At each of the 25 longitudinal-study schools, Special Strategies research teams spent a minimum of three days on site for each of five semesters from fall of 1990 through fall of 1992. This resulted in a total of at least 30 person days of on-site data gathering at each longitudinal site. During those visits, researchers interviewed administrators, teachers, parents, and students. During the site visit, detailed observational data on classroom processes were collected. At each school, three students were identified during year one for extended observation. Those students were followed through detailed "whole school day" observations at each subsequent school visit. Based on their fall 1990 pretests, augmented by principal judgments, students generally were chosen in three categories: students with reading comprehension pretests in the bottom 17 percent nationally, in the 18-35 percentile range, and between the 36th and 50th percentiles. In the case of a student transferring or dropping out of school in years one or two, back-up students were identified and, to the extent practical, followed. These detailed records often provided remarkable windows into the longitudinal impact of programs on schooling, including the organization, curriculum, and instruction students received.
As noted previously, two to four replication sites were visited for each program that had fewer than four longitudinal sites. Each replicate was visited only once and therefore studied in less detail: no independent quantitative data were gathered at the replication sites. Qualitative data gathering was targeted at specific components of programs and addressed the question of program replicability in diverse contexts. Data on replication sites contributed to conclusions about program implementation, and are explored more explicitly in the Final Report.
A unique feature of Special Strategies was its link to the nationally representative Prospects data set. (Puma, Jones, Rock, & Fernandez, 1993). The Prospects/Special Strategies data sets intersections are shown in Figures 4 and 5 (For a full description of the intersections, see the Special Strategies Final Report.) Prospects quantitative data gathered at each site included norm-referenced achievement test scores,5 archival data from student records, and questionnaires given to administrators, teachers, parents and students. Student questionnaire data were gathered at the third grade and above. Prospects data permit comparisons between Special Strategies outcomes and those achieved at more nationally representative sites (see Final Report, Chapters Twelve through Fifteen).
Figure 6 (on page 17) provides data on the quantitative
data gathering for Special Strategies and Prospects. Note that while the
overlapping designs and schedules for the first- and third-grade cohorts provided some
potential for using Prospects samples as Special Strategies controls, the
ninth-grade Special Strategies sample had no comparable Prospects sample,
and hence no control group.
5 The norm-referenced test chosen for
Prospects, and hence used in Special Strategies, was the Comprehensive Test
of Basic Skills, Fourth
Edition (CTBS-4, CTB/McGraw-Hill, 1989).
