Results from a Survey of Master's Degree Education in Applied Mathematics
A Final Report to the Alfred P. Sloan Foundation from the Society for
Industrial and Applied Mathematics (SIAM)
December 31, 2003
Jim Crowley
SIAM Executive Director
Lee Seitelman
SIAM Survey Director
With funding from a Sloan Foundation grant, SIAM undertook a survey of applied mathematics master's programs on July 1, 2002. The goal of this effort was to understand the nature of the programs currently offered, and to present a workshop at a SIAM meeting at which the results of this survey could be presented, and the emerging professional science masters (PSM) programs could be spotlighted.
The Vision
The goals of this effort were to:
- Conduct a study of established master's degree programs in applied mathematics.
- Determine the principal characteristics of these programs (including disciplinary emphasis, starting point, goals).
- Identify master's degree programs with strong industrial connections.
- Make information available to institutions interested in developing programs, and to prospective students considering careers in applied mathematics.
- Disseminate this information at one or more SIAM meetings, through symposia on the creation and nurturing of professional science masters programs, and job fair activities for their graduates.
The interest in PSM programs developed, in part, from the need to augment traditional Ph.D. opportunities with programs more closely aligned with the needs of those nonacademic organizations that are expected to provide professional opportunities for an increasing fraction of professional graduates. The survey was expected to provide valuable information and guidance that could be used to assist students seeking directions for their programs of study, and universities seeking to provide those resources.
What Happened?
A survey was conducted by e-mail, with follow-up by additional e-mails, telephone solicitations and Web site searches to gather additional information. Preliminary results were determined, using the services of an external contractor, Marketing Partners, Inc. Presentation of the preliminary survey results, and a PSM workshop and job fair activities, originally scheduled for Toronto in June, 2003, were postponed until October, due to a series of SARS outbreaks in that city. Finally, presentation of the final survey results, and additional discussion about PSM programs, will take place in January, 2004 at the Joint Mathematics Meetings in Phoenix, AZ.
I. The Survey
The Society for Industrial and Applied Mathematics (SIAM) survey of applied mathematics master's programs was sent by e-mail to a broad cross-section of academic departments with interest in applied mathematics. A second e-mailing, telephone follow-up to many departments, and SIAM staff exploration of several departmental Web sites, completed the solicitation phase. The survey instrument (attached) was developed by Eleanor Babco of the Commission on Professionals in Science and Technology (CPST) for the Sloan Foundation in concert with the survey directors from three other professional societies (AIP, AGI, SIM), and input from ACS.
A total of 56 departments completed the questionnaire. Of these, 27 identified their programs as business oriented, including six of the eleven PSM programs in mathematics funded by the Sloan Foundation.
In light of the fact that the survey was sent to 1347 departments, the 4.2% response rate was initially considered disappointing. However, a closer inspection of the situation presented a different picture.
The proposal to Sloan discussed at length the intrinsic interdisciplinary nature of applied mathematics, and its historical connections with other fields. In an attempt to be as complete as possible, the e-mail list for the SIAM survey was purposely chosen to cast as broad a net as possible. The 1347 recipients included all the departments (including some in Computer Science, Statistics, and Industrial Engineering, among other fields) that SIAM knew to have an interest, however limited, in applied mathematics and related topics. Casting a "broad net" virtually guaranteed a low response rate.
To more fully appreciate the situation, one need only realize that data base information supplied by Eleanor Babco of CPST – from Joan Borrelli at NSF – identifies only 168 master's degree-granting institutions in the United States in mathematics and applied mathematics, only 38 Ph.D.-granting applied mathematics departments, and only 12 applied mathematics departments that award the master's as their highest degree. CPST's own listing of science master's programs in mathematics shows only 26 programs. Set against any of these standards, the response rate is quite reasonable.
Nevertheless, many of the better known names in applied mathematics departments – what might be called the "brand-name" institutions – did not respond to the survey, despite repeated invitations to do so. The absence of their data diminishes our appreciation of the success of the survey.
Further, the response of the Sloan Foundation-funded PSM schools was much less than we had hoped. Only 6 of the 11 programs in applied mathematics responded to the survey, despite repeated e-mails and telephone solicitations. After considerable prodding, the others provided only summary statistics on enrollments and graduates for the academic years 2001-02, 2002-03 and 2003-04. Eight of the eleven programs graduated no students in 2001-02, and another graduated only one.
It would be inappropriate to attempt to draw meaningful conclusions at this time about the efficacy of the PSM model in applied mathematics. Follow-up canvassing by e-mail of the Sloan programs, and review of the appropriate web sites, revealed that only two of these programs – Michigan State University (Industrial Mathematics) and Georgia Institute of Technology (Quantitative Computational Finance) – are sufficiently mature that they had significant numbers of graduates before 2003. Because the number of graduates in the Sloan programs was expected to more than double in 2003, and increase thereafter, a follow-up canvass in two or three years would likely provide useful information about not only the training of the students, and the establishment and ongoing development of the programs, but also the value of this training for eventual professional success, i.e., for enabling the students who graduate from these programs to move into the managerial and leadership positions for which the PSM is clearly intended to prepare them.
The data developed from the survey is attached.
II. The Dissemination
For a variety of reasons (late dissemination of the announcement of the survey in SIAM News, initial e-mail of the survey in mid-February, the necessity for repeated follow-up to solicit input, etc.), the final input for the survey was not received until about July 1. Not surprisingly, this delayed data reduction and interpretation. The primary data is attached to this report.
III. A Dissemination Bonus: Ensuring Accessibility of the Data
SIAM is committed to making the results of the survey as accessible as possible, by posting this information on its web site. The target date for this information to be online is January 31, 2004. Data from the survey will be complemented by links to each of the responding programs to enable the interested academician, or student, to quickly and easily locate useful information about existing programs. Finally, e-mail will be sent to mathematics and applied mathematics departments across the United States and Canada to alert them to its availability.
Lessons Learned
A number of valuable and instructive lessons can be learned from the survey.
First, the survey was much too long. This single fact had to have been a major cause of the glacial pace at which the responses came in. With many different responsibilities competing for the limited time of the department chairperson, the prospect of dealing with a questionnaire with many pages had to be a daunting, and not especially welcome, prospect. The response rate would likely have been considerably higher, and the follow-up considerably reduced, if the survey had been replaced by a two-part questionnaire, in which the first part merely asked if the institution had a master's degree program in applied mathematics, and the second asked for all the additional information, if the answer to the first question was affirmative.
Second, a common Excel spreadsheet format for the survey responses would have been helpful, both for the individual surveys, and to provide the logical infrastructure for integrating the results of the several surveys. (The second author made this suggestion to CPST in October, 2002, but it was never implemented.) Because cross-disciplinary comparisons were clearly of interest, from the very outset, to the social scientists associated with the Sloan Foundation, data collection should have been organized in a way that would most effectively support their needs.
Third, a better "top-down" overview of the integrated end product that was desired was needed. SIAM was unaware of a plan for a series of meetings between representatives of the several professional societies involved in the survey until the beginning of July, 2002, despite the fact that SIAM's proposal was not submitted to the Sloan Foundation until late in June, 2002. (A similar situation applied to the October, 2003 meeting.) Accommodating such change necessitated the rearrangement of time and budgetary priorities. Prior knowledge of the "game plan" would have allowed for careful attention to these needs in the planning and proposal process.
