Division of Science Resources Studies
Professional Societies Workshop Series

Welcome and Introductions

Jeanne Griffith, Director, Division of Science Resources Studies, NSF

I would like to add my words of welcome to those of Alan [Rapoport], and I am thrilled to see how many people have come out for this meeting. We thought it was an interesting and important topic, but we were not anticipating this many people, so this is really terrific. I see a lot of familiar faces out there and we welcome all of you.

This is the fourth in a series of workshops that the Division of Science Resources Studies (SRS) has held with professional associations and other interested parties to try to participate in a conversation about data-related issues in science and technology. We try to have these meetings on a fairly regular basis, just to get together and have some discussion on important issues. I want to thank the members of the planning committee: Alan, the head of that committee; Carolyn Shettle; Joan Burrelli; Dick Morrison; and Deborah Collins. They put in a tremendous amount of work with this and I am very grateful for all of that work. I also appreciate the time that all of you are taking to attend this meeting and the time that all of our speakers have taken to prepare for this. It is a very important meeting for SRS, and we are really grateful for the efforts that people have made, and for people showing up on a gorgeous day.

Just a little background about the Division of Science Resources Studies, because I think some of you may be less familiar with us than others. We are part of the Directorate of Social, Behavioral, and Economic Sciences. The major goals of our division are somewhat different from those of other divisions in the National Science Foundation. Our division is charged to produce high-quality, timely, and relevant data and analyses about the domestic and international science, engineering, and technology enterprises. We are the statistical organization for the National Science Foundation. We also are charged to make those data and analyses accessible to a very wide community of users, and we define those users as people who make or influence, or are interested in policies and issues that affect our nation's science and engineering enterprise.

There have been a number of surveys on human resources, both education and labor force, on the nation's investment in research and development, on public opinion and public knowledge about science and technology, and on the facilities and instrumentation associated with science and technology.

On graduate education in particular, which is our focus here today, we have a number of specific surveys. I won't name each one of them, but we have different surveys that examine different aspects of the graduate experience, and the population of scientists and engineers. We put out several major reports on the basis of those surveys. I hope you were able to see the xhibit out front that displayed a number of those reports, which include special topical reports and compendia on various aspects of science and engineering, and focus specifically on graduate education. Our data are also available on the Web, and there are stacks of cards out there that have our Web address on them. I hope you will help yourself to those cards and help yourself to our Web site.

I have some background information that I want to give you about this workshop, but before I do that I would like to introduce Bennett Bertenthal, who is the assistant director for the Directorate of Social, Behavioral, and Economic Sciences. He is a professor of psychology at the University of Virginia, and we are fortunate to have him at the National Science Foundation and at this meeting today.

Bennett Bertenthal, Assistant Director, Directorate for Social, Behavioral, and Economic Sciences, NSF

I, too, would like to take this opportunity to welcome all of you to this workshop. As you have just heard, I have been at NSF only for a brief period of time, less than a year, although given the number of hours that I spend in this building, I have calculated that it comes out to more than two years of work time already. Prior to coming here, I was a professor of developmental psychology and cognitive science at the University of Virginia for the past 18 years. For a large part of that time, I was the director of a large interdisciplinary training program and had an opportunity to experience a number of dramatic trends in graduate education, some of which were very positive, some of which were somewhat surprising and of concern to me and my colleagues. Two issues in particular had to do with the difficulty of attracting the best and the brightest to graduate schools, and once they had decided to come, keeping them in graduate school.

This second issueretention of graduate studentsrepresents the main topic of this meeting, and I am pleased that SRS has organized this workshop, because it is high time that we begin taking stock of this very critical issue. In the few minutes' time that I have, I would like to offer some of my own reasons for supporting the need for additional data and analysis. Let me do so by sharing with you some intriguing data that was provided by a historian of science, Derek De Solla Price.

This figure is a semilog of the cumulative number of scientific journals founded worldwide as a function of time. What is represented by this main function is a positive straight line, which represents a pure exponential. And as you can see, the first scientific journals were published around 1680. For those of you who are not really in tune with what was happening then, that was about the time that Newton published his Volume Principea. Over the next two and a half to three centuries, the number of publications increased by a factor of 10, every 50 years until Price made this curve in 1950. The number of journals founded up to that point was about 100,000. The curve extrapolates to one million by the end of the century. We are now near this point, but as you and I both know, we are nowhere near one million journals at this time. In fact, currently only about 40,000 journals exist in the world. This may explain why so many of us lament that by the end of the week there is nothing left to read. Whether you have that problem or not, I think you will agree that sometime between the 1950s and now, those 300 years of exponential growth have come to and end.

Now Price claimed that any quantitative measure of science would look the same. David Goodstein, who is an eminent physicist and currently the provost at Cal Tech, put this claim to the test by plotting the number of U.S. physicists receiving their Ph.D.s. The first Ph.D. was awarded around 1870 and then, as you can see in this graph, exponential growth began up to about 10 per year in 1900. In around 1930, we reached 100 per year. By 1970, we were producing 1,000 per year. Extrapolation of this curve to right now should have led to 10,000 a year and 1 million by the middle of the next century. But that is not what happened. What happened was that growth stopped abruptly in 1970, and it has been fluctuating around 1,000 Ph.D.s in physics per year ever since.

By now, you are probably beginning to recognize that something exceptional happened around 1970. Many hypotheses for this cessation of exponential growth are possible, and I am sure each has its advocates. Although ultimately we would like to understand why this change occurred, for the moment I would simply like to underscore the significance of recognizing this phase shift in the history of science and the education of graduate students. Without the benefit of this recent quantitative analysis contributed by both Price and Goodstein, we might still be denying that a substantive change took place. It is difficult, however, to refute this claim when we have such compelling data.

For similar reasons, the provision of new data will provide us with a far better understanding of the status of graduate education. Currently, we are in need of much more data about the status of graduate students, how quickly they matriculate, and how likely they are to receive their Ph.D.s. I hope that this workshop will prove to be an important step in meeting the needs for additional data. I wish all you much success in your efforts today, and I look forward to hearing about the outcome of this meeting. Thank you.

Jeanne Griffith:

Thank you, Bennett. This workshop, as I mentioned, is the fourth in a series of workshops that the Science Resources Studies Division has held, and the purpose of these workshops is to share information on data collection and analysis, activities of our division that are of mutual interest to professional societies. Another purpose is to strengthen the ties between SRS and professional societies, because we have strong common interests in understanding the different phenomena related to science and engineering in the country. The first three workshops focused on labor market outcomes. They each had slightly different perspectives, but that was the broad focus in those three workshops. For this one, we decided to focus more on the graduate student experience and on attrition, in particular.

A number of events over the last couple of years, both in the National Science Foundation and in the science community, have really focused our attention on the graduate experience. I want to review for you a few things that have been happening in SRS related to our graduate education activities. There was a workshop examining the feasibility of using university accounting and financial systems to identify students by name and demographic characteristics, so that we could identify students who were receiving NSF support and perhaps include them in our surveys, and certainly have more information on them. There was a roundtable discussion on strategies and tactics for increasing the participation and the success rates for women and members of historically underrepresented minority groups , as well as persons with disabilities in science, mathematics, and engineering. Claudia Mitchell-Kernan, who is our keynote speaker this morning, chaired that roundtable. In the near future we are going to hold a conference on innovative practices and policies that graduate institutions are using to attract and retain graduate students, as well as policies that these institutions are pursuing to change the graduate experience.

We have also done a number of things to modify our data programs, and we think that these will have important payoffs down the road in providing more information about the graduate experience. We have expanded the sample of our biannual survey of doctorate recipients to include all recent Ph.D.s who have received NSF fellowships. We are also enlarging the sample of that survey to include a larger number of recent Ph.D.s and we will try to speed up the processing of the recent Ph.D. parts of that sample so that it will be available more quickly for analysis by the user community. We are analyzing the 1995-96 National Post-Secondary Student Aid Study, which is conducted by the National Center for Education Statistics (NCES), to get information about the overall sources of support used by graduate students, and to look at those sources of support by the social and demographic subgroups in the population.

NSF has very strong concerns, obviously, about graduate education. The education, the training of graduate students in science and engineering is obviously a matter of great national importance. It is relevant to the missions of all of the science agencies, but it is of special concern at NSF. We support students in science and engineering directly through graduate fellowships and traineeships and indirectly through research assistantships on NSF grants, which are given for supporting various kinds of research. We are also committed to increasing the participation and the success rate of historically excluded groups in science and engineering education. We are looking very carefully at programs that can help to increase the entrance and success of women or historically underrepresented minority groups or persons with disabilities. We want to identify viable strategies and tactics to attain those objectives.

We are particularly concerned with graduate student attrition because we are concerned that doctoral education is costly, not only for the society in general but also for the student. It is a very important undertaking, and we need to know more about it. There was a conclusion in a report that was recently released by the National Research Council, called The Path to the Ph.D.: Measuring Graduate Student Attrition in the Sciences and Humanities. There is a quote from that report, which I would like to read to you: "Accurate and reliable statistics that document enrollment patterns, degrees conferred, attrition rates, and postgraduate plans are vital to educators and policymakers for monitoring the performance of the graduate education system. They are also useful to prospective graduate students in making choices about whether and where to enroll." But the problem is, we do not know very much about graduate student attrition. We do not know very much about people who begin a science and engineering graduate program, but who end up dropping out or terminating their studies short of a Ph.D..

Besides not knowing much about the overall extent of doctoral student attrition, we do not know anywhere near enough about factors that influence whether or not students complete their doctoral degrees. We do not know about the impacts that such decisions have on their future earnings and labor force experiences. Therefore, I am very much looking forward to this workshop, which will focus on the research, the data, and the data needs that we have for determining the extent and the causes and consequences of graduate student attrition. Hopefully, we will be able to look at the abilities and the limitations of existing information to answer our questions in this area. We will be able to identify some further information that may need to be collected, and I hope that we can identify some roles for the Division of Science Resources Studies and for the professional associations, i.e., what SRS might do to improve the state of information.

I would like at this point to introduce our keynote speaker, who is going to speak about "Is Attrition Important and Why?" We are very fortunate to have as our keynote speaker Dr. Claudia Mitchell-Kernan. She is vice chancellor of academic affairs and dean of the Graduate Division at UCLA. She is an anthropologist and has joint academic appointments in the Departments of Anthropology and Psychiatry and the Bio-Behavioral Sciences. She is also a member of the National Science Board and is chair of the Indicators Subcommittee, which oversees a good deal of the work that our Division does, so we are delighted to have her here this morning.

Previous Page Workshop HomePage Next Page

attrition uses $ limitations
qualitative approaches professional future directions

SRS HomePage