gains in the percentage of students

responding “yes” were seen in grades 3-5, as shown in Figure 22.

Figure 22. Nutrition behavior: Percentage of students responding “yes” to “Did you eat breakfast

this morning?”

***p<.001;** p<.01

Physical Activity

The Physical Activity unit included knowledge content on the benefits of

exercise, types of exercise, and the effects of physical activity on the body;

lessons also incorporated attitude and behavior content.

Grade 2 Grade 3 Grade 4 Grade 5

Nutrition Knowledge

Pre-test Score

Post-test Score

***

20%

40%

60%

80%

100%

Grade 2 Grade 3 Grade 4 Grade 5

Did you eat breakfast this morning?

Pre (Yes)

Post (Yes)

***

Significant pre-test to post-test gains in physical activity knowledge were

seen in all four grades, as shown in Figure 23.

Figure 23. Physical activity knowledge, pre-test vs. post-test

***p<.001; ** p<.01

Bullying Prevention

The Bullying Prevention unit included content on friendships, bullying

behavior, the experience of being bullied, and bullying prevention.

Figure 24 shows that in grades 3-5, students report significantly more

frequent experience of being bullied in the post-test.

Figure 24. Experience of being bullied scores, by grade (reverse coded—higher is better)

**p<.01; *p< .05; x=p<.10

Grade 2 Grade 3 Grade 4 Grade 5

Physical Activity Knowledge

Pre-test Score

Post-test Score

***

0.0

1.0

2.0

3.0

4.0

Grade 2 Grade 3 Grade 4 Grade 5

Experience of Being Bullied (Higher Score = Better)

Pre-test Score

Post-test Score

City Connects Progress Report 2014

Although at first glance, this appears to be a negative trend, the higher

reported levels of victimization may in fact be due to greater awareness of

what bullying is and an enhanced capacity to recognize it.

Healthy Choices

In the Healthy Choices unit, lessons include decision-making, bike and

car safety, fire safety, preventing illness from spreading, and (for grades

4-5) the effects of smoking and alcohol on the body.

Significant pre-test to post-test gains were seen for knowledge items in all

four grades, as shown in Figure 25.

Figure 25. Healthy Choices knowledge, pre-test vs. post-test

** p<.01

Grade 2 Grade 3 Grade 4 Grade 5

Healthy Choices Knowledge

Pre-test Score

Post-test Score

Impact on Schools

Principal satisfaction

In spring of 2013, CCNX surveyed principals about their satisfaction with

the program.

Of those who completed the survey, in both Boston and

Springfield, 100% reported they were satisfied with CCNX as a whole,

and with School Site Coordinator (SSC) work with students and families.

Principals were also highly satisfied with SSCs’ work with teachers: 88%

in Boston and 100% in Springfield. Another strong indicator of principal

satisfaction was that in both Boston and Springfield, 100% of principals

would recommend CCNX to a principal in another school.

In addition to being satisfied with the School Site Coordinators’ work, 86%

of Boston and 100% of Springfield principals indicated that the delivery of

student support has improved at their school as a result of CCNX.

An interesting trend emerged in the 2013 survey: increasingly, principals

cited the importance of CCNX on academic achievement and standardized

test performance. The trend is illustrated in the first rows of Table 5.

Table 5. Percentage of principals rating City Connects as helpful at addressing each area in

their schools

Source: City Connects 2012 and 2013 principal surveys

As shown in Table 5, over the past two years, a growing percentage of

principals reported they believe CCNX is helpful at addressing student

academic achievement and performance on statewide test (MCAS) scores.

Principals also recognized the helpfulness of CCNX with such important areas

as student behavior and teachers’ ability to support students in the classroom.

27 In Boston, 26 principals and assistant principals received the survey and 16 completed it; in

Springeld, 15 received it and 9 completed it.

Principal Ratings of the

Impact of CCNX on Specic

Areas

Boston

2011-12

Boston

2012-13

Springeld

2011-12

Springeld

2012-13

Student academic achievement

(i.e., grades)

82% 100% 100% 100%

Student MCAS performance 64% 100% 75% 100%

Student classroom behavior 100% 100% 100% 100%

Student health and well being 100% 100% 100% 100%

Teacher ability to focus on

instruction

90% 92% 75% 89%

Teacher ability to support

students in the classroom

100% 100% 75% 100%

City Connects Progress Report 2014

The 2012-13 survey asked principals via open-response, “What was the

most valuable thing about having City Connects in your school this year?”

Example responses included:

“The way in which children are supported, either through direct

service or by advocating for and ensuring they get services.”

—Boston principal

“Connecting with outside providers to obtain services for our students.”

—Springfield principal

“City Connects has placed high-quality professionals in our school

who provide an outstanding level of effort and support to our

students, families, and teachers.”

—Boston principal

Teacher satisfaction and impact on teaching

In the 2012-13 anonymous survey of teachers at CCNX schools, 98% percent

of the 175 teachers responding in Boston and 91% of the 106 teachers

responding in Springfield answered “yes” to the question “Are you satisfied

with City Connects?”

Additionally, 97% of Boston and 91% of Springfield

respondents would recommend CCNX to a teacher in another school.

In addition to these basic satisfaction questions, the survey asked teachers

to report their satisfaction with the support SSCs provide to teachers, the

school as a whole, and students. Table 6 summarizes findings.

Table 6. Percentage of teachers satised with the supports the School Site Coordinator provides

in each area

Are you satised with the support City connects provides to:

Boston

(N=165)*

Springeld

(N=103)*

Teachers

( e.g., conducting Whole Class Reviews and assisting with behavior

challenges in the classroom)

92% 85%

The School

(e.g., coordinating the Student Support Team, outreach to families,

partnering with community agencies, their presence on the playground)

98% 92%

Students

(e.g., securing services, providing individual support, running lunch groups)

96% 90%

*Not all teachers responded to this question set. Source: City Connects 2013 teacher survey

28 All 384 teachers in the Boston City Connects schools were invited to participate in the survey;

of those who received the survey, 175 (46%) completed it. In Springeld, of the 207 teachers

who received the survey, 106 (51%) responded. Not every teacher responded to every

question.

Table 6 shows high levels of satisfaction across all types of support. The

slightly lower percentages in Springfield are due in part to the fact that

middle school teachers, new to CCNX in 2012-13, reported slightly lower

levels of satisfaction than elementary teachers, who were in their second

year of implementing the model. For example, although the table reports

85% satisfaction with support to teachers for Springfield overall, the

percentage rises to 88% when only elementary school teachers’ responses

are analyzed. This pattern is consistent with that seen in implementation

over time in Boston.

One of the most important components of School Site Coordinators’ work

with teachers is the Whole Class Review process. This process gives School

Site Coordinators the information they need to tailor services for students.

CCNX has learned that the Whole Class Review process enhances most

teachers’ non-academic knowledge of their students, which in turn informs

their work in the classroom. In 2012-13, the majority of teachers agreed that

the Whole Class Review process enhanced their awareness of their class as a

whole and students as individuals; see Table 7.

Table 7. Percentage of teachers who agree with each statement about the Whole Class Review

process, 2012-13

I agree that:

Boston

(N=136)*

Springeld

(N=76)*

a. The Whole Class Review process enhanced my awareness of the

dynamics of my class as a whole.

87% 91%

b. The Whole Class Review process enhanced my awareness of my

students as individuals.

89% 91%

c. The Whole Class Review supported my ability to identify new options

for working with my students.

86% 87%

d. The Whole Class Review was helpful to me. 84% 89%

e. My instructional practices were enhanced as a result of the Whole

Class Review

68% 84%

f. The Whole Class Review process added to my knowledge of the

non-academic aspects of my students’ lives (e.g., neighborhood and

family context).

88% 87%

g. The Whole Class Review process increased my empathy for students. 81% 86%

*Not all teachers responded to this question set. Source: City Connects 2013 teacher survey

The results in Table 7 suggest that teachers find the Whole Class Review

process led to new knowledge and awareness, both of individual students

City Connects Progress Report 2014

and of the dynamics of their class as a whole. Findings from this section

of the survey begin to shed light on the complexity of how the CCNX

intervention affects the work of schools, showing that the majority of

teachers not only report higher levels of awareness of the context of

students’ lives, but also an increase in empathy for students.

These quantitative findings are echoed and expanded upon in open-

response survey items. Teachers were asked, “What would you say to

a colleague is the most important benefit to City Connects?” A majority

of responses cited benefits to students, and many of the comments

revealed that the close collaboration with teachers that leads to greater

understanding of students as individuals was viewed as integral to these

benefits. Example comments include:

“Our City Connects coordinator is … not only resourceful and

proactive because she establishes and strengthens relationships

with community partners, but she simply knows and cares about

the students. Talking with her about individuals is always

helpful to understanding how students think, get motivated, and

ultimately learn.” —Boston teacher

“The connection they make with the students and their families and

the information they pass on to the teachers. Their willingness

to listen and consider our concerns. They provide a place where

students feel safe and valued.” —Springfield teacher

“It is wonderful to have someone who knows about behavior plans

available for support. I have learned about making them more

realistic for children, and more in line with promoting success.”

—Boston teacher

“The school has gained a watchful eye toward the emotional

and academic needs of all students, and also an experienced

[professional] that can function between the worlds of school

counselor and teacher.” —Springfield teacher

“City Connects helps us access services within the community. When

students have specific needs City Connects almost always has a

resource to help…” —Boston teacher

Teachers reported they recognize the services CCNX arranges are

appropriate to the individual student. The survey asked teachers whether

they agreed (yes/no) with the statement, “The support and enrichment

services most students receive are a good fit for their needs.” Agreement

was high: in Boston, 93%, and in Springfield, 90%. At the same time, the

teacher findings and comments above reveal that for teachers, the way

these specific matches are established between students and particular

supports matters. In the CCNX model, the connections are made teacher

by teacher and student by student. Through this process, teachers gain

insights into their students and their classes that deepen their empathy

and help inform their strategies for reaching individual students.

Impact on Community Agencies

Number of 2012-13 partnerships

During the 2012-13 academic year, CCNX collaborated with 380 community

partners in Boston and 179 in Springfield. Services to students and

to schools were provided by (a) community agencies; (b) community

institutions and businesses; and (c) universities. Table 8 displays the

number of each type of community partner by year for the past two years.

Table 8. Number of CCNX community partners, by year and by partner type

Community Partner Type

Boston Springeld

2011-12 2012-13 2011-12 2012-13

Community Agency 148 192 46 62

Community Institution/Business 83 102 56 91

University Partner 44 38 10 12

Other 16 31 6 14

Total across Partner Types 291 380 118 179

As Table 8 shows, the number of partnerships increased in both Boston

and Springfield from 2011-12 to 2012-13. The table reflects growth in

partnerships for both districts, and the increase in partnerships in

Springfield is partly due to the expansion of the model to middle schools,

where in some cases, students need different services than elementary

school students do.

City Connects Progress Report 2014

Community partner satisfaction

In the 2012-13, 132 community partners working with CCNX schools

responded to an online survey. Of the responding partners, 105 identified

themselves as working primarily in Boston and 27 in Springfield. In

both cities, most respondents reported (via yes/no response) that they

are satisfied with their partnership with CCNX: 99% in Boston and 92%

in Springfield. Most also indicated that CCNX is effective at identifying

students’ needs: 99% (Boston) and 88% (Springfield). Almost all would

recommend CCNX to another agency: 99% (Boston) and 96% (Springfield).

Partners were asked to rate their levels of satisfaction when working

with CCNX schools and non-CCNX schools; they rated their satisfaction

overall and along several dimensions of partnership.

For both Boston

and Springfield, an important difference across CCNX and non-CCNX

schools was the degree of satisfaction, with higher percentages of partners

indicating they are “very satisfied” with CCNX school partnerships than

non-CCNX school partnerships; see Table 9.

Table 9. Percentage of community partners who are very or somewhat satised with

dimensions of partnership with CCNX vs. non-CCNX schools

*Not all partners responded to every question. Source: City Connects 2013 community partner survey

Table 9 reveals several dimensions of good collaboration for which

respondents were more often “very satisfied” or “somewhat satisfied” with

29 These dimensions of partnership included communication, referrals, follow-up, meeting

goals, tailoring services, providing student support, and cultural sensitivity. Participants

were asked to use a four-point scale to denote level of satisfaction: very satised, somewhat

satised, somewhat dissatised, and very dissatised. Not applicable was also listed as an

answer choice.

Boston Springeld

CCNX

(N =88)*

Non- CCNX

(N = 61)*

CCNX

(N = 21)*

Non- CCNX

(N = 18)*

a. Communication with primary contact 99% 67% 90% 89%

b. Referral process (e.g., identifying students that would benet

from your services)

98% 67% 95% 81%

c. Follow-up on service delivery (e.g., checking to ensure the

student(s) received the service)

95% 62% 75% 65%

d. Eectiveness of your partnership in reaching goals 100% 65% 71% 76%

e. Providing you with feedback that would improve service

delivery, when appropriate

90% 60% 79% 56%

f. Providing opportunities for you to provide feedback to the

school

93% 61% 79% 53%

g. The cultural competence of your primary contact in the school 98% 86% 100% 67%

CCNX schools than with non-CCNX schools. Differences are particularly

notable in the referral process, the opportunities provided for partners to

give feedback to the school, and the cultural competence of the primary

contact in schools. While Table 9 reflects somewhat higher reported

rates of satisfaction in Boston than in Springfield, this difference across

districts is less notable in Table 10, which reports only the “very satisfied”

responses.

Table 10. Percentage of community partners who are very satised with dimensions of

partnership with CCNX vs. non-CCNX schools

*Not all partners responded to every question. Source: City Connects 2013 community partner survey

As Table 10 shows, in both Boston and Springfield, many more community

partners were “very satisfied” with their work in CCNX in contrast to non-

CCNX school partnerships.

The survey included open-ended questions that offered community

partners the option to comment on their partnership with CCNX.

Responses included the following:

“At the schools where I have been able to make contact with a

City Connects Site Coordinator, I have had a much easier time

identifying students who would be a good fit for our program

and ensuring that those students follow through with the process.

When I have made contact with City Connects, it has been a highly

positive experience.”

—Boston community partner

Boston Springeld

CCNX

(N =88)*

Non- CCNX

(N = 61)*

CCNX

(N = 21)*

Non- CCNX

(N = 18)*

a. Communication with primary contact 81% 15% 62% 22%

b. Referral process (e.g., identifying students that would benet

from your services)

60% 16% 68% 13%

c. Follow-up on service delivery (e.g., checking to ensure the

student(s) received the service)

69% 12% 60% 12%

d. Eectiveness of your partnership in reaching goals 65% 18% 57% 12%

e. Providing you with feedback that would improve service

delivery, when appropriate

58% 16% 47% 17%

f. Providing opportunities for you to provide feedback to the

school

55% 16% 47% 18%

g. The cultural competence of your primary contact in the school 78% 39% 75% 11%

City Connects Progress Report 2014

“The City Connects staff members at our school are incredibly

knowledgeable, helpful, and committed to the students as well as

to working with partners at the school. They’ve taken the time

to learn about our program and to ensure that they are making

appropriate referrals to our program. They consistently follow up

with us and are … collaborative and easy to work with.”

—Boston community partner

“City Connects workers have helped identify students whose skills

and needs make them good candidates for our…programs.”

—Springfield community partner

“I LOVE the City Connects Coordinators! They have gone above and

beyond in connecting students to my program. They have helped

students/families with the registration paperwork, sent it to me in

a timely fashion and do check-ins to make sure we are ‘all set.’ I

can’t say enough about them!”

—Springfield community partner

Conclusions

City Connects has shown that optimized student support can be delivered

in a high-impact, cost-effective way. By making use of existing structures

in the public schools, and by leveraging the rich resources of the city’s

community agencies, City Connects is able to link students to the services

and enrichments that match their individual strengths and needs.

The successful expansion of City Connects to a new city—Springfield,

MA—demonstrates that the model is scalable. Measures such as the

number of Whole Class Reviews completed, community partnerships

established, and services delivered indicate that City Connects is being

implemented in Springfield with high fidelity to the model. The initial

student outcomes analysis reported here suggests that City Connects

contributes to improved academic achievement after only one year of

implementation. Expansion to a new city with even higher levels of

poverty and lower tax revenue has provided encouraging evidence of the

robustness of the model, suggesting it is a feasible way to address out-of-

school factors impacting achievement in urban settings across the state

and country.

Students enrolled in City Connects elementary schools benefit long after

they have left the intervention itself and move into middle school and high

school. As shown in this report, students enrolled in City Connects schools

outperform their non-City Connects peers on measures of academic

achievement and life chances, such as middle school report card scores,

standardized test scores, attendance at selective high schools, chronic

absenteeism, and rates of school dropout. Careful attention to the unique

skills, talents, and needs of each student makes a difference.

City Connects Progress Report 2014

References

Berliner, D. C. (2009). Poverty and potential: Out-of-school factors in school

success. Boulder and Tempe: Education and the Public Interest Center &

Education Policy Research Unit. Available: epicpolicy.org/publication/

poverty-and-potential

Blow, C. M. (2011, October 28). American’s exploding pipe dream. The

New York Times. Retrieved November 22nd, 2011 from www.nytimes.

com/2011/10/29/opinion/blow-americas-exploding-pipe-dream.html?_

r=2&hp

Bryk, A.S., Sebring, P.B., Allensworth, E., Luppescu, S., & Easton, J. Q.

(2010). Organizing schools for improvement: Lessons from Chicago. Chicago:

Chicago University Press.

Coleman, J.S., Campbell, E.Q., Hobson, C.J., McPartland, J., Mood,

A.M., Weinfield, F.D., et al. (1966). Equality of educational opportunity.

Washington, DC: Office of Education, U. S. Department of Health,

Education and Wellness.

Coley, R.J., & Baker, B. (2013). Poverty and education: Finding the way

forward. The ETS Center for Research on Human Capital and Education.

Available: www.ets.org/s/research/pdf/poverty_and_education_report.pdf

Dearing, E. (2008). The psychological costs of growing up poor. Annals of

the New York Academy of Sciences (Special Issue: Scientific Approaches to

Understanding and Reducing Poverty, S. G. Kaler & O. M. Rennert, Eds.),

1136, 324-332.

Harrington, M. (1962). The other America: Poverty in the United States. New

York: Macmillan.

Imbens, G.W. (2000). The role of the propensity score in estimating dose-

response functions. Biometrika, 87(3), 706-710.

Levin, H., Belfield, C., Muennig, P., & Rouse, C. (2006). The costs and

benefits of an excellent education for all of America’s children. Teachers

College, Columbia University.

Phillips, M., Brooks-Gunn, J., Duncan, G., Klebanov, P. & Crane, J. (1998).

Family background, parenting practices, and the black–white test score

gap. In C. Jencks and M. Phillips (Eds.), The black-white test score gap.

Washington, D.C.: Brookings Institution Press.

Raudenbusch, S., & Bryk, A. (2002). Hierarchical linear models:

Applications and data analysis methods. Second ed. Thousand Oaks, CA:

Sage Publications.

Raudenbusch, S., & Liu, X. (2001). Effect of study duration, frequency of

observation, and sample size on power in studies of group differences in

polynomial change. Psychological methods, 6(4), 387-401.

Rothstein, R. (2010). How to fix our schools. Issue Brief #286. Washington,

DC: Economic Policy Institute, October 14, 2010. Available: www.epi.org.

Rothstein, R. (2013). Does “poverty” cause low achievement? The Economic

Policy Institute blog, October 3, 2013. Available: www.epi.org/blog/

poverty-achievement/

Walsh, M.E., & DePaul, G. (2008). The essential role of school-community

partnerships in school counseling. In H. L. K. Coleman & C. Yeh

(Eds.),Handbook of school counseling (pp. 765-783). Baltimore: MidAtlantic

Books & Journals.

Walsh, M. E., & Murphy, J. (2003). Children, health, and learning. San

Francisco: Jossey-Bass.

City Connects Progress Report 2014

City Connects Sta • (2012-13)

Mary E. Walsh, Ph.D.

Executive Director, City Connects

Kearns Professor, Department of Counseling, Developmental and

Education Psychology, Lynch School of Education, Boston College

Director of the Boston College Center for Optimized Student Support

Claire Foley, Ph.D.

Associate Director and Director of Research Reports, City Connects

Visiting Professor in Linguistics, Boston College

Beverly Ross Denny, M.B.A.

Director of New Initiatives, City Connects

Leah Lindsay

Director of External Relations, City Connects

Jennifer Coyle, M.A.

Administrative Officer, City Connects

Mary Howard

Manager, Office Administration, Center for Optimized Student Support

Brian Ward, M.A.

Technology Consultant, City Connects

City Connects Research Sta • (2012-13)

George Madaus, Ed.D.

Director of Evaluation, City Connects

Boisi Professor Emeritus, Department of Educational Research,

Measurement and Evaluation, Lynch School of Education,

Boston College

Founding Director, Boston College Center for Testing, Evaluation,

and Educational Policy

Anastasia Raczek, M.Ed.

Associate Director of Evaluation, City Connects

Terrence Lee-St. John, M.Ed.

Research Associate

Consultants • (2009-present)

Henry Braun, Ph.D.

Boisi Professor, Department of Educational Research,

Measurement and Evaluation, Lynch School of Education,

Boston College

Director, Boston College Center for Testing, Evaluation and

Educational Policy

Eric Dearing, Ph.D.

Associate Professor, Department of Counseling, Development, and

Educational Psychology, Lynch School of Education,

Boston College

Norman Hursh, Sci.D.

Associate Professor of Counseling Psychology, Boston University

City Connects Progress Report 2014

Maureen Kenny, Ph.D.

Professor, Department of Counseling, Development, and

Educational Psychology, Lynch School of Education,

Boston College

Julie Paquette MacEvoy, Ph.D.

Assistant Professor, Department of Counseling, Development, and

Educational Psychology, Lynch School of Education,

Boston College

Graduate Research Assistants • (2012-13)

Bercem Akbayin, M.Ed. Shea Kelly

Chen An, M.Ed. Rebecca Ledford

Sarah Backe, M.A. Francesca Longo

Teresa Battaglia Erin Sibley, Ed.M.

Jillian Boudreau Evan Michel, MA.

Michael Capawana, M.A. Amy Orecchia, M.A.

Kelly Collins Stephanie Paulk, M.A.

Jaime Denelle Katherine Shields, Ed.M.

Interns • (2012-13)

Katherine Fragapane Kathryn Raskin

School Data Liaisons

Kamalkant Chavda, Ph.D., Boston Public Schools

Paul Foster & Chekesha Lockett, Springfield Public Schools

External Evaluators • (2009-present)

Albert Beaton, Ed.D.

Former Augustus Long Professor Education, Lynch School of

Education, Boston College

Director of design, research, and data analysis for the National

Assessment of Educational Progress (NAEP)

Chairman of IEA’s Technical Advisory Committee (1989-1993)

Jennifer Hebert-Beirne, Ph.D.

Vice President of Research and Development, Women’s Health

Foundation, Chicago

Former policy planner and analyst, Chicago Department of Public

Health

Thomas Kellaghan, Ph.D.

Director, Educational Research Centre, St. Patrick’s College,

Dublin

Fellow of the International Academy of Education

President, International Association for Educational Assessment

(1997-2001)

Patrick McEwan, Ph.D.

Whitehead Associate Professor of Critical Thought, Department of

Economics, Wellesley College

Paul Holland, Ph.D.

Professor Emeritus at the Graduate School of Education and

Department of Statistics, University of California, Berkeley

Fellow of the American Educational Research Association

Elizabeth Reisner, M.Ed.

Founder and Manager, Policy Studies Associates, Inc.,

Washington D.C.

City Connects Progress Report 2014

Implementation Team • (2012-13)

Patrice DiNatale, M.Ed.

Director of New Practice

Alison Stahl, M.S.W.

Director of Implementation

Audra Vernon, M.Ed.

Implementation Systems Coordinator

Program Managers • (2012-13)

Raisa Carrasco-Velez, M.P.A.

Group A Program Manager, Boston

Raghida Jeranian, M. Ed.

Group B Program Manager, Boston

Julie Donovan, M.S.W.

Group A Program Manager, Springfield

Boston School Site Coordinators • (2012-13)

Brendan Adams, M.Ed., Jackson Mann K-8 School

Pam Belford, M.S., Dever-McCormack K-8 School

April Bouzan, M.Ed., Edison K-8 School

Courtney Bruno, M.S., Mission Hill K-8 School

Elizabeth Centeio, M.Ed., Dudley Street Neighborhood School

Amy Cluff, M.S.W., Jackson Mann K-8 School

Emilie Cromer, M.S., Dever-McCormack K-8 School

Claire DeRosa, M.S., Elihu Greenwood Leadership Academy

Mark Griffin, M.A., Edison K-8 School

Lilly Guttenplan, M.S.W., John F. Kennedy Elementary School

Joseph King, M.Ed., Holland Elementary School

Ursula Lucas, M.S., Trotter Elementary School

Lauren Melone, M.A., Clap Innovation School

Nikki Robinson, M.A., Winship Elementary School

Milord Mirville, M.A., Mason Elementary School

Sarah O’Connor, M.S.W., Gardner Pilot Academy

Sara Pizzute, M.S.W., Eliot K-8 School

Adam Prisby, M.Ed., Holland Elementary School

Catherine Riede, M.Ed., Quincy Elementary School

Derronda Williams, M.A., Bates Elementary School

Nicole Young, M.S.W., Quincy Elementary School

Springeld School Site Coordinators • (2012-13)

Antonia Cardaropoli, M.A., Chestnut Accelerated Middle School

Caitlin Casey, M.S.W., Gerena Community School

Meagan Graham, M.S.W., Kiley Middle School

Jennifer Grant, M.S.W., Brightwood Elementary School

Stephanie Sanabria, M.A., Kennedy Middle School

Karen Schreiner, M.A., Zanetti Montessori Magnet School

Doris Schuh, M.Ed., Gerena Community School

Katie Singerman, M.Ed., White Street Elementary School

Lauren Zanotti, M.Ed., Brookings Elementary School

City Connects Progress Report 2014

Coordinators Implementing the New Balance

Foundation Health & Wellness Program • (2012-13)

Portia Abernathy, M.A., M.Ed., Mason Elementary School

Caitlin Kelly, M.Ed., Edison K-8 School

Consultant • (2012-13)

Michele Montavon, Ph.D.

Director of Health Education, Worcester Public Schools

Information Technology Support

Ronald Ko, M.S.

Barry Schaudt, Ph.D.

Artwork, Graphics, and Layout

Kevin Keane, Genius Pool

ó ó ó