Publishing Quality Research vs. Deluging Scientific Information
It is widely believed in the field of scientific research that too many articles are being published. Many research papers are cranked out mainly to pad their author’s resumés, who focus on quantity, not quality. While no doubt this has always been true, there is some evidence that the problem is getting worse. According to a recent study the growth rate in scientific publications was 2–3% per year in the period 1919-1939, but has now reached a growth rate of 8–9%. The current rate corresponds to more than doubling the number of scientific publications every ten years, which does seem a lot higher than the increase in the number of scientists doing research.
Publications: Is More Better?
If the increasing number of publications were high quality, no one would complain. But this seems not to be the case. About 30% of scientific papers are not cited by other publications, and a fair number of the thinly cited papers are only cited by the papers’ authors. These statistics indicate that there are a lot of papers out there that are not considered worthwhile by other researchers.
How to Encourage Quality Publications
Researchers publish second-rate papers because they think padding a resume will impress the decision makers when it is time for grant tenure or promotion. If the powers that be make it clear that they are more interested in quality than quantity, this would go a long way to cut down on the number of bad publications. It would also encourage scientists to do quality research.
How to Rate the Quality of Publications
Suppose a university tells its researchers that from now on the quality not the quantity of publications will be the primary consideration for promotion. Now, how does it measure quality? Should they use a metric such as Impact Factor? Maybe, though I have mixed feelings about this. The metric chosen should not only give a reasonable measure of quality but be hard to “game” so that a researcher cannot artificially inflate his metric score. I don’t have a definite suggestion as to which metric to use but surely some formula which factors in quality would be better than a system which simply counts the publications.