Really Cool
    Minority Data

Study of Instructional Improvement (SII) (ICPSR 26282)

Published: May 20, 2010

Principal Investigator(s):
Deborah Loewenberg Ball, University of Michigan. School of Education; David K. Cohen, University of Michigan. School of Education; Brian Rowan, University of Michigan. Institute for Social Research

Version V1

To meet the growing need for high-quality research on whole-school approaches to instructional improvement, researchers at the University of Michigan School of Education, in cooperation with the Consortium for Policy Research in Education (CPRE), conducted a large-scale, mixed method, longitudinal Study of Instructional Improvement to investigate the design, implementation, and effects on student achievement of three of the most widely-adopted whole-school school reform programs in the United States: the Accelerated Schools (ASP), America's Choice (AC), and Success for All (SFA). Each of these school reform programs sought to make "comprehensive" changes in the instructional capacity of schools, and each was being implemented in schools in diverse social environments. Each program, however, also pursued a different design for instructional improvement, and each developed particular strategies for assisting schools in the change process. In order to better understand the process of whole-school reform, Study of Instructional Improvement (SII) developed a program of research to examine how these interventions operated and to investigate their impact on schools' instructional practice and student achievement in reading and mathematics. The research program had 3 components: a longitudinal survey of 115 schools (roughly 30 schools in each of the 3 interventions under study, plus 26 matched control schools), case studies of the 3 interventions under study, and detailed case studies of 9 schools implementing the interventions under study (plus 3 matched control schools). Across all components of the SII study, the research examined alternative designs for instructional improvement, alternative strategies for putting these designs into practice in local schools, and the extent to which alternative designs and support strategies promote substantial changes in instructional capacity and student achievement in reading and mathematics. The most comprehensive component of SII was a large-scale, longitudinal, multisurvey study of schools. The use of survey research methods was intended to track the course of schools' engagement in comprehensive approaches to instructional improvement and to investigate the conditions under which this led to substantive changes in instructional practices and student achievement in reading and mathematics. The study design called for each school to participate in the study for a period of three years, although some schools voluntarily provided a fourth year of teacher, leader, and school-level information (no additional student-level data). In addition, survey researchers conducted interviews, primarily a telephone protocol with a parent or guardian of each cohort student in order to gather information on students' family background and on students' home and community environments. Researchers also gathered data from school leaders and others about the policy environments in which the schools are located. Another component of the research program involved the development of detailed case studies of a small number schools participating in the study. The case studies gathered observational, interview, and documentary evidence to better understand how instructional change processes unfolded in different school settings. Case studies were conducted in 12 schools operating in differently configured state and district policy environments. In each environment, researchers selected schools participating in one of the interventions under study as well as a "matched" control school. Finally, case study data was used to chart key similarities and differences in the design and operations of the interventions under study, to analyze how different design features affect operating strategies, and to better understand the general problem of how intervention programs can work to devise and "bring to scale" a feasible scheme for improving instruction in local schools.

Loewenberg Ball, Deborah, Cohen, David K., and Rowan, Brian. Study of Instructional Improvement (SII). Ann Arbor, MI: Inter-university Consortium for Political and Social Research [distributor], 2010-05-20.

Export Citation:

  • RIS (generic format for RefWorks, EndNote, etc.)
  • EndNote

Atlantic Philanthropies

Hewlett Foundation

United States Department of Education

National Science Foundation

Co-Nect Schools

University of Michigan

University of Pennsylvania

University of Washington

2000 -- 2004

2000 -- 2004

Due to Stata limitations, variables with characters for values will not have labels, however these labels are available in the SPSS and Stata setup files and in the ICPSR codebook.

For part 1 through part 46, the variable "BRTHDATE" has been blanked to protect respondent confidentiality.

To relieve the potential misalignment of files, the SII files have been merged to logically reflect chronological participation of students across a three-year period. This pertains to the assessment data, teacher log files, the Student Rating Form, and Student Motivation Form. However, each of these files contains a variable flag named "year" to indicate the exact school calendar year in which data were collected.

SII is described as a four-year study, but a "phased" or staggered collection design is utilized, where two separate student cohorts (Kindergarten through second grade and third grade through fifth grade) are followed longitudinally over three-year periods. Cohort A is designated as Kindergarten through second grade and Cohort B is designated as third grade through fifth grade.

Although SII was designed as a quasi-experiment with three "treatment" groups and a matched control group, after the SII sample was constructed, subsamples of treatment and comparison schools were not perfectly matched. For this reason, it is recommended that statistical analyses oriented to establishing the effects of "treatment" participation on outcomes using "control" schools as the counterfactual be conducted using some form of propensity score matching.

Although most sections of the Teacher Questionnaire (TQ) and School Leader Questionnaire (SLQ) are longitudinally designed and contain precisely the same items year-to-year, the data producer made some minor adjustments to the SLQ between survey administrations, and the PCK items for the TQ were changed during each administration to develop the battery of items necessary for scaling work. Variable positions (and variable names) changed year-to-year to accomodate these unavoidable shifts. Please see Appendix B and Appendix C for a variable cross-reference list for the TQ and SLQ, respectively.

For part 1 through part 60, various date and time variables have been converted to string. The 'MISSING' codes for these variables display as "59:59" or "09/09/9999", and the code for 'DON'T KNOW' displays as "08/08/8888". Please see the "Processing Notes" section of the ICPSR codebook for more information.

For further information regarding this study please refer to the SII Web site.

In order to construct a composite socioeconomic status (SES) measure, SII replicated the procedures commonly used in the development of education databases sponsored by the National Center for Education Statistics. It is important to note, however, that the SES measure developed by SII is not norm-referenced to a national school population. Instead, a standardized coefficient of SES represents a student's status compared only to other students in the SII population.

The main purposes were two-fold. First, to know the circumstances under which different intervention designs and strategies could be expected to produce changes in particular elements of instructional capacity in schools. Secondly, to know which elements of instructional capacity, when present in schools, worked to produce higher levels of student achievement in reading or mathematics.

Sampling efforts yielded 115 schools located in 45 different school districts, in 15 states, and 17 metropolitan areas. Overall, 31 America's Choice (AC) schools, 30 Success for All (SFA) schools, 28 Accelerated Schools Project (ASP) schools, and 26 comparison schools. For more information please refer to the "Sample and Study Design" section of the ICPSR Codebook.

All United States public elementary schools that began their affiliation with ASP, AC or SFA in the 1998-1999, 1999-2000, or 2000-2001 school years.

experimental data

face-to-face interview

self-enumerated questionnaire

telephone interview

Please see Appendix A of the ICPSR Codebook for response rates.



2010-05-20 ICPSR data undergo a confidentiality review and are altered when necessary to limit the risk of disclosure. ICPSR also routinely creates ready-to-go data files along with setups in the major statistical software formats as well as standard codebooks to accompany the data. In addition to these procedures, ICPSR performed the following processing steps for this data collection:

  • Created online analysis version with question text.
  • Checked for undocumented or out-of-range codes.


  • Data in this collection are available only to users at ICPSR member institutions.

ICPSR logo

This study is provided by ICPSR. ICPSR provides leadership and training in data access, curation, and methods of analysis for a diverse and expanding social science research community.