Technology Annotated Report Excerpts

Excerpt 1 [Delaware]

Quantitative Analysis:
Describes assessments and variables used in analysis

STUDENT ACHIEVEMENT

Reading and mathematics achievement tests were administered to students in the first and second grades who participated in the target school program. A pretest was given to students in early fall 1997 and a posttest was administered in late spring 1998. Over 300 first-graders at four schools across twenty classrooms and approximately 250 second-graders at three schools across fourteen classrooms participated in the achievement testing. Students with missing or incomplete data were excluded from the analyses. The following sections provide scaled scores, percentile ranks, and stanines. Scaled scores are raw scores that have been converted to make scores in a given content area comparable from form to form and level to level. Percentile ranks range from a low of 1 to a high of 99 and indicate the percentage of the reference group obtaining scores equal to or less than that score. The reference group is a national sample of students at the same grade taking the test at a comparable time of the year. A percentile rank of 50 denotes average performance. Stanines are derived from percentile ranks and also indicate a student's relative standing in a reference group. Stanines are normalized, standard scores that range from a low of 1 to a high of 9. A stanine of 5 denotes average performance. An assumption made in these analyses is that while a student's scaled scores should significantly increase in any given school year, a student's standing in relation to the reference group should not significantly change between the fall and the spring.

First Grade Achievement. Table 8 (contact Delaware Challenge Project) provides the mean scaled score, percentile rank, and stanine by semester for first grade reading and mathematics achievement. As would be expected in any given academic year, student scaled scores on both the reading and mathematics achievement tests increased significantly from the pretest to the posttest. Because no local comparison group was available, the analysis of the gains in student achievement over the school year is based primarily on norm-referenced scores. Unlike scaled scores, one would not necessarily expect significant gains in norm-referenced scores over the course of the school year.

Excerpt 2 [Delaware]

Quantitative Analysis:
Describes coding procedures for cross-case analysis

INTRODUCTION

Classrooms in two schools participated in an in-depth study of the program's implementation. The study addressed the following four program components:

• Classroom Component - Teacher Usage
• Classroom Component - Home Connection
• Home Component - Student Usage
• Home Component - Parent Role

The initial data collection activities generated case-specific data to each of these orienting components. Subsequent coding of interview and observation data yielded a within-site coding matrix. Cross-case comparisons were made to examine the coding from each individual site by orienting variable, that is the role of classroom component, home component, etc. A cross-site coding procedure was then used to generate the findings as delineated in this report. Assertions were built based upon thematic findings common across both cases. Each of the cross-case data analyses was designed to enhance the generalizability of the findings.

For the purpose of this analysis, the two cases will be called Pacific Elementary School and Atlantic Elementary School. These names in no way relate geographically or otherwise to the two schools studied. Further, the results outlined in this section not only relate to classroom observation and interviews at these two case study schools, but when appropriate are supplemented with parent and staff interview data across all participating Challenge Grant schools.