Home > The Study > About > Methodology

Download PDF of this section

Methodology

The Department of Public Policy at the University of Connecticut emphasis on conducting complex and policy-relevant research is in part a function of its educational mission. The Department includes the Graduate Program in Survey Research Methods, which offers graduate degrees in survey research. Students and faculty are regularly involved in conducting original research to improve survey research methodologies and to make professional contributions to the field.

The Department and The Center for Survey Research (CSRA) conduct approximately 100 survey research projects per year. All of these projects are externally funded and range from large-scale, complex national policy evaluation projects to local community surveys. The Department specializes in designing and implementing survey projects that require complex methodologies and rigorous standards of data quality. Clients are state agencies (in Connecticut, New Jersey, Minnesota and Washington), federal agencies and commissions ( U.S. Department of Homeland Security, U.S. Department of Education, National Institutes of Health, EPA, U.S. Department of Labor, the Commission on Philanthropy and Civic Renewal ), private and public foundations (the Ford Foundation, the Freedom Forum, the Chase Foundation, the First Amendment Center, the Criminal Justice Policy Foundation), and news media organizations ( The Hartford Courant ).

CSRA's data collection facilities  

CSRA maintains a variety of in-house resources for designing and implementing high quality, complex survey research. The professional staff consists of 12 full-time researchers, including a statistician, two computer-assisted telephone-interviewing (CATI) system programmers, a telephone center operations manager, a business manager, and seven professionally trained survey research analysts (several of whom hold doctorates with specialized training in survey research methodology). CSRA also employs 12 to 15 graduate research assistants annually, all of whom are studying in the Department of Public Policy's Graduate Survey Research Methods Program.

Questionnaire design  

Dr. David Yalof and Dr. Kenneth Dautrich served as Principal Investigators for the study. These scholars led a team that included CSRA researchers and outside scholars and experts to develop the questionnaires and survey methods.

The mode of administration was a key factor in the design of the survey instrument. The study utilized self-administered questionnaires administered in schools around the country. Questions were designed with this method in mind and rigorously tested to ensure that all respondents would understand each question the same way. Four high school based populations were surveyed with a unique questionnaire. These populations included students, teachers, principals, and school buildings. For the school building survey, the principal identified the most appropriate administrator at the school to answer questions about characteristics of the school building.

Prior to being administered to the total sample, the questionnaire and field protocols were pre-tested with three schools. This exercise provided researchers with valuable feedback from respondents, which was used to modify and finalize the questionnaires.

The four questionnaires used in the study are included in the appendix to this report.

Sample design  

The survey consisted of a multi-stage cluster sampling design conducted in two stages across three separate units of analysis: schools, students, and faculty. Four separate surveys were administered. Initially, a survey was administered to a random sample of 544 separate high school buildings around the country. In the second wave of the study, 327 of the initial set of school buildings completed a more intensive level of data collection. In these schools, separate surveys were administered to administrators, faculty, and students. The samples of schools, school principals, teachers, school administrators, and high school students are representative of the populations from which they were drawn.

The sample for this project was based on a database licensed from Educational Directories of Schaumberg, Illinois . This database, published in print format as "Patterson's American Education," is a comprehensive single source database of both public and private high schools in the US . Table 1 displays the sample frame:

The sample was stratified to proportionately represent public and non-public schools of different sizes. In the first stage of the sample design, schools were selected with equal probabilities of selection. In the second phase, the survey was designed to interview all faculty and students at these schools. Consequently, the sample was designed to provide self-weighting data at the institutional level for schools and at the individual level for faculty and students.

	This web site is produced by J-IDEAS
	J-IDEAS is funded in part by the John S. and James  L. Knight Foundation's High School Initiative and Ball State University.		J-IDEAS Department ofJournalism Ball State University, Muncie, Ind. 47306 (765) 285-8923 Terms of Service