Assessing Research-Doctorate Programs: A Methodology Study

Committee Task Review and revise the methodology used to assess the quality and effectiveness of research doctoral programs. Explore new approaches and new sources of information about doctoral programs and new ways of disseminating these data. Recommend whether to conduct a full assessment using the methodology developed in by the committee

History of NRC Assessments 1982 “Assessment of Research-Doctorate Programs in the United States” Lyle V. Jones (Co-Chair) Gardner Lindzey (Co-Chair) 1995 “Research-Doctorate Programs in the United States: Continuity and Change” Marvin L. Goldberger (Co-Chair) Brendan Maher (Co-Chair)

Perceived Strengths of Prior NRC Assessments Authoritative source Comprehensive Clearly stated methodology Temporal continuity Widely quoted and utilized

Perceived Weakness of Prior NRC Assessments Spurious precision of program rankings Confounding of research reputation and educational quality Soft criteria for assessments of programs Ratings based on old data

Weaknesses continued… Poor dissemination of results for some audiences Taxonomy categories out of date Validation of data inadequate

Design of the Methodology Study Formation of a committee. Definition of tasks. Panel meetings to define questions, discuss methodology. Panels:  Taxonomy and interdisciplinarity  Quantitative measures  Student processes and outcomes  Reputation and data presentation Pilot trials of questionnaires, taxonomy.

Recommendations Spurious precision issue: The committee recommends a new statistical methodology to make clear the probable range of ranking for each assessed academic unit.

Alternative Approach to Rankings to Convey Rating Variability Draw ratings at random. Calculate rating for that draw. Repeat process enough times to reach statistical reliability. Present distribution of ratings from all the draws.

Recommendations continued… Research versus education issue: –Drop reputational estimate of education quality as not independent of the reputational estimate of program quality. –Add quantitative indicators of educational offerings and outcomes.

Program Measures and a Student Questionnaire Questions to programs –Size –Student characteristics and financing –Attrition and time to degree –Competing programs

Program Measures and a Student Questionnaire continued… Questions to students in selected fields –Employment Plans –Professional Development –Program Environment –Infrastructure –Research Productivity

Recommendations continued… Soft criteria issue: Add quantitative measures concerning research output, citations, student support, time to degree, etc.

Examples of Indicators Publications per faculty member Citations per faculty member Grant support and distribution Library resources (separating out electronic media) Laboratory Space Interdisciplinary Centers

Recommendations continued… Poor dissemination issue: –Add analytic essays to archival book output. –Add updateable current web output. –Add electronic assessment tools. –Add links from professional societies.

Recommendations continued… Taxonomy issue: –Update 1995 taxonomy. –State clear criteria. – Consult professional societies, administrators and faculty. – Allow for two academic categories (rated programs and emerging fields). –Named subfields to help universities classify their programs. –Allowed faculty to be in more than one program. –Included two sub-threshold humanities fields (classics and German) to maintain continuity.

Recommendations continued… Validation issue: Conduct pilot studies and institute checks, both by institutional respondents and by external societies.

Pilot Institutions University of Maryland Michigan State University Florida State University University of Southern California Yale University University of Wisconsin at Milwaukee University of California, San Francisco Rennsalear Polytechnic Institute

What’s next Obtain financing for the full study from both federal and foundation sponsors. If funding is obtained: –Full study would begin in Spring, 2004 –Data collection in 2004/2005 for previous academic year. –Final report in summer 2006

Conclusion The study that the Committee recommends is a BIG undertaking in terms of survey cost and the time of graduate programs and their faculty. Why is it worth it? It will provide faculty, students and those involved with public policy an in-depth look at quality and characteristics of those programs that produce our future scientists, engineers, and those who help us understand the human condition.

Committee Jeremiah Ostriker, Princeton, (Astrophysics), Chair Elton Aberele, U. of Wisc (Ag) John Brauman, Stanford U. (Chem) George Bugliarello, PolyNY (Eng) Walter Cohen, Cornell U. (Hum) Jonathan Cole, Columbia U. (Soc Sci) Ronald Graham, UCSD (Math) Paul Holland, ETS (Stat) Earl Lewis, U. of Michigan (History) Joan Lorden, U. of Alabama- Birmingham (Bio) Louis Maheu, U. de Montréal (Soc) Lawrence Martin, SUNY-Stony Brook (Anthro.) Maresi Nerad, U. Wash (Sociology & Education) Frank Solomon, MIT (Bioscience) Catherine Stimpson, NYU (Hum)

Sub Committee – Panels STUDENT PROCESSES AND OUTCOMES QUANTITATIVE MEASURES TAXONOMY AND INTERDISCIPLINARITY REPUTATIONAL MEASURES AND DATA PRESENTATION Joan Lorden (Chair) University of Alabama-Birmingham Catherine Stimpson (Chair) New York University Walter Cohen (Co-Chair) Cornell University Frank Solomon (Co-Chair) Massachusetts Institute of Technology Jonathan Cole (Co-Chair) Columbia University Paul Holland (Co-Chair) Educational Testing Service

Additional Panel Members STUDENT PROCESSES AND OUTCOMES Adam Fagen, Harvard Univ. (Bioscience, grad.student) George Kuh, Indiana Univ. (Education) Brenda Russell, Univ. of Illinois-Chicago (Bioscience) Susanna Ryan, Indiana U. (English, Woodrow Wilson Fellow) QUANTITATIVE MEASURES Marsha Moss, Univ. of Texas (Institutional Research) Charles E. Phelps, Univ. of Rochester (Provost & Econ.) Peter D. Syverson, Council of Graduate Schools

Additional Panel Members TAXONOMY AND INTERDISCIPLINARITY Richard Attiyeh,UCSD (Econ.) Robert F. Jones, AAMC (Bioscience) Leonard K. Peters, VPI (Computer Science) REPUTATIONAL MEASURES AND DATA RESENTATION David Schmidley, Texas Tech (President & Bioscience) Donald Rubin, Harvard (Statistics)

Project web-site