User:El Sandifer/Craft of Research

The below is summarized and paraphrased from The Craft of Research: Second Edition, by Wayne C. Booth, Gregory G. Colomb, and Joseph M. Williams. This is currently among the reigning standards of research manuals across academic fields.

Sourcing and the reliableness of sources is a tremendously complex issue; there's a reason most universities have an entire class on writing the research paper. No policy page can possibly solve your sourcing problems for you. In the end, these are principles - actual guidance will come from the consensus of your fellow editors.

Reliable sources
(From pages 77-78)

Booth, Colomb, and Williams frankly admit that "there is no formula for testing the reliability of a source." Furthermore, they note that a review committee for the Journal of the American Medical Association found that even peer-reviewed papers display a high rate of error. No sourcing guideline can ever be a substitute for careful and reasoned judgment.

That said, Booth, Colomb, and Williams suggest that a reliable source ought meet at least one of the following criteria:

Publication by a reputable press
For scholarly sourcs, this includes most university presses, as well as some commercial presses. An exhaustive list of these presses is impossible to generate; check with the relevant WikiProject for the area if you have any doubts or need guidance.

Other reputable presses include government reports, news articles from the major papers, or, for subjects of local interest, from local papers, etc.

Peer-review
A peer-review process is the single best marker of scholarly reliability. Most books from reputable presses will also meet this criteria, though many edited collections will not. Most good scholarly journals utilize peer review, though there are some exceptions.

Reputable author
If the author of a source is a reputable scholar in his field, we ought take the source seriously. There is no white line test to be employed on this. Credentials and C.V.s can, however, often be checked online, giving a good basis for trying to figure out if the scholar in question is reputable. Again, checking this with the relevant WikiProjects is a good idea.

Currency
It is important to use the sources that are currently important to a field. The degree to which this changes varies by field. In fast-moving scientific fields, current sources are often a few months or years old at best. In the humanities, sources can be current for decades or even, in the case of philosophy, centuries. And, of course, for primary sources on historical topics, currency is eternal.

Note that currency describes a source's importance to the work currently going on in a field, not how recently the source was published. A paper written yesterday by a known crank physicist is not a current source in the field of physics. Once again, this is a judgment that should be made in consultation with other editors in the field.

Online sources
(From pages 83-85)

In general, a print source is preferable to an online source. However, there are some circumtances in which online sources might be necessary or even preferable to traditional print sources. In such cases, the general rules for evaluating sources still apply. What follows are some notes specific to online sources.

Equivalence to print versions
Many journals and newspapers, as well as a number of government agencies repost their content verbatim on the Internet. Such sources are reliable, and, because they can be linked to directly, are in many cases preferable to print versions that require more work to follow up on.

Supplements to print
Many journals and news sources offer expanded versions of articles, datasets and figures that could not be included in print versions for monetary reasons, and supplements such as conversations between authors and readers or responses from other scholars in the field. Such extensions of reliable print sources are also reliable.

Recent publication
Many things make it to the Internet before they make it to libraries or print newspapers. In such cases, it's acceptable to cite the Internet preprints. Be careful that you are citing material that is actually going to see print and not un-reviewed drafts.

No print equivalent
Many journals, news organizations, and other such sources publish some of their material exclusively online, and a number of significant publications exist only online. Such sources are important aspects of their field, and ought not be ignored.