Over the last decade, report after report after report has examined the state of doctoral education and concluded, among other things, that it is essential to shorten the amount of time it takes candidates to earn their degrees.

New data from the National Science Foundation suggest that the repetition of that message, and steps graduate educators and students alike have taken to confront the problem in recent years, have begun to make a difference, though arguably a small one.

The report, " Time to Degree of U.S. Research Doctorate Recipients," finds that the median "total time to degree" for recipients of research doctorates in 2003 continued the general pattern of a slight decline since a peak in the mid-1990s. (Total time to degree is the number of years between when a candidate received his or her baccalaureate degree and the Ph.D.) The 2003 figure was 10.1 years, which is down from a high of 10.8 years in 1996.

The "registered time to degree," which looks at the number of years that Ph.D. candidates actually spent enrolled in graduate school, less reported periods of nonenrollment, was 7.5 years in 2003, and has actually been slowly rising since the late 1970s, when it was nearly a year shorter. The median age of a doctoral recipient over all has stayed constant for the last three years, at 33.3 years, down from more than 34 in 1994.

The study, which is drawn from the Survey of Earned Doctorates that is produced annually by the science foundation's Division of Science Resources Statistics and five other federal agencies, also provides detailed breakdowns of time to degree by academic discipline. Below is a table that shows the median total time to degree, the registered time to degree, and the age at the time of the doctorate for recipients in various fields:

The NSF study also examines the time to degree for Ph.D. recipients based on their main source of funds for their graduate education. Those who received research assistantships had the shortest time to degree, followed closely by those with fellowship or dissertation grants and those with teaching assistantships. Not surprisingly, the 28 percent of doctorate recipients who reported using their own funds or other non-institutional forms of support -- including other non-academic jobs -- lagged considerably.