Offenberg+and+Clark+1998

02/19/02 SCH DIST PHILA RES&EVAL **LIBRARY POWER** Impact on Student Achievement As Measured By The Stanford Achievement Test, 9th Edition

Report Prepared By

Robert M. Offenberg Research Associate and Thomas J. Clark Director

Research and Evaluation [School District of Philadelphia, 1998]


 * Library Power in Philadelphia**

The Philadelphia Library Program began with a cohort of 11 schools midway through the 1994-95 academic year. A second cohort of 19 schools was added the following year. Library Power funds have been used to provide professional development for teachers, school librarians and administrators, to renovate school libraries, and to match School District funds for books and other library materials. The School District of Philadelphia provided a full-time, certified library media specialist in each Library Power school: kept Library Power school libraries open throughout the day with schedules that allowed students to go to the library individually, in small groups, and with their classes; and covered labor costs for renovating school libraries.

For a one-time refurbishing cost of $9.000, Library Power schools created a physical resource center that enriched the entire school community--typically 700 or more students, many parents, and 30 or more staff. An annual expenditure of $6,000 updated the library collections to include current nonfiction and research materials that directly support the curriculum and fiction materials that reflect the cultural diversity of the school community and meet the reading needs of the students. An expenditure of approximately $58,000 per year provided each school with a certified school librarian who spent time working collaboratively with teachers to improve student achievement.

To achieve the Library Power goals, the initiative has a specific set of strategies: • to redesign and refurbish the library; • to update book and software collections; • to implement flexible scheduling and independent checkout; • to establish collaborative planning between librarians and teachers; • to provide extensive professional development opportunities for librarians, teachers, and administrators; and • to encourage the creation of partnerships among schools, public libraries, community agencies, academic institutions, and parent groups.

The Philadelphia Education Fund recently published a study entitled "Library Power in Philadelphia: Final Report from Seven Case Studies" (Useem & Coe, 1997).* The report included findings from continuing case studies at seven schools, project wide focus group interviews, and analysis of documents such as student work, curriculum maps, collaboration logs, and collection maps. The study showed that in schools where Library Power approaches full implementation, one Is likely to see a bright, well-furnished student-centered library used by individual students, groups, and whole classes engaged in research or other constructivist learning activities; increased circulation of books boosted by the independent checkout system; a revitalized librarian who assumes an array of teaching roles in ways that provide greater support for teachers; a collection that matches current classroom teaching topics; and closer relationships with community groups. In short, the library provides a vibrant center for collaborative teaching and learning [and] is headed by a librarian with a broader and more professionalized role. In addition, a new district-wide network of librarians provides intellectual and emotional support to its members and supplies for school district with a cadre of leaders {or related district-wide professional development activities (Useem & Coe, 1997). 1


 * Library Power in Philadelphia:Final Report from Seven Case Studies by Useem and Coe is available to download full-text from ERIC documents at



viewed 2/2/2010

**Impact on Student Achievement**
 * as Measured by the Stanford Achievement Test, 9th Edition**

The possible effects of Library Power on student achievement as measured by the SAT-9 were examined using regression analysis techniques. Spring 1997 Reading scores for both the multiple choice and open ended sections were used as the criteria in separate analyses for each of the grades for which data were available (grades 3, 4, 7, and 8). Scores from the Spring 1994 administration of the CTBS-4 and the percentage of students on AFDC and/or Food Stamps were used as covariates, essentially removing the effects of these variables on the 1997 achievement scores (both of these variables were highly correlated with the SAT-9 scores). Caution should be exercised in interpreting the results since most schools were in only their third year of program implementation; by its very nature this type of program will require considerable time for its effects to diffuse throughout the school. Other variables not being controlled in this analysis should also be examined for their effects (e.g. other programs, administrative support, teacher cooperation, etc.).

We found trends favoring Library Power schools over comparable schools on the multiple choice portion of the SAT-9 Reading when we examined scatter plots of the grade by grade analyses; of the 66 comparisons across the four grades (each of the 33 Library Power schools had two grades with SAT-9 data), 44 showed a higher than expected score for the Library Power schools (significant at the .05 level using the Chi Square statistic). In the eight individual grade analyses (multiple choice and open ended for each of the four grades) we found no statistically significant difference on the SAT-9 Reading between Library Power schools and comparable schools. However, we did notice that as the grade level increased, so did the number of Library Power schools scoring higher than expected, although none of the grade by grade analyses revealed statistically significant differences.

Another interesting finding involved the third grade multiple choice analysis, in which there was a significant interaction between poverty level and difference from the expected achievement level for the Library Power schools; student[s] in higher poverty Library Power schools scored higher than expected while students in low poverty schools scored lower than expected. This finding suggests that when mature, Library Power may provide the most benefit for children in high poverty schools. 2

Performance of Library Power Schools on 1997 SAT-9 Multiple Choice Reading Test Compared to Predictions from City-wide Trend

Library Power Schools Compared with All Others: 3rd Grade Spring 1997 SAT-9 Reading Multiple Choice

Library Power Schools Compared with All Others: 4th Grade Spring 1997 SAT-9 Reading Multiple Choice

Library Power Schools Compared with All Others: 7th Grade Spring 1997 SAT-9 Reading Multiple Choice

Library Power Schools Compared with All Others: 8th Grade Spring 1997 SAT-9 Reading Multiple Choice

Performance of Library Power Schools on 1997 SAT-9 Open End Reading Test Compared to Predictions from City-wide Trend

Library Power Schools Compared with All Others: 3rd Grade Spring 1997 SAT-9 Reading Open Ended

Library Power Schools Compared with All Others: 4th Grade Spring 1997 SAT-9 Reading Open Ended

Library Power Schools Compared with All Others: 7th Grade Spring 1997 SAT-9 Reading Open Ended

Library Power Schools Compared with All Others: 8th Grade Spring 1997 SAT-9 Reading Open Ended