The National Longitudinal Study (NLS) of the High School Class of 1972 Data Base: Implications for Use in Doctoral Training Programs in Higher Education

33 critical policy question for many multi-campus systems is how to effect the most appropriate match between student and institution. Even if one cannot change the institutional system, one can get a bigger "bang for the educa­ tional buck” by efficacious matching of students and institutions. In ef­ fect, such matching is designed to capitalize on interaction effects. The only part of the book where I look at interaction effects is in the chapter on career development. College impact was assessed separately for groups of students who had been sorted into relatively homogeneous categories according to their career aspirations at the time of college entry. As it turned out, some variables affected career progress similarly for all groups (main ef­ fects), but a number of variables turned out to have unique effects in parti­ cular groups (interaction effects). Clearly, future research on college im­ pact will have much more direct relevance to policy if it focuses more inten­ sively than I did on interaction effects. It is somewhat remarkable that two respected scholars of higher educa­ tion— Anderson and Chickering— would read the same volume and draw final conclusions that seem to be so diametrically opposed. Anderson feels that my policy perspectives, if taken seriously, would ". . .be completely disrup­ tive of American higher education as we know it." Chickering, by contrast, concludes that ". . . if the directions of change to meet the challenges to be faced by higher education during the next twenty years continue to ignore these findings; it is hard to be optimistic about the contribution of col­ leges and universities to individual development and to more general social needs." Regardless of which perspective one prefers, Anderson has initiated a dialogue about policy and values in higher education which should be expan­ ded and explored more intensively by all of us who study higher education. THE NATIONAL LONGITUDINAL STUDY (NLS) OF THE HIGH SCHOOL CLASS OF 1972 DATA BASE: IMPLICATIONS FOR USE IN DOCTORAL TRAINING PROGRAMS IN HIGHER EDUCATION* by James L. Morrison Professor of Education University of North Carolina at Chapel Hill The National Longitudinal Study of the High School Class of 1972, spon­ sored by the National Center for Education Statistics, provides a rich data base for studies of the vocational, educational, and personal development of high school graduates and the psychological and social factors that contri­ bute to that development. The purpose of the data collection effort is (1) to establish an empirical basis for developing educational policy concerning access to educational and vocational opportunity and (2) to extend our know­ ledge of human development in those years of transition from high school to adulthood. Peng, Stafford, and Talbert (1977) have reviewed the studies using NLS data. Most studies to this date have been primarily descriptive, not analy­ tical. The major reason appears to be that relatively few researchers, par­ ticularly those in academic settings usually more concerned with analytical *Revised version of a paper presented to the Annual Meeting of the Associa­ tion for the Study of Higher Education, Chicago, Illinois, 1978. 34 studies, are aware of this data source. This paper will first describe the NLS sample and data collection efforts, briefly review the nature of informa­ tion available for each data collection effort, and outline possible implica­ tions that this data base has for use in graduate training programs in higher education. NLS Sample and Data Collection The population for the NLS consists of all 1972 twelfth graders enrolled in all public, private, and church-related high schools in this country. The sample design is a highly stratified, two-stage (schools as first stage; stu­ dents as second stage) probability sample. The school sampling frame was stratified by type of control, region, school size, proximity to postsecon­ dary institutions, percentage of minority group enrollment, community income level, and degree of urbanization. The reader should see Bailey (1976) for a more complete description of the sample design. The study was initiated in the spring of 1972. In the base-year survey (n = 16,683), each senior was requested to complete a lengthy questionnaire and to take a 69 minute test battery. In addition, administrators in survey high schools (n = 1,070) were requested...