Non-historian scholars regularly have a need for dependable accounts of particular moments in the past. How can a non-historian locate and then best use scholarly histories?
Non-historians need to be aware that historians produce history using historiography.
Historiography is an empirical methodology, it deals with the outside world using fallible human apparatus.
Historiography is a discursive methodology, it deals with the close reading of large sets of known to be fallible and incomplete texts. This means that history tends towards complex explanations of complex phenomena which are not definitive. This also means that an idea that a non-historian has of the "actually existing past" should be very carefully revised after having read appropriate histories. Historians work in a high context environment, which means that plucking the eyes out of a work is counter productive.
Historiography has a small horizon for reasonable explanation and interpretation. Historians seek to make bridging claims in areas lacking evidence as small as possible. This means that historians produce detailed works. This also means that historians tend not to produce large theories, as large theories are trivially falsified.
Historians mainly produce results in two modes: monographs and journal articles. Non-historians should be aware that historical journal articles are spread across multiple journals, and topical organisation by journal is not necessarily high. This means that searches need to be broad. Additionally, the tendency of historians to produce specific readings of dense texts sets means that you may find many journal articles that partially deal with your problem. Most historical journals are peer reviewed to the standards of the field, if in doubt check the journal publisher's claims regarding peer review or Ulrich's serials database.
Historians also produce monographs. These tend to be through presses of varying quality, and the sign of quality if both long term (60 year) citation and review in journals. Searching for reviews in journals should be relatively easy. Searching for monographs on the subject should be easy. Most historical monograph titles contain a summary description of the chief topic of the work. Reviews can come out up to five to seven years after publication, but most reviews come out within 3 years of publication.
Before making claims about what historians think, you should locate sources, test their usefulness (peer review, review of monographs). Finally historians regularly write "review articles" for journals that cover changes in major problems in history. These are the key to "what do historians in general believe about x." They're unlikely to provide an anecdote, but relying on history for anecdotes is almost certainly a misuse. Anecdotes are often given without the high context required.
The biggest difference between the way a non-historian and a historian do history is the type of sources they use.
A historian will attempt to use primary sources, meaning sources written by actors or reporters at the time of events. For example, letters, newspapers and other contemporary accounts.
A non-historian will tend to use secondary sources, meaning books and articles written by historians, possibly at a much later date.
For example, let's imagine we are studying the Medici family. A historian will try to use letters and documents from 14th and 15th century Florence, all written in Latin or Italian, of course, to understand the Medici. A non-historian will use a book like "The House of Medici: Its Rise and Fall" by Christopher Hibbert to learn the history.