"What is the difference between Dr. X and God?" goes an old joke about a prominent scientist. The answer: "God is everywhere, while Dr. X is everywhere except in his lab."

Such is the image of scientists who run highly successful research labs. They face so many urgent demands on their time, and have so many skilled postdocs and graduate students in their labs, that they find it counterproductive to actually work at the bench. To learn whether there is more than a germ of truth in this impression, we ran a brief survey of HHMI's 329 scientists to see how some of the nation's leading biomedical researchers allot their time—their most precious commodity. We received answers from 236 of them, or 72 percent.

A surprisingly large number (129, or 55 percent) of those who replied said they spent no time at all on Item 1, "doing experiments with your own hands in the lab" (which is expected to be the dominant activity of scientists at Janelia Farm). Several of those who answered "none" to this question took the trouble to express the frustration this caused them. "My answers are a sad commentary on what happens to the time of successful scientists," wrote one. "None—except when I am on leave," wrote another. A third added, "None. Are you kidding? I wish!!!"

Some of those who are mathematicians or structural biologists explained that for them, "doing experiments with your own hands in the lab" meant doing computational experiments or solving 3-D structures.

Another 58 scientists reported spending between 1 and 5 hours a week doing experiments with their own hands in the lab. Only a handful managed to carve out more time for such work. One scientist who wrote that he spent between 20 and 30 hours a week at it commented, correctly, "I'll bet my reply is different from average."

Regardless of how much time they worked at the bench, most of the scientists said they spent roughly 10 to 20 or 20 to 30 hours a week on Item 2, "talking to people who are doing experiments in your lab, or looking over their data." One researcher noted that much of this time was actually spent "talking to people about things that are not directly related to experiments. There is a lot of informal teaching, career advice, psychotherapy, and discussing things related to getting jobs and grants (like writing letters of recommendation, helping to prepare talks and proposals, talking about what constitutes the right job, etcetera)"—activities that would naturally take more time in a larger lab.

The remainder of the workweek was split almost evenly. The median scientist spent roughly 16 hours on Item 3, "writing grants applications, writing/editing papers, or thinking about results and future research plans directly related to experiments in your lab," as well as on Item 4, "teaching; writing review articles; attending scientific conferences, faculty or committee meetings; reviewing grants or papers; consulting for companies; or doing other scientific or professional tasks not directly related to experiments in your lab." However, some of the researchers spent an astounding 50 percent of their time on this final group of activities.

Several researchers questioned the accuracy of such a survey. "Since schedules vary widely from month to month," respondents' answers must necessarily be guesses, wrote one. He explained that "in the part of the year when I am teaching, or writing a long review, these activities will take up nearly all my time," but during other periods the allotment is quite different.

Finally, one scientist questioned whether such measurements made sense, or even had to: "Scientific research is not a job that one measures in hours. I am not sure when I am working during the day and when I am not. The good news is that the time doesn't seem to matter."

—Maya Pines

Illustration: Mick Wiggins

Download this story in Acrobat PDF format.
(requires Acrobat Reader)

Reprinted from the HHMI Bulletin,
June 2003, pages 24-28.
©2003 Howard Hughes Medical Institute