Science advances as a whole because of communication. Scientists communicate their findings to each other in both written form and in presentations at conferences. At a deeper level, colleagues discuss their research in a multitude of informal ways, of course. But, the formal process of peer-reviewed publication is the primary means by which research is disseminated within the “halls of science.”
Outside the walls of the ivory tower, however, the picture is very different. In the glare of TV lights, scientists communicate by press conference if their findings are particularly timely, or popular science websites and magazines may translate published results into a more palatable form. The problem with these mass-market distillations of published results is that they are often presented third-hand.
A broader understanding of the peer-review process in general, but that of scientific publication in particular, would help to demystify the process, to the benefit of everyone. So here, in as clear a fashion as possible, I want to convey the essence of peer-reviewed publication.
Then, following this entry, very soon, I am going to publish the first in a regular feature where I translate the results of a published article for a general audience. I will mostly stick to my areas of lessened-ignorance: engineering, physics and astronomy, and planetary and environmental sciences, but I may stray (at my own risk) into the realms of biology, psychology, or whatever suits my fancy.
There are 5 primary elements of a peer-reviewed publication (no, not Intro, Methods, Results, Discussion, and Conclusion; those section headings are very much of secondary or tertiary importance), each will be presented and explained below. Importantly, though the contents of publications are as various as the sciences themselves, all (that I’ve encountered at least) contain these 5 elements.
But first, a definition; peer-reviewed publications refer to written reports of research that have been approved by a panel of fellow scientists, preferably those deemed experts in the field. Approval does not mean that the reviewers believe the author’s conclusions, rather it merely states that they feel the paper presents sound science.
1 - Results
Peer-reviewed publications must present results of research. Such research could be laboratory or field data gathering, theory development and/or testing, pre-existing data analysis, or new technique development, among many many others. The results are what the paper is truly about.
2 - Interpretation
A machine can generate data or even results, but only humans can interpret them. This is where the training and expertise of the scientist is most important. It is also where the scientist is most often wrong. A good research publication does not merely show the world results, it offers the view of the author who has considerable experience with their own research.
3 - Data
The primary “product” of a publication is the results, but absolutely essential to their credibility and broader acceptance is an honest and thorough presentation of the data (in as raw a form as possible). They are presented to the reader via tables, graphs, maps, etc. Here, the reader must be given enough data so that, if desired, he or she can reach their own conclusions as to the validity of the results and their interpretation.
4 - Reproducibility
Controversial results absolutely require that another team or individual can reproduce the results presented in the publication (preferably with no further communication with the author). This means that an author must fully describe their methodology and provide raw data wherever possible. Even non-controversial research benefits greatly from the requirement of reproducibility, because it allows others not as familiar with the field to enter.
5 - Citations
Very few scientists truly work without the benefits of those whom have come before. Citations, notations somewhere in the text that enable the reader to find the source cited, provide credit to those whose work enabled the results presented in the publication. They are also very useful in providing the detail necessary for the condition of reproducibility, as the author can rely on the published work of others for most of the descriptions of their methodology or data.
In my field, the grad students write all the papers, and the grad students end up peer reviewing them as well in leui of their advisors. Do you find this true in your field as well?
Tom,
The grad students do write the papers, in large part. As far as I’m aware, the two professors I’ve talked explicitly to about reviewing papers do the reviews themselves. Have you done any reviewing for your prof. yet, or do other people in your group do this?
Anthony