In most multi-campus studies of college impact that have been conducted over the past four decades, investigators have relied on ordinary least squares (OLS) regression as the analytic method of choice. Recently, however, some investigators have advocated the use of Hierarchical Linear Modeling (HLM), a method specifically designed for analyses…

Critiques written by J. Bruce Francis, Gerald Gurin, Arthur W. Chickering, and G. Lester Anderson of Astin's "Four Critical Years" are discussed. Points of agreement and contention between the reviewers concern the effects of college on students. (SF)

The Cooperative Institutional Research Program, the largest ongoing study of higher education in the United States, is discussed in terms of its conceptual framework, technical aspects of data collection, and uses of the data in research and national policymaking. (Author/MSE)

Examines the design of service learning research using psychologist-philosopher Ken Wilber's fourfold scheme. By interfacing Wilber's two dichotomies (interior versus exterior and individual versus group), researchers can access a comprehensive framework for conceptualizing the full range of service learning outcomes to study. Wilber's framework…

The purpose of this article is to describe a system for protecting the anonymity of subjects in longitudinal research and for maintaining the security of data files. The basic system comes from the American Council on Education Cooperative Institutional Research Program. Data is collected from questionnaires, and converted to magnetic tape.…

First of a 2-part series concerned with: different types of inferential errors, single- vs. multi-institutional studies (MIS), longitudinal vs. cross-sectional data, statistical designs, measurement error effects, and measuring student-environment interaction effects. The best data is obtained from longitudinal MIS on student inputs and outputs,…

Originally published in July 1984, presents and describes a student development theory based on student involvement. Asserts that the theory can be used both by researchers to guide their investigation of student development, and by college administrators and faculty to help them design more effective learning environments. Discusses implications…

While any method of data collection is subject to certain unique types of systematic errors, the investigator has available several techniques for detecting such errors or for minimizing their effects on his conclusions. (Author/JB)

This study explored how logit, probit, and more traditional linear regression techniques compare in predicting college student retention, using data provided by registrars from a national sample of colleges and universities. Results showed little practical difference among the techniques, despite theoretical advantages offered by the first two.…

Fall 1982 data on first-time, full-time freshmen are presented as part of the Cooperative Institutional Research Program. The normative data are presented separately for women and men, and for 38 groupings of institutions. The major stratifying factors are institutional race (predominantly black versus predominantly white), type (two-year college,…

Methodological difficulties inherent in the design of studies of college impact are reviewed and an attempt is made to provide an improved research design. The discussion of research problems focuses on those of inferring causation. A characteristic three-component model of student development, comprised of student output, student input, and…

This report presents survey data (from 216,362 students at 403 higher education institutions) on the characteristics of students attending American colleges and universities as first-time, full-time freshmen. The normative data presented are reported separately for men and for women and for 35 different groupings of institutions. The major…

This report presents survey data (from 194,182 students at 382 higher education institutions) on the characteristics of students attending American colleges and universities as first-time, full-time freshmen. The normative data presented are reported separately for men and for women and for 35 different groupings of institutions. The major…