Project Description: Schoolwide Projects examined the implementation of Chapter 1 schoolwide projects in a large urban school district in order to describe the extent of use, the designs employed, and the effects of the schoolwide projects in 61 schools. Longitudinal achievement effects have been reported for 40 elementary schools.
Many observers of Chapter 1 programs have noted the disruptive impact of pullouts, the waste of materials and time expended to keep non-eligible children from benefiting from Chapter 1 services, and the limitations on use of effective programs imposed by the requirement that only test-eligible children may be served. Despite this strong rationale for schoolwide projects as an alternative use of Chapter 1 funding, it is by no means certain that schoolwide projects, such as reducing class sizes, would be any more effective than traditional Chapter 1 services (e.g., pullout programs for test-eligible children). This project studies the implementation and effectiveness of Chapter 1 schoolwide projects in a large city school district. It addresses two questions: (1) What are the schools that are exercising this option actually doing? and (2) What are the effects of these schoolwide projects on student achievement? Question 1 is addressed in Activity 20005 and question 2 is addressed in Activity 20006.
Project Director: Winfield, Linda W.
Institution: Johns Hopkins University
PHONE (410) 516-8800
FAX (410) 516-8890
Activities
Activity ID: 7112-20005
Name of Activity: Descriptive Study of Chapter 1 Schoolwide Projects
Description of Activity: This activity examined the implementation of Chapter 1 schoolwide projects in a large urban school district in order to describe the extent of use, the designs employed, and the effects of the schoolwide projects in 61 schools. Longitudinal achievement effects have been reported for 40 elementary schools. Descriptions of Chapter 1 schoolwide plans and practices are obtained for 61 elementary schools in Philadelphia. Survey instruments are designed and administered and observations are conducted to learn what each of the schools is doing.
The project was designed as a 3-year study and compared the effectiveness of alternative programs implemented under schoolwide funding with schools using more traditional Chapter 1 models. The project has been completed.
Study Design: The study design is not specified.
Unit of Analysis: The unit of analysis is the school.
Generalizability: The study examines the universe of elementary schools in Philadelphia with Chapter 1 schoolwide projects. Results are generalizable to these schools only.
Sample Description: The study is based on a purposive sample of 61 elementary schools in urban Philadelphia with Chapter 1 schoolwide projects. These schools are among the poorest schools in the city.
Dependent Variables: Dependent variables are not specified.
Independent Variables: Independent variables are not specified.
External Variables Controlled: Control procedures/variables are not specified.
Statement of Finding(s): Schoolwide Chapter 1 programs implemented in six schools increased students' reading and math achievement.
Description of Finding(s): Schoolwide Chapter 1 programs implemented in six schools increased students' reading and math achievement. In one of the schools, students' reading scores rose six percent and their math scores rose ten percent after the schoolwide program's first year.
The researchers found that these six schools used their Chapter 1 funds in various ways. They lowered their teacher-student ratio in math and reading by adding another teaching position, eliminated split-grade classes, and cut out pull-out classes by providing all instruction within the regular classroom. Five of the six schools created a first-grade transition class for students who had not attended kindergarten or who had been retained.
Are data from the study available? Yes
Name of Activity: Evaluations of Alternative Chapter 1 Schoolwide Projects
Description of Activity: This project examined the implementation of Chapter 1 schoolwide projects in a large urban school district in order to describe the extent of use, the designs employed, and the effects of the schoolwide projects in 61 schools. Longitudinal achievement effects have been reported for 40 elementary schools. In this activity, evaluations are conducted to determine the effects of the Chapter 1 schoolwide programs in 61 urban Philadelphia elementary schools.
The project has been completed.
Study Design: The evaluations that comprise this study compare schools using schoolwide projects with schools that have traditional Chapter 1 programs, matched on historical achievement level, economic background of parents, and ethnic makeup. The comparisons provide a basis for the assessment of student achievement in the various types of schools. A matrix sampling procedure in which each child takes only a portion of an extensive test battery is employed, yielding meaningful groupl-evel data for a wide range of learning domains. A subsample of students is given an individually-administered reading inventory. In addition to the assessments, the schoolwide and non-schoolwide schools are visited and described, and changes introduced by the transition from traditional Chapter 1 to the schoolwide program are documented.
Unit of Analysis: The unit of analysis, presumably either the individual student or the school, is not specified.
Generalizability: This study examines the universe of elementary schools in Philadelphia with Chapter 1 schoolwide projects; therefore, findings are generalizable to those schools. In addition, a matched sample of 61 Philadelphia schools without schoolwide projects is selected for comparison. Because sample selection techniques are not fully described, the generalizability of findings pertaining to these schools cannot be determined.
Sample Description: This study is based on the universe of 61 schools with schoolwide projects and a purposive sample of 61 matched schools without schoolwide projects in urban Philadelphia. Schools are matched on the basis of historical achievement level, economic backgrounds of parents, and ethnic makeup of students. The matrix sampling procedure is not specified. The procedure for selecting a subsample of students to take the individually-administered reading inventory is also not specified.
Dependent Variables: The dependent variable is student achievement, including measures of higher order and basic skills obtained from the matrix sampling procedure and reading skills obtained from administration of the reading inventory to individual students.
Independent Variables: The independent variable is Chapter 1 program type, schoolwide or traditional.
External Variables Controlled: The matching procedure for schools controls for historical achievement, ethnicity of students, and family economic background.
Statement of Finding(s): Achievement results fluctuate by grade level, and age, gender, and race/ethnicity influence achievement in various ways.
Description of Finding(s): This activity conducted longitudinal evaluations of 40 elementary schoolwide programs compared to 20 elementary schools that were comparable in level of poverty but were not schoolwide program sites. The researchers report that achievement results fluctuate by grade level, and age, gender, and race/ethnicity influence achievement in various ways. The following brief findings by grade level are all statistically significant unless otherwise noted.
First-graders in schoolwide program sites performed the same as first-graders in non-schoolwide sites. Older first-graders performed less well than younger first-graders. African-American children performed slightly higher than white students.
Second-graders in schoolwide program sites performed better than second-graders in non-schoolwide. Again, older second-graders scored less well than younger. Girls did better than boys. African American and Hispanic students performed less well than white students.
Third-graders in schoolwide program sites performed lower than third-graders in non-schoolwide. Again, older students, boys, African American, and Hispanic students scored lower than their respective counterparts.
Fourth- and fifth-graders in schoolwide program sites scored higher than fourth- and fifth-graders in non-schoolwide sites, but the results were not statistically significant. The prevailing patterns for age, gender, and race/ethnicity continued, but the race/ethnicity fifth-grade results were not significant.
The researchers note that the findings provide "some tentative evidence of the long-term effects of being in a schoolwide project." It is encouraging that both the fourth- and fifth-grade achievement effects were positive even though not significant, indicating long-term effectiveness. Also encouraging is that race effects, although still negative in fifth grade, are no longer significant. But the negative effects for third-grade schoolwide program students are puzzling, and the strong age, gender, and race-ethnicity effects that run throughout the grades are discouraging.
The researchers also examined how the implementation of specific components of schoolwide programs affected student achievement in grades one through five in a single year (1989). They examined survey data, received from principals at 40 sites, which described the major components that each site had implemented, and constructed scales to reflect the components.
No clear-cut effects on student achievement emerge from the various components. As with longitudinal achievement comparisons, different patterns of effects appear at each grade level (one through five).
The effects of the parent involvement component are typical of the overall findings in their complexity. At second and fourth grade, parent involvement is positively related to student achievement, but no effect is seen for third grade. In fifth grade, the percentage of parents involved in school programs is negatively related to student achievement outcomes; however, the percentage of parents attending school conferences is positively related.
One aspect of site-based management -- teachers' involvement in decision making about how human resources are used -- comes close to having reasonably clear effects. This component includes teacher input in decisions about how they (and students) are assigned to classes, how and what new teachers or aides are hired, and so on. This component had a positive impact on student achievement at all grade levels except the fourth.
The researchers emphasize that their findings indicate the complexity and interaction of organizational variables in changing how schools deliver services to students in Chapter 1 programs. The findings indicate some modest evidence of potential long-term achievement effects of schoolwide projects for serving disadvantaged students.
Are data from the study available? Yes
![[ Previous Project ]](/icons/prev.gif){kind=icon}
![[ Center Description ]](/icons/up.gif){kind=icon}
![[ Next Project ]](/icons/next.gif){kind=icon}