Abstract

This study has been done to help me answer my own questions, and to provide a methodology that will be useful in evaluating other volunteer programs. While my research focuses on the impact of AmeriCorps on reading achievement, the logic of the model could be extrapolated to evaluate many programs. My objectives in this study are to determine a dollar value of the services rendered by AmeriCorps volunteers, and to then figure out how effective the volunteer is at helping to increase reading achievement levels in several major Ohio urban school districts. To assign a dollar value to the volunteers' labors for comparison purposes, they are equated with teacher's aides, and evaluated accordingly. I perform a cost-benefit analysis of three urban AmeriCorps programs. To determine the impact of the volunteer, deemed "serviceeffectiveness" I run a regression model. With reading achievement at the fourth grade level (based on results of the Fourth Grade Proficiency Test) as the dependent variable, the model can be seen as a production function. The independent variables include AmeriCorps' presence, student-to-teacher ratio, expenditure per student, location by city (either Columbus or Cleveland), percentage of married couple families in the tract of a given school, and the percentage of high school graduates over age 25 in a tract. The key findings of my study are somewhat limited by the explanatory power of my data. In a perfect world, I would have had access to information about achievement on the individual level. However, confidentiality rules limit that access, and I instead use information on a more aggregate level, for the school and the census tract in which it may be located. I found that the City Year Columbus volunteer's cost per hour was actually less than that of the paid worker, while the AmeriCorps for Literacy and Math volunteer's cost per hour was marginally more expensive. These results are based on data acquired directly from AmeriCorps program administrators. I also found that AmeriCorps' presence in schools had an unexpected association: where they were present, there were lower levels of reading achievement. The overall modeling sample consists of 61 schools, all located in census tracts with as much or more poverty occurrence than the average of those schools with AmeriCorps presence. While these findings are important, they must also be considered with respect to the explanatory power of the data .

Advisor

Burnell, James

Second Advisor

Fitz Gibbon, Heather

Department

Urban Studies

Publication Date

1998

Degree Granted

Bachelor of Arts

Document Type

Senior Independent Study Thesis

Available for download on Thursday, January 01, 2150

Request Access

Share

COinS
 

© Copyright 1998 William G. Kopka