PROGRAM DESCRIPTION
The training program in Quantitative Methods at the University of Illinois is located in the Department of Psychology. The core faculties of the program are members of the Division of Quantitative Psychology, and remaining faculty have primary affiliations with other divisions and with other departments on campus.
The purpose of the program is to train pre- and post-doctoral students in the various sub-specialties of quantitative psychology, including measurement, statistics, decision theory, evaluation research methodology, experimental design, psychological scaling, and mathematical modeling. The training is research oriented, with emphasis on formal theory, basic research, and applications in such fields as mental health, educational assessment, and personality psychology. Trainees are encouraged to take courses and participate in research in at least one substantive problem area in conjunction with their work in quantitative psychology.
The support for the Quantitative Program comes from an N.I.M.H. training grant awarded in 1997, entitled "Quantitative Methods for Behavioral Research". The primary purpose of this program is to train specialists in quantitative methods with an emphasis on applications to mental health settings.
The Quantitative Psychology program is one of seven graduate training programs within the Psychology Department. The areas of specialization represented by these programs are: Biological, Clinical/Community, Cognitive, Developmental, Visual Cognitive& Human Performance, Social-Personality-Organizational, and Quantitative. The nine primary faculty of the Quantitative Program include specialists in psychological measurement, mathematical modeling, research design, assessment, behavioral statistics, judgment, decision making, and psychological scaling. In addition, approximately fifteen other faculty members from Psychology, Statistics, Educational Psychology, and other social and behavioral science departments contribute to the training of our students by offering relevant courses in quantitative methods and applications, (e.g. Sampling, Attitude Measurement, Clinical Assessment, Psycho physiological Measurement, and Time Series Analysis) and by cooperating in the supervision of predoctoral and postdoctoral trainees’ research.
The Quantitative Program has been in existence since
the mid 1950’s. Since its inception, it has enjoyed a reputation as one of
the most comprehensive and influential programs in the world. The training
program subsumes four separate levels of training: (a) an M.S. degree in Applied
Psychological Measurement; (b) an M.S. in Applied Statistics with specialization
in Behavioral Science (a joint program with the Statistics Department); (c)
a Ph.D. degree in Quantitative Psychology; and (d) a postdoctoral training program
in Quantitative Methods. In addition to the specialists trained in these programs,
approximately one-third of the students from other programs in the Psychology
Department pursue minors in quantitative methods. Graduates of our program
are leading scholars on the faculties of many of the top research universities
in the
In 1980, with the support of an NIMH training grant, a postdoctoral training component was added to our program. For more than two decades, this program has attracted applications from individuals from a wide variety of substantive specialty areas wishing to enhance their skills through advanced training in quantitative methods. Judging from (a) the positions secured by trainees who completed such training, (b) their own statements about the impact of that training on their research and career development, and (c) the quality and quantity of research subsequently completed by these trainees, the postdoctoral training program has been, and continues to be, very successful. It has been renewed by NIMH every five years since 1980.
The program supports three postdoctoral trainees each academic year. Most of these trainees spend two years in the program. A typical postdoctoral trainee devotes about 60% of the time and effort during the first year to course work and 40% to collaborative research with program faculty. In the second year, a larger portion of the trainees’ time is devoted to research, usually involving application of newly acquired skills and methods to problems of interest to the trainee. Postdoctoral trainees are encouraged to become involved in the ongoing research programs of one or more program faculty, as well as to initiate independent research. It is the hope that this model will help the trainee become a self-motivated, independent scholar.
All program faculty, students, and postdoctoral trainees participate in a biweekly research seminar where proposed, ongoing, and recently completed research is presented and discussed. Each semester, three or four visitors who are doing important work in quantitative methods or applications present their work to the seminar.
In our view, the training of researchers concerned with the role of basic psychological processes involved in mental illness and health requires a foundation of course work in research design, psychological measurement, multivariate models, and other statistical methods suited to analyzing these complex phenomena and processes. Likewise, in both experimental and “quasiexperimental” paradigms for evaluation of mental health treatment, prevention, and delivery programs, specification and measurement of antecedent, process, and outcome variables are basic to understanding program operation and to inferring program effectiveness. Multivariate statistical models are needed for making inferences about relationships among program, participant, and outcome variables, and for generalizing findings across times, persons, and program types. A typical evaluation study involves problems in attitude measurement, behavioral assessment, determination of cost effectiveness, and measurement of change. In settings where randomized experiments are not possible, measurement considerations and models of data analysis are especially important; in those situations, careful application of factor analytic methods, regression models, or structural equation models is often appropriate, but requires a well-trained specialist. Also, there is a need for basic research to develop new techniques and models, and to investigate ways in which existing methods and models can be adapted and improved for investigation of behavioral and mental disorders, assessment and diagnosis of psychopathology, and related problems.
The training program offers opportunities for advanced training and research in several new areas of specialization, including: (a) behavioral decision-making; (b) social network analysis and modeling; (c) analysis of categorical data; (d) combinatorial approaches to data analysis; (e) measurement models for psycho physiological phenomena and methods for the analysis of psycho physiological data. Each of these specialties reflects the research and teaching interests of one or more faculty members. Given below are brief descriptions of two of these new areas of specialization; decision making and social network analysis.
Decision Making. During the last few years the foundation has been laid for developing a major new research and training emphasis in decision theory and behavioral decision-making, with applications to clinical diagnosis and related problems in medical and mental health settings. A discussion follows of the general nature and significance of this new program component.
Decision making is the major activity of many health service professionals. From choices among diagnostic tests and procedures to final diagnosis and treatment, the judgments and decisions of physicians, clinical psychologists, and other health care specialists affect the lives and well being of many. Although many medical decisions have been routinized, virtually all are characterized by uncertainty – all information may not be available, information may not be completely reliable, and the outcomes of particular courses of action may not be perfectly predictable. Decision making under uncertainty has been a rapidly expanding area of psychological research for the past 25 years. During this time, both prescriptive theories to study how decisions ought to be reached and descriptive theories that portray how decisions actually are reached have been developed.
The application of certain prescriptive theories (e.g., decision analysis, or Subjective Expected Utility theory) has become widely accepted among medical educators and researchers. Such applications, however, are not without problems (e.g., in the area of utility measurement) and are often rejected by practicing physicians who fee that these normative formal approaches fail to utilize important information such as holistic impressions. Some researchers have approached this question by examining how expert medical decision makers go about solving problems. This work attempts to isolate the cognitive components that differ between experts and novices. As in other areas thus investigated, it seems that the main expert-novice difference lies not in perceptual or memory skills but in a different and more efficient knowledge representation of the problems.
Another methodological contribution of psychological decision theory to the medical diagnosis issue has been signal detection theory, originally derived from studies of radar operators. Clinical diagnosis can be construed as a signal detection problem, in which observers must distinguish true signals from noisy background.
Cognitive limitations that affect human information processing in many ways have been shown to lead to the development of cognitive shortcuts, or rules of thumb, which most often lead to good judgments but can also result in biases. Such heuristics and resulting biases have been shown to occur in clinical judgment and decision situations as readily as in the psychology laboratory. Furthermore, decision heuristics seem to be descriptive of certain aspects of physicians’ information processing.
There is great potential for psychological decision research to contribute to the understanding and improvement of clinical judgments and choices. Moreover, there is a growing consensus among medical and health educators about the need to provide practitioners and administrators in medicine and public health with more thorough and more extensive training in the theories and techniques developed by decision theorists. Demand for such training currently seems to exceed availability. We are fortunate to be able to provide training in this important and rapidly growing area.
There is campus-wide interest and participation in the area of decision theory at the University of Illinois. Faculty in the training program and department with interests in judgment and decision problems include: David Budescu (decision making and judgment), James Davis (group decision processes; jury decision making) Lawrence Jones (contextual effects in judgment), Janet Sniezek (decision making processes in organizations), Christopher Wickens (decision making in man-machine environments), and Robert Wyer (social judgment; role of memory processes in decision and judgment).
The primary faculties of the training program are from the Quantitative and Clinical Divisions of the Psychology Department. Our training program produces specialists in the areas of behavioral decision-making, analysis of categorical data, combinational approaches to data analysis, measurement models for psycho physiological phenomena and methods for the analysis of psycho physiological data. The faculty participating in the training of these students is the largest group of specialists in measurement and multivariate modeling in the country. Most of them combine technical expertise with backgrounds in applied research. There are also numerous faculties in psychology and other departments who offer relevant courses and conduct research of potential interest to trainees.