Bob Filler
Good science is an intellectually rewarding enterprise, characterized by fun and frustration in search of new knowledge, pursued with commitment, enthusiasm and vigor. It's an exciting, challenging, and stimulating endeavor. Its hallmark is thoroughness in execution and intellectual honesty. Let me elaborate on these points.

Thoroughness in execution requires a clear statement of purpose, a comprehensive survey of what is known on the subject in the literature, and a well-designed plan of attack. These cannot be sloughed over. These must be done first.

In experimental chemistry, for example, which I've been involved in since 1943, we should expect a well crafted procedure, careful observation, and a dedicated unbiased recording of the events and results, written at the time, not later.

I'd like to emphasize this last point, observation. Students don't always realize that it isn't good enough to set up the experiment and take down what is observed. You should be looking very closely. I learned that the hard way.

Many years ago, when I was a graduate student, I discovered something that had nothing to do with the research I was doing, something small. I went to various professors. They said, "Don't waste your time, get on with the show." So I dropped it. Two years later, a young assistant professor at Georgia Tech elaborated on the type of observation I made and became famous. I don't know if he became rich-he is a professor-but he certainly established a name for himself. If you see something out of the ordinary, follow it. Those tangents are sometimes where the interesting science is. Another example:

You've heard of Teflon. That was serendipity, finding the most delightful thing in the most unexpected place. That discovery, absolutely accidental, led to a remarkable advance in polymer technology. Many of you use Nutrasweet. This, of course, is a di-peptide (two amino acids put together) called aspartame. It was discovered in Skokie about fifteen years ago A researcher went home and noticed that his food was sweet. His wife said, "I put nothing in it:" He said, "Oh yes you did." It turned out that he hadn't washed his hands completely when he left the lab. He went back and realized he had discovered a non-caloric synthetic sweetener. His discovery has had a significant impact on nutrition.

Now, about intellectual honesty. Many of you have heard about scientists who falsified results, usually under pressure, for fame or fortune or both. I cannot over-emphasize this particular point to students. It is the easiest thing in the world to fall into. Yesterday, Ireceived a copy of R & D Magazine. One article reports that of ninety-nine industrial chemists who responded to a survey, about half reported that they were personally acquainted with falsifying data. They didn't say if they were the ones who did it, only that they knew of an industrial laboratory where someone had committed fraud in order to keep profits high-in order to beat out somebody else.

So, I'm concluding with just two simple messages: first, observe carefully and pursue any observation that looks important; and second, keep your intellectual honesty.

Porter Johnson
Good science is novel. New insights are introduced, new discoveries are made, new models are constructed, or new views of the world are created-new paradigms. A great discovery creates an industry. Perhaps a physical industry, such as micro-electronics, which came as a result of the discovery of the transistor. Or, perhaps, just an industry of people who are studying the problem in question. Great research papers catalyze other significant research papers and many people begin to work on the subject as a result of a breakthrough.

Good science is cumulative as well as novel. Good science is reproducible. All important scientific discoveries will be checked somehow, somewhere, someday. If it's not important, you don't have to worry about whether it's checked. But if it's at all important, it's going to be checked. If you make a significant claim in science, someone else is going to check your work.

Good science is open. Secret recipes, secret procedures, secret ideas are held in suspicion by scientists and for good reason, because usually they don't work.
Good science is international. Science thrives on open communication. Science is a model for international communication. Indeed, the scientist Sakharov has played a fundamental role in the liberalization of policies within the Soviet Union. His model was communication among scientists.

Now I want to talk about fraud and deception. Scientists are like children. As individuals, they are easily deceived by hucksters. Because science thrives in an atmosphere of open communication, scientists don't look for secrets, hidden mechanisms, or hidden messages. For example, during the 1970's, many European scientists sought and found significance in "spoon bending" and other tricks by the Israeli psychic Uri Geller. Geller's procedures were debunked by a magician-The Amazing Randi. Scientists themselves were willing to accept what Geller said at face value because they're not used to trick explanations. They can be thrown off by elementary forms of deception.

Scientists do occasionally embellish data or results. They do that with the full understanding that if they are caught, and if they are wrong, their careers are kaput. They know it's wrong, and still they sometimes do it, and it's unforgivable.

Funding patterns in the United States and elsewhere may actually encourage some forms of deception. Big science creates special problems of fraud. But deception and fraud have always been present in science. So. I think it's important for scientists to convey a sense of ethics to students. Research students should learn from the professors what's O.K. and what's not O.K. We teach our students that they should not put their name on a paper or a research report until they're sure that it's correct because it's their name and their reputation that is at stake. They have to defend their work to us and we have to defend our ideas and conclusions to them. Science thrives in this open atmosphere of questioning, with an adversarial relation between the person who is claiming something and the one who seeks justification for the claim.

Ben Stark
Good science could be said to have two components. First is the choice of a problem considered "important," that is, the answers to which are expected to provide particularly useful or fundamental information. Second is the execution of the experiments or theoretical work required to solve the problem. This second component is the one usually associated with poor science, in particular accusations and instances of scientific fraud. It is possible, however, that the first component might also be a source of questionable practices.

Notable recent cases in which execution, or absence of execution, of experimental work has been the source of fraud, have included clear and extreme abuses such as fabrication of data or plagiarism of data already published. Fraud can also take much more subtle forms, such as minor alteration of data. The message to students to counter such behavior is that any alteration of what Mother Nature provides is absolutely prohibited.

Of course, you can be completely ethical and still produce data which are artifactual, too sparse, or too widely scattered to give clear answers to the questions asked. Sometimes experiments are so difficult that such limitations cannot be avoided. There is no ethical problem if these limitations are clearly noted when the data are presented. Whenever possible, however, we should stress to students the importance of trying to overcome these limitations. The necessary practices range from simple ones (repeating experiments a sufficient number of times) to more difficult ones (constantly examining and, if necessary, changing the experimental design so that artifacts can be eliminated and scientific questions answered more directly). An important corollary of these practices is never to presuppose so strongly what the results of an experiment should be that the experimental design or interpretation is even subconsciously skewed against alternative explanations.

Much more difficult is formulating criticisms and recommendations concerning the choice of an area of inquiry to pursue (the first component of good science). As recently as when I was an undergraduate (in the late 60's and early 70'sJ, one of the real advantages of a career in pure science, our professors told us, was that you could work on anything that interested you, no matter how obscure. The rationale was that there was no way to tell beforehand which area of research would prove important later, so all were worthwhile.

That philosophy may still be true in theory, but in the real world of the '90's it is inoperative. The demand for federal support for scientific research has increasingly outstripped the supply. The consequence is fierce competition for grant money from the NSF and NIH. It is now very difficult to get funding for unusual or heretical ideas. The fundable proposals still must have some novel aspects but they cannot be too far from what are considered important mainstream ideas.

While many mainstream ideas are, in fact, fine, this channeling of research can cause scientists, who tend to be parochial anyway, to think in narrow ways. This, in turn, can put great pressure on their students to think the same way. The net result may be to stifle investigation in directions lightly regarded within the mainstream, including those that may ultimately turn out to be important. I have seen a heretical idea put forward by a graduate student and met with an attitude varying from indifference to opposition by his advisor. When his advisor finally accepted the idea and tried to present it to the scientific community, he too met with indifference and opposition. Just a few years later, the idea was hailed as one of the greatest of the last several decades and, ironically, is now part of the mainstream.

What should we teach the students of today about how to do good science? It can be distilled into a few simple axioms: We cannot presuppose the solutions to the mysteries left for us to solve. We cannot presuppose which of these mysteries are the most important to solve. In working towards a solution, we must try to be as objective, honest, and careful as possible. Given pressures today, this may be increasingly difficult, but we have to do it anyway. Mother Nature expects no less.