Public understanding of science III: Training graduate students

We tried to catch up with our neighboring countries by rightly building up our S&T capabilities. Programs included increasing the R&D funding and the technical manpower. After a decade, however, established indicators showed no progress in S&T performance. Obviously, the S&T programs missed a basic objective – to improve research.

For instance, in 1991-1996, the DOST budget increased four-fold or from P854 million to P3.4 billion. It doubled to P1.7 billion in 1992. Further, holders of graduate degrees in science and engineering also significant increased. The UP College of Science alone granted 509 advanced degrees from 1985 to 1994, with 133 Ph.Ds. Our S&T capability, however, did not improve. Total yearly output of publications in international journals in 1981-1995 by national R&D institutions remained the same.

Scientists are again calling on the government to allot more money to improve S&T. It’s about time. I hope now programs would give more weight on the proper way of doing research. "You need to know how to do research properly before you can begin to think about commercializing discoveries."

If we hope to achieve our objectives this time around, we can no longer ignore the established norm of adequate peer review and verification of research results. Peer review and verification have more than three centuries of developing tradition. They have served science well. We can apply them in training graduate students. This can focus on thesis work, which is meant to be training in research. Since research is not completed until results are properly published and the best indicator of a good graduate training is publication of the thesis, we should include publication in the graduate program.

Below are some pointers to improve the training of graduate students in science. They are also useful for graduate faculty and those involved in R&D.

Choosing a research problem

Funding agencies usually dictate the areas of research to support and often leaves grant applicants no choice but to work on problems out of their main interest. Output quality would then be below that of the research problem of their choice. Also, grant recipients would lose the chance to sustain their creativity and develop competitive ability. In our present state of S&T, we are yet aiming for a critical mass of competitive scientists.

Another problem is about the type of research, whether to go basic or applied. Poor countries are more on applied research, whereas basic studies are common in developed countries. In applied research, time for benefit is short, predictability is high, payoff is small, and scope of applicability is narrow. But in basic research, time for benefit is long, predictability is low, payoff is big, and scope of applicability is wide. These differences perhaps explain why rich nations get richer, whereas poor countries find it hard to develop.

Many of our researchers often ignore basic knowledge when doing applied research. Although relevant studies are available in the journal literature, they are seldom cited in most grant proposals and published papers. Hence, results from applied research often fail in field applications. Developed nations have shown that the greater are the needs for applied research, the more important become basic studies.

The big payoff and wide scope of applicability of basic research deserve more comment. Great scientific discoveries that changed our ways of life were not intended. This serendipitous nature of basic research has given us wonder drugs, lasers, computers, biotechnology, and many others. Every scientist who contributed to the development of biotechnology, for example, did not have biotechnology in mind. This stresses the point of leaving the choice of problems to scientists. As the Nobel Laureate Joshua Lederberg once said in connection with the development of biotechnology, "It would have been tragic were any industry to have had a veto in deciding what would truly be of greatest industrial consequence."

Literature search

A major cause of poor research performance is the failure to search the journal literature. A common complaint is the lack of journals in the library. No library in a poor country can afford to subscribe to the journal needs of researchers. But references can be accessed through indexes that cover important journals. Recently, the Google Scholar has added a speedy way of accessing literature. Whereas literature review should be exhaustive, literature citation should be selective.

Libraries should therefore subscribe to indexes and limit other subscriptions to important review journals. Indexes vary in number and quality of covered journals. The most widely used indexes in the natural and social sciences are issued by the Institute for Scientific Information (www.isinet.com). Examples are the different editions of Current Contents (CC).

Choosing journals and articles for references will improve the quality of the data gathering and the publication arising from it. An indicator of article quality is the reference list. The integrity and quality of an article depends on the quality of the bibliography added to it. If the reference list of an article is more than 80 percent gray literature, and if you are a researcher, don’t waste time reading the article.

Data gathering

Data gathering is the first main part of research; the second is proper publication. It is important to think of publication as an objective when gathering data. This improves data gathering as one thinks of peer review, references and methods, adequacy of data, controls, replicates, etc.

The best time to develop proper work habit is during the data gathering part of training. Much of the quality and usefulness of results depends on work habit. There is no substitute to observing the professor in his laboratory, or in the field, to learn the scientific habit. Unfortunately for our country, we rarely find a scientist whose only absence in his laboratory is holding class.

One work habit that should be corrected early in graduate training is sloppiness. Errors are caused by careless handling of data. If they leaked through the peer review, errors are propagated and can result in serious damage to the scientific literature. The other cause of errors is fraud. Sloppiness gives rise to unintentional errors; fraud, to intentional errors. Their damaging effects on science are the same. But we are often more concerned about fraud than sloppiness, when sloppiness is much more prevalent than fraud. And sloppiness is easier to correct during graduate training, but the tendency to fraud can be hereditary.

Manuscript and publication

Normally, the first experience of publication comes from the thesis. The master’s thesis or a chapter of the doctoral thesis should therefore be a scientific manuscript, written following a "guide for authors" of a CC journal. A master’s program may require only such publishable manuscript. But a Ph.D degree should require at least a published chapter of the thesis. There is no sense in writing the thesis differently, then rewriting it for publication.

Three guides will be needed in preparing a manuscript: (a) a good book on scientific writing and publishing, (b) a "guide for authors" of a chosen journal, and (c) sample articles from the journal. A good book on scientific writing and publishing will show how to prepare the manuscript – stating the Title, writing the Abstract, Introduction, and Materials and Methods, designing Tables, preparing Illustrations, etc. A "guide for authors" and sample articles will show the details. You should be followed strictly.

Presenting a paper orally

The purpose of oral presentation is to inform the audience and to improve the manuscript before publication. Scientists in the audience, by their questions and comments, can improve the paper. Some are helpful to young researchers, particularly to those serious in their work, which they can see in the manner of presentation. It is then important that one learns the basic rules of presenting a paper orally. These are described in a good book on writing and publishing scientific papers.

Presentation should cover main points only. Tables and figures will show results clearly and briefly. Prepare summary tables and figures selected from those in the manuscript. For example, reducing the number of columns and rows of tables will allow bigger prints. Avoid presenting tables and figures as they appear in the manuscript. Unlike in the written paper, projected images are seen only briefly. Well-prepared and properly sequenced slides will allow smooth presentation, with you talking less, and letting the slides convey the message. Some presenters read everything printed on the slide, forgetting that everyone in the audience can also read.

Mechanisms of quality control

A manuscript usually undergoes a preliminary or informal peer review. Reading of a manuscript by a professor or colleague is an example. Another is presenting a paper in a seminar or scientific meeting as described above.

Then the manuscript goes through the standard device for quality control – the peer review. For adequate peer review of a manuscript, good journals have normally three referees active in research on the same specific subject. We do not have enough such reviewers in the country to adequately review manuscripts, one reason why researchers should publish in international journals.

Once the paper is published, it goes through the second and final stage of quality control – the peer verification. This occurs if the paper appeared in a journal covered in widely used indexes. International peers can then do verification studies by repeating experiments to disprove or confirm published results. This explains why improperly published results have largely failed in our technology generation.

Research proposal

This topic is taken last to show that knowledge of research is needed in making a research proposal. If one knows the research process as outlined above, then preparing a proposal should be easy. Problems may still be met, however, until we have enough scientists to review proposals.

The three elements of a research proposal are (a) the funding agency and peer review, (b) scientific merit, and (c) proponent’s capability. I will only reason for the proponent’s capability. This is most crucial for advancing S&T, and it should be the main concern of funding agencies to avoid wasting funds. Evaluation of research proposals is simplified if publication experience is made the criterion in screening applicants. For example, only those with publications in CC-indexed journals should qualify for research grants as in developed countries.

With this requirement and funds available, we can expect every proponent to produce a scientific paper, which is what we expect from research, and what we need to advance science, so that we can have more useful technologies, and finally move on to real development. Whereas science alone cannot save the Philippines, the Philippines without science cannot be saved. This was also said of Africa.

* * *

The author is a retired professor of marine science at UP Diliman, a former chief of the Southeast Asian Fisheries Development Center in the Philippines, and a former chancellor of UP Visayas. E-mail him at [email protected].