It contains the advice or opinions of one or more Wikipedia contributors. This page is not an encyclopedia article, nor is it one of Wikipedia's policies or guidelines, as it has not been thoroughly vetted by the community. Some essays represent widespread norms; others only represent minority viewpoints.
This page in a nutshell: Both research articles and review articles are permissible as sources for Wikipedia articles, but they are useful in different ways. Original research articles do not share many of the same issues that purely primary sources raise, but the choice of which research articles to include should be guided by how they are cited in secondary sources to ensure their notability.
I've encountered a lot of discussion on the proper role of original research articles in scientific articles, usually in relation to their identification as primary sources. My view is that both original research articles and secondary sources such as reviews are both permissible, but serve different purposes. Research articles establish verifiability, but not things like neutral point of view, proper balance, or notability. Review articles are much better at satisfying the latter three, but since they tend be less in-depth than the original literature, they are somewhat less useful for verifying details. It is important to cite both types of literature in order to satisfy both verifiability and NPOV/balance/notability.
Special considerations for scientific articles
The concept of primary and secondary sources originated in the humanities, but does not perfectly apply to works in the sciences. Wikipedia editors often make an absolute identification that original research articles are primary sources while review articles are secondary sources, but I believe the situation is actually considerably more nuanced.
Are research articles primary sources?
According to WP:PSTS, "primary sources are very close to an event, often accounts written by people who are directly involved, offering an insider's view of an event, a period of history, a work of art, a political decision, and so on."
Original research articles contain the results of an experiment the authors have performed, along with a review of previous literature relevant to the experiment, justification for their experimental design, and analysis and interpretation of the results contained in the paper. While the description of the results is definitely a primary source, the discussion and interpretation of the results arguably has some of the character of a secondary source. The literature review in the introduction of most journal articles is unmistakably a secondary source.
Most scientific articles are peer-reviewed by independent scientists and published by a third-party publisher. People often confuse primary sources with self-published sources, and many of the perceived restrictions on primary sources are actually guarding against using self-published sources such as personal blogs and vanity presses, which are considered to be unreliable. Wikipedia:Verifiability specifically states that peer-reviewed journal articles are to be considered reliable sources.
Are review articles secondary sources?
Also according to WP:PSTS, "secondary sources are second-hand accounts, generally at least one step removed from an event. They rely on primary sources for their material, making analytic or evaluative claims about them."
Review articles contain no original data, but summarize the results of a large number of other research papers, placing them into a larger context and often drawing comparisons between them and discussing the big-picture direction of the field in question. This would seem to easily classify these as secondary sources.
However, review articles are almost always written by the same researchers who publish original research articles in the field, and authors of review articles often include their own research articles in their reviews. If the authors are not uninvolved in the field, the same balance and NPOV issues are present as with the original research articles.
This last point is obviated with scientific third-party news reports, such as "Perspectives" and "News and Views" pieces which appear in scientific journals like Science and Nature, as well as articles in newsmagazines published by scientific societies such as Chemical & Engineering News. As a matter of current practice, these tend to be rarely cited on Wikipedia, but I find these to be invaluable for neutrally establishing the context and impact of an original research article. They can also be expected to contain accurate science as well, which is not always the case with popular media news reporting.
Proper uses of primary and secondary sources
Primary sources establish verifiability but not NPOV, balance, or notability.
Secondary sources establish neutral point of view, proper balance, and notability, but are less good at satisfying verifiability and at providing attribution for the original source of the idea or discovery.
The choice of which primary sources to include should be guided by how they are cited in secondary sources. Most Wikipedia articles on current topics of research are written by someone with expertise in that field, and they will often have a sense of which are the most notable papers in that field. While this expert sense of notability is extremely useful in writing an article, there is the possibility that this sense is subtly biased by the expert's own experiences and professional associations. On Wikipedia, this unintentional bias is mitigated by having multiple editors reach consensus, but some topics of current research are new or esoteric enough that there might be very few, or even only one, editor who is expert enough in that field to make that call. Therefore using the secondary sources as a guide to which primary sources are notable is often the only way to keep unintentional bias from creeping into the article.
Secondary sources are, however, somewhat problematic from an attribution standpoint, as the original source of the idea or discovery is not being cited. This can be provided by in-text attribution, but it is often more concise and more convenient to the reader to provide the full citation to the original primary source. In addition, secondary sources are often less good at satisfying verifiability because they often lack the same detail as the primary sources.
The best method is to write the article from secondary sources, but also include citations of research articles which are mentioned in those secondary sources. Sections covering things like background, history, and applications should only use secondary sources; more technical sections should ideally have each statement cited by both a secondary and a primary source.
Material such as an article, book, monograph, or research paper that has been vetted by the scholarly community is regarded as reliable. If the material has been published in reputable peer-reviewed sources or by well-regarded academic presses, generally it has been at least preliminarily vetted by one or more other scholars.
It is also important, however, for our articles to clearly indicate the person who first discovered an astronomical object, first proved a theorem, first performed an experiment, or was otherwise responsible for the idea being discussed.... Articles should provide attribution for experiments, theorems, astronomical objects, and similar topics, when the original discoverer is known. Many editors prefer to supply the original source for an idea when providing this attribution.
Secondary or tertiary sources are needed to establish the topic's notability and to avoid novel interpretations of primary sources, though primary sources are permitted if used carefully.... A scientific paper documenting a new experiment is a primary source on the outcome of that experiment.... Unless restricted by another policy, primary sources that have been reliably published may be used in Wikipedia, but only with care, because it is easy to misuse them. Any interpretation of primary source material requires a reliable secondary source for that interpretation. A primary source may only be used on Wikipedia to make straightforward, descriptive statements that... are [verifiably] supported by the source.
