Mission

The Joint Program for Survey Methodology is a unique educational structure for the United States. It blends together faculty from diverse disciplinary backgrounds and multiple institutions, all devoted to training in state-of-the-art methods in survey research. The Joint Program is in essence, a department of survey methodology. Its primary goal is to educate the current and future professionals of the Federal Statistical System, but it has the wider goal of offering training to all qualified students, regardless of the employment sector of interest to them. Starting in the Fall, 1993, it has offered an MS degree in Survey Methodology. Since the Fall of 1999, it has also offered three non-degree programs, including two certificate programs and a citation program. A PhD program in survey methodology was added in the Fall of 2000.
 

History

The idea for the Joint Program was part of an 1990 initiative of the Federal Statistical agency heads, the then current head of the OMB Statistical Policy Office, and then chair of the Council of Economic Advisors. The idea of a graduate educational and research unit serving the Federal agencies was prompted by widespread belief that recruiting staff with the interdisciplinary knowledge needed for large-scale surveys and censuses was not being facilitated by the traditional disciplinary graduate degree programs. For example, while products of statistics departments were well-versed in advanced statistical estimation, they typically had little practical knowledge of complex sample design or of survey instrument development. The mismatch between the disciplinary organization of most universities and the technical staffing needs of the system required a new academic organization. This problem is not peculiar to the United States, and other countries have built educational institutions within the government statistical agencies themselves. The legislative initiative called for a graduate education and research center offering courses in the DC area.

In December, 1992, after an open competitive process, the National Science Foundation awarded a $4.1 million five year cooperative agreement to the University of Maryland at College Park. Maryland had joined with the University of Michigan and Westat, a survey organization in Rockville, MD, to propose the Joint Program.

The NSF support was used to build a new department on the College Park campus, with contributions from three organizations simultaneously. The Westat contribution includes instruction in the core and elective courses of the curriculum. The Michigan contribution includes the permanent location of two faculty members on the College Park campus, the construction of a two-way video/audio telecommunication system for transmitting courses back and forth between the Ann Arbor campus and the College park campus, and the commitment to offer a second site of the Summer Institute courses in the DC-area. 

Many meetings between Joint Program faculty and leaders of the Federal Statistical agencies have occurred since December, 1992, and several agencies have established policies regarding their financial and other support of current staff attending the credit-bearing courses of the Joint Program. Some agencies mount internal competition for selection to the MS degree program, giving the winning applicants tuition support and half-time release for the program. Others support course-by-course tuition costs. JPSM is now funded primarily by the Interagency Council on Statistical Policy.
 

Educational Vehicles

There are several teaching vehicles now in progress:

• The MS degree in Survey Methodology, started on Fall, 1993, has both statistical and social science areas of concentration.
   
• The PhD degree in Survey Methodology, started in Fall 2000, also has both statistical and social science areas of concentration.
   
• The JPSM Citation in Introductory Survey Methodology, started in Fall 1999, is a non-degree program consisting of one semester-long course and eight short courses.
   
• The Graduate Certificate in Intermediate Survey Methodology, started in Fall 1999, is a non-degree program consisting of six semester-long courses.
   
• The Graduate Certificate in Survey Statistics, started in Fall 1999, is a non-degree program consisting of six semester-long courses.
   
• Noncredit short courses for survey professionals, started in 1993, cover a wide range of topics and are presented by leading researchers in the field. The Joint Program offers ten to twelve courses per year.
   
• DC- area offerings of the long-standing Michigan Summer Institute in Survey Research Techniques began in Summer, 1993.
   

The MS Program

The MS program is a 46 credit hour, two year program. The program offers a statistical science concentration for those interested in specializing in sample design, estimation in complex samples, variance estimation, statistical measurement error models, and statistical adjustments for missing data. The social science area is designed for students who will specialize in questionnaire design, computer assistance in data collection, effects of mode of data collection, cognitive psychological insights into measurement, and efforts to reduce various non-sampling errors in surveys.

The two areas share a core curriculum that includes a two term sequence in survey design, collection, and analysis here the students actually plan and conduct a survey. The core also includes courses in applied sampling, data collection design, a course on the design and functioning of the Federal Statistical System, a survey design seminar where design and analysis consulting skills will be honed, a "randomized and nonrandomized research design" course, blending classical experimental design with quasi-experimental or observational study design, and a course in total survey error perspectives on survey quality.

The statistical science area has some additional courses that resemble those in traditional Master's programs in statistics departments (e.g., probability and mathematical statistics). Other courses are novel: a course in inferential issues in complex survey analysis, a course treating weighting and imputation, ratio and regression estimation, small area estimation, and sampling methods for rare populations. Similarly, the social science area has some courses that were constructed from scratch: an advanced course in questionnaire design, a course in the cognitive and social theoretical foundations of survey measurement, a course in practical methods of analyzing complex survey data, a course in survey management, reviewing budgeting, administrative, and organizational issues of large scale surveys.

To target the working student many of the courses are held in the late afternoon or evenings. Some courses are held within the statistical agencies themselves to reduce the burden on working students.
 

The Future

It is no small matter to build an organization that is a collaboration of two educational institutions, one commercial organization, and faculty in several other locations, designed to serve ten large statistical agencies but also forty diverse other agencies with statistical functions. The Joint Program is just beginning. It has, however, the potential of not just increasing the quality of technical staff in the Federal Statistical System but also of enriching the field of survey statistics and methodology itself.