You are seeing this message because your web browser does not support basic web standards. Find out more about why this message is appearing and what you can do to make your experience on this site better.

www.HFRP.org

The Harvard Family Research Project separated from the Harvard Graduate School of Education to become the Global Family Research Project as of January 1, 2017. It is no longer affiliated with Harvard University.

Terms of Use ▼

Program Description

Overview Teach Baltimore (TB) is an academically intensive summer program that recruits and trains university students to provide approximately eight weeks of reading and writing instruction to low-income elementary students. Founded by two Johns Hopkins University students, TB is sponsored by the Johns Hopkins University and is entirely run by college students and recent graduates. TB has three main goals: (1) to prevent summer learning loss and improve the reading skills of high-poverty students, (2) to transform collegiate volunteerism into a focused and effective commitment, and (3) to create a successful prototype that can be easily and cost effectively replicated.
Start Date 1992
Scope local
Type summer/vacation
Location urban
Setting public school
Participants kindergarten and elementary school students
Number of Sites/Grantees 5 sites
Number Served approximately 2,100 in 1992–2002, 293 began the 3-year intervention in 1999, 145 began in 2000
Components TB is run by Johns Hopkins University college students and recent graduates, with training and support provided by experienced teachers in the Baltimore City Public School System. The TB summer begins with 2 weeks of preservice training in which volunteer instructors become familiar with the curriculum as well as strategies for quality lesson planning, effective behavior management, and successful parent involvement. Following this training period, the 8-week program begins. Each day starts with youth receiving breakfast, followed by 3 hours of intensive reading and writing instruction. Morning instruction is followed by lunch, then a series of afternoon activities focused on hands-on science and math, recreation, art, and drama. In addition, youth take part in weekly afternoon field trips and cultural enrichment experiences. The program provides a maximum youth–instructor ratio of 7 to 1.
Funding Level approximately $293,000 in 1999; $231,000 in 2000; $253,000 in 2001
Funding Sources Aaron and Lillie Straus Foundation, the Abell Foundation, the After-School Institute, Annie E. Casey Foundation, Baltimore City Public School System, Baltimore Community Foundation, Baltimore Orioles, Baltimore Sun, Best Buy Youth's Foundation, Black & Decker Corporation, the Campbell Foundation, Inc., Civic Works AmeriCorps, Clayton Baker Trust, the Commonweal Foundation, Corporation for National Service, Developmental Studies Center, Ensign Markland Kelly Foundation, Enterprise Foundation, ERIC Clearinghouse on Urban Education, Erickson Foundation, Goldsmith Foundation, James Riepe Family Foundation, Jean and Sidney Silber Foundation, John J. Leidy Foundation, Johns Hopkins University, Joseph Meyerhoff Fund, Junior League of Baltimore City, Kaiser Permanente, Lockhart Vaughan Foundation, Lois H. & Charles A. Miller Foundation, Inc., Margaret O. Cromwell Family Fund of the Baltimore Community Foundation, Maryland Institute College of Art, Maryland Space Grant Consortium, Maryland State Department of Education, Mercantile Bank, the Morris Goldseker Foundation, Morton and Jane Blaustein Foundation, Morton & Sophia Macht Foundation, Open Society Institute – Baltimore, Otto-Whalley Family Foundation, Reading Is Fundamental, Shriver Consortium, Smith Richardson Foundation, St. Paul Companies, Inc., Staples Foundation for Learning, State Farm Companies Foundation, T. Rowe Price Foundation, Thomas Wilson Sanitarium, U.S. Department of Education, United Way of Central Maryland, UPS Foundation, Verizon Foundation, William G. Baker, Jr. Memorial Fund


Evaluation

Overview A 3-year random assignment study of TB looked at the long-term impact of multiple summer interventions on the reading achievement of low-income youth.
Evaluators Geoffrey Borman, and N. Maritza Dowling, University of Wisconsin–Madison

Laura Rachuba, Gina Hewes, and Laura T. Overman, Johns Hopkins University

Matthew Boulay, Jody Kaplan, and Ron Fairchild, Teach Baltimore
Evaluations Profiled Can a Summer Intervention Program Using Trained Volunteer Teachers Narrow the Achievement Gap? First-Year Results From a Multi-Year Study

The Longitudinal Achievement Effects of Multi-Year Summer School: Evidence From the Teach Baltimore Randomized Field Trial
Evaluations Planned None
Report Availability Borman, G., Rachuba, L., Hewes, G., Boulay, M., & Kaplan, J. (2001). Can a summer intervention program using trained volunteer teachers narrow the achievement gap? First-year results from a multi-year study. ERS Spectrum, 19(2), 19–29.

Borman, G. D., & Dowling, N. M. (2006). The longitudinal achievement effects of multi-year summer school: Evidence from the Teach Baltimore randomized field trial. Educational Evaluation and Policy Analysis, 28, 25–48.

Available at: www.summerlearning.org/resources/index.html


Contacts

Evaluation Geoffrey D. Borman, Ph.D.
Associate Professor, Educational Leadership and Policy Analysis,
Educational Policy Studies, and Educational Psychology
University of Wisconsin-Madison
1161D Educational Sciences Building
1025 West Johnson Street
Madison, WI 53706-1796
Tel: 608-263-3688
Fax: 608-265-3135
Email: gborman@education.wisc.edu
Program Jody Kaplan
Program Coordinator
Teach Baltimore
The Bunting-Meyerhoff Interfaith and Community Service Center
3400 North Charles Street
Baltimore, MD 21218
Tel: 410-261-1882
Fax: 410-261-1212
Email: teach@jhunix.hcf.jhu.edu
Profile Updated July 13, 2007

Evaluation 1: Teach Baltimore: Can a Summer Intervention Program Using Trained Volunteer Teachers Narrow the Achievement Gap? First-Year Results From a Multi-Year Study


Evaluation Description

Evaluation Purpose To examine TB’s effectiveness in preventing summer learning loss for low-income public school students.
Evaluation Design Experimental: Program applications were distributed to all kindergarten and first-grade students in 10 high-poverty urban Baltimore schools during the spring of 1999. Applicants were assigned randomly to either a program or control group. In total, 293 youth were assigned to the program group, and 135 were assigned to the control group.

Two sets of analyses were conducted to look at preliminary (1st-year) program effects. First, all youth assigned to the program group were compared to all youth in the control group. Second, program effects were then estimated for all youth in the program group with attendance rates of 75% or higher.

“Treatment compliance,” or the degree to which youth assigned to the treatment group actually attended the program, varied. Among kindergarteners, 20% never attended, 34% attended between 1% and 75% of the time, and 46% attended more than 75% of the time. Among first graders, 23% never attended, 31% had attendance rates between 1% and 75%, and 46% attended over 75% of the time. The evaluators used a statistical method called latent variable mixture modeling to determine a subset of the control group that can be considered “would-be compliers” if they had been offered the program treatment.

Program and control youth were statistically equivalent in terms of gender, race/ethnicity, eligibility for the free lunch program, school attendance rate, and parents’ years of education. The only significant differences between the two groups for the background data were that first-grade program youth had a higher completion rate on the district-administered pretest (p <. 01), and for the kindergarten cohort, the control group had significantly higher (p < .05) pretest scores than program youth. Applicants and nonapplicants were also statistically equivalent in terms of gender, race/ethnicity, eligibility for the free lunch program, and parents’ years of education. The only significant differences between the two groups for background data were that applicants in both cohorts had higher 1998–1999 school year attendance rates (p < .001), and for test scores, the kindergarten cohort of applicants had significantly higher (p <. 05) pretest scores than nonapplicants.
Data Collection Methods Document Review: Education levels of program and control group youth’s parents were obtained from TB’s program application forms.

Secondary Sources/Data Review: Demographic data on treatment and control youth’s gender, race/ethnicity, free lunch eligibility status, and 1998–1999 school year school attendance rates were abstracted from the Baltimore City Schools’ Pupil Information File.

Tests/Assessments: For kindergarten students, the evaluators administered Level 10, Form B of the Comprehensive Test of Basic Skills, 4th Edition (CTBS/4) reading test, regardless of application status. For first graders, evaluators accepted the results of Level 11, Form B of the CTBS/4 reading test, which was administered to all Baltimore City first graders during the same week. The evaluation staff re-administered the Level 10 and 11 CTBS/4 during the fall to the kindergarten cohort (then in first grade) and the first grade cohort (then in second grade). In the re-administration evaluators used the CTBS/4’s Form A so as to minimize spurious gains associated with closely spaced pretests and posttests.
Data Collection Timeframe Data were collected during the spring and fall of 1999.


Findings:
Summative/Outcome Findings

Academic In the overall sample, program youth made greater gains in reading achievement than control youth in both cohorts, but neither result achieved statistical significance.

For the treatment compliance subsample, all program subgroups (kindergarteners above the mean on the pretest, kindergarteners below the mean on the pretest, first graders above the mean on the pretest, and first graders below the mean on the pretest) showed positive effects, with the exception of low-achieving kindergarteners, although only the result for high-achieving kindergarten students reached statistical significance (p < .05).

Evaluation 2: The Longitudinal Achievement Effects of Multi-Year Summer School: Evidence From the Teach Baltimore Randomized Field Trial


Evaluation Description

Evaluation Purpose To examine the 3-year longitudinal effects of a multiyear summer school intervention designed to counteract the cumulative effect of the summer slide on low-socio-economic-status youth’s reading outcomes.
Evaluation Design Experimental: The total sample includes 686 youth from 10 high-poverty, urban schools. This sample is composed of 2 year-specific cohorts: a group of 428 kindergarten and first-grade students who joined the study during the spring of 1999 and 258 kindergarten youth who joined the study in spring 2000. For the first cohort of 293 TB youth, the intervention began during the summer of 1999, after they had completed kindergarten or first grade. The opportunity to register for the program was offered to all kindergarten and first-grade students from the 10 Baltimore City schools during the spring of 1999. From the 1999 cohort of youth who registered for the program, the evaluation team randomly assigned 293 youth to receive the three-summer intervention and 135 to the control condition. For the second cohort of 145 TB participants, the intervention started during the summer of 2000, after they had completed kindergarten. From the summer 2000 registrants, evaluators randomly assigned 145 to receive three summers of intervention and 113 to the no-treatment control group.

Two sets of analyses were conducted to look at program effects. First, all youth assigned to the program group were compared to all youth in the control group with respect to reading learning trajectories. Second, program effects were then estimated for all youth in the program group with attendance rates at or above average over two or more summers.

“Treatment compliance,” or the degree to which youth assigned to the treatment group actually attended the program, varied. Approximately two thirds of the youth attended two or three summers of the program and nearly one third attended all three summers of the program. The evaluators used a statistical method called latent variable mixture modeling to determine a subset of the control group that can be considered “would-be compliers” if they had been offered the program treatment.
Data Collection Methods Document Review: Education levels of program and control youth’s parents were obtained from TB’s program application forms.

Secondary Sources/Data Review: Demographic data on treatment and control group youth’s gender, race/ethnicity, free-lunch eligibility status, and 1998–1999 school year attendance rates were abstracted from the Baltimore City Schools’ Pupil Information File.

Tests/Assessments: During the spring, the evaluation team administered Form B of the Comprehensive Test of Basic Skills, 4th Edition (CTBS/4) reading test to all treatment and control youth. The evaluation team began by administering the grade-appropriate test to kindergarten students and first graders. During the fall, staff re-administered the same test level to each cohort but used the alternate form, Form A, so as to minimize potential spurious gains due to relatively closely spaced pre- and posttests. Using the grade-appropriate form for each cohort, the same procedure was used during each ensuing year of the study. Over the course of the full 3 years of the study, treatment and control youth took the norm-referenced reading achievement tests six times, at the beginning and end of each of the three summers. Evaluators analyzed outcomes for the Total Reading test and the two subtests that make up the Total Reading score: Reading Vocabulary and Reading Comprehension.
Data Collection Timeframe Data were collected in 1999, 2000, and 2001.


Findings:
Summative/Outcome Findings

Academic Overall, TB youth did not demonstrate significantly greater learning rates than control youth on any of the three outcome measures (Total Reading, Reading Vocabulary, and Reading Comprehension).

No treatment effects on learning rates were found between TB and control youth of parents with varying levels of education on any of the three outcome measures.

When comparing treatment compliers to control would-be compliers, the effect of attending at least two of three summers at an average or better attendance rate was estimated as equivalent to 50% of one grade level in vocabulary, 40% of one grade level in comprehension, and 41% of one grade level in total reading relative to controls who were similar in profile to those TB youth who were regular program attendees (p < .05 for all three outcomes).

 

© 2016 Presidents and Fellows of Harvard College
Published by Harvard Family Research Project