In your OS400 proposal, you are required to make use of the traditional scientific literature (journals, peer-reviewed symposia, scholarly texts), and to keep your use of "internet sources" below 10% of your total citations. But what are "internet sources"? How can you tell if a symposium is peer-reviewed? What is is the difference between a scholarly text and another kind of text? Read on, friend.
In this context, the term "internet source" does not refer to sources that can be found on the internet (essentially all peer-reviewed scientific journals post their articles on the internet, for example), but to sources that have no fixed existence separate from the webpages where they appear. For example, the Maine Lobsterman's Association website has a page that describes the work they have done to test gear that is safer for right whales--this material was written to appear on the website, it appears no where else, and it can be changed at any time. This is an internet source. In contrast, the website sciencedirect.com, as part of its database of the journal Marine Policy, has a webpage that provides the abstract for an article entitled "Reducing effort in the U.S. American lobster (Homarus americanus) fishery to prevent North Atlantic right whale (Eubalaena glacialis) entanglements may support higher profits and long-term sustainability." Even though this journal article is available on the internet, it also appears in paper copies of the journal, it was written and reviewed as part of a larger project of furthering understanding of fisheries management policy, and it is permanantly fixed in its content, that is, the authors can't go back and change any part of it--it is a source that can be found on the internet, not an "internet source." Internet sources should always include a "date accessed" in the citation; this is never necessary for a published source that you happen to have accessed via the internet, because content of the source is fixed at the moment of publication.
Peer review is a process by which scholarly materials (journal articles, books, volumes or special journal issues based on symposia or conferences) are critiqued by the professional peers of the author(s) to determine if they should be published.
The most serious kind of peer review (which is standard for most prestigious journals) is called double blind peer review. In double blind peer review the reviewers of an article don’t know who wrote it, and the authors don’t know who their reviewers were. (In many cases, these people can make educated guesses, but the process discourages people from being blinded to an article’s faults/merits by the fame/obscurity of the author.)
The peer-review process for a journal usually is structured something like this:
In many academic disciplines, the papers people present at conferences are assumed to be works-in-progress. Accounts of such meetings are sometimes published in documents called Proceedings or Transactions (e.g. Proceedings of the Biochemical Society). Sometimes the organizers of a conference will submit papers from the conference to peer review and then publish them in a single volume. And often papers that are presented in draft form at conferences are later published in peer-reviewed journals. In some academic disciplines, the peer-review process is used to determine the papers that will be presented at a conference, in which case everything presented has passed that high standard of authoritativeness. This all means that words like "symposia" or "proceedings" can't tell you one way or the other whether something was peer reviewed. Whenever you are using conference or symposium papers, posters, or proceedings, it is important to find out if and when peer-review was involved in the process. The introductory material to the book or issue will tell you.
Specialist encyclopedias are a powerful tool for the early stages of exploring a literature review, because they can quickly point you to the most important or most comprehensive resources on a topic, as well as allowing you to see the full picture of the topic before narrowing your focus. Your actual topic is by necessity quite narrow, but you will best understand the little sub-community of the scholarly conversation that you are a part of when you can see it in its context. In an ideal world, every research process would (1) begin with an encyclopedia article, (2) use that article's suggestions for narrower resources, and then (3) focus further using the bibliographies of those resources, thus allowing you to assimilate that bigger context.
You can tell at what audience an encyclopedia is aimed by the presence and type of scholarly machinery. By this phrase I mean the tools that allow a reader to go deeper into a topic, to make connections within the broader scholarly conversation. An online Encyclopedia Brittanica article, for example, has been carefully written and fact-checked to be sure that it is authoritative and reliable, and it contains cross references to other articles in the Encyclopedia Brittanica, but it contains no "Works Cited" list or list of suggestions for further reading. It is intended to be the end point of your learning about the topic in question. For the general interest reader or a high school student, this is perfectly adequate. An article in a scholarly encyclopedia, in contrast, is specifically intended to be an entry point into a topic, rather than the final word. It will, therefore, contain citations and, in the best cases, suggestions for further reading that are specifically chosen to be a logical next step as a person dives deeper into a topic.
The resources below are all scholarly specialist encyclopedias in the sense that they each have a subject index and the individual articles all contain cross references and citation lists and/or lists of suggested further reading. The articles are intended to be entry points into a topic, rather than end-point, definitive accounts of that topic.
Practical Handbook of Marine Science by 
Encyclopedia of Ocean Sciences by 
Encyclopedia of Tidepools and Rocky Shores by 
Encyclopedia of Aquaculture by 
The Encyclopedia of Beaches and Coastal Environments by Journals that specialize in review articles can be a very useful place to conduct literature-review work. A review article is a scholarly article that provides an overview and synthesis of a whole area of empirical research, rather than investigating its own particular empirical question (what is often called a “research article” or “original article”). Not all journals with “Review” in their titles limit themselves to review articles; many provide a mix of original and review articles. If a review article exists on your topic or a very similar topic, a lot of the work of your literature review will have been done for you, with all the most important citations listed in one place.
Identifying the most important journals in a discipline is an important technique for finding useful articles. For any given research topic the most important journals will not include the whole of the literature you should consider, but they are an efficient entry point, because they take you to the heart of the scholarly conversation. In general, people send their best work to the highest ranked, most prestigious journals. This doesn’t mean good work can’t be found elsewhere, but the more prestigious the journal, the closer its articles are likely to be to the center of the conversation.
Judging the quality of a journal is not a straightforward thing. Because people tend to send their best work to what they perceive as the most prestigious journals, there is a complicated mixture of popularity, age/venerability, selectivity, association with powerful people/departments, and a whole host of other difficult-to-measure factors involved in that gut-instinct judgment. When someone has been reading in an academic discipline for a long time they develop a taste in journals and they come to have an intuitive understanding of what journals are the best or the most popular places to publish particular kinds of work. You will develop this intuition over time, but in the meantime, there are several heuristics to consider that can allow you to make reasonable guesses. But before we get into those heuristics, a few words about journal prestige and traditional privilege.
Journal prestige and journal quality operate in a feedback loop with each other. If a journal is prestigious, people will publish their best work there, raising the quality of the journal, which further raises its prestige. This means that there is a lot of inertia in journal prestige, and another way to describe inertia in social systems is privilege.
Below I will describe several different methods of guessing about the quality of a journal when you don't otherwise know anything about it. All three of these methods are deeply enmeshed in the traditional power structures of academia, which have historically maintained and continue to maintain the privilege of the already-privileged. While many of the explicit mechanisms that have systematically excluded women, Black, indigenous and other people of color, non-Christians, and others from academia have been dismantled in recent decades, the effects of these changes have been very slow to change the make-up of academic communities, particular at the highest levels. Much of the power in academia--tenured professorships at prestigious universities, the editorial boards of important journals, etc.--remains in the hands of people who belong to historically dominant groups (white people, rich people, men, people who have received traditional Western educations, etc.).
The result of this reality is that using journal prestige as a guide in your literature review will systematically exclude a disproportionate number of voices of the marginalized. This does not mean you should not use prestigious journals as a short-cut in your literature review process, but that you must keep in mind this effect, and think carefully about perspectives that may not be included in the mainstream of the scholarly conversation.
Total citation counts are a common way to judge individual articles. You will make use of these counts if you conduct forward-looking citation tracing. Aggregate statistics for citations at the journal-level are also kept, and these can be a rough guide to the most important journals within a given discipline.
Impact Factor (IF) is the most commonly used citation metric at the level of a journal, employed in Clarivate products like Web of Science. The IF value for a given journal is the annual average number of citations per article for a journal's previous two years of articles. Because it is an average across the whole journal, very popular articles can exist alongside articles that are rarely cited. And articles that become highly cited over long periods of time (such as those from slower-moving disciplines) are disadvantaged by the metric. For these reasons, IF statistics are not comparable across disciplines.
Up-to-date IF statistics are usually behind paywalls, but relatively recent historical data can usually be found via Google, and meaningful changes in these kinds of numbers don't happen very quickly.
The process of producing a journal is expensive: editors, designers, and printers need to be paid, paper copies of the journal are mailed, there is often a physical office that must be rented or owned, staffed with administrative help, stocked with coffee, etc. Journals vary in how they get the money to pay for these expenses.
Traditional journals charge subscription fees for access to the journal (whether for print subscriptions or digital access). In many cases, even this money is not enough to sustain a journal, and it is subsidized by the its publisher or parent organization. In most cases, the contract the author signs with the journal limits the author's ability to disseminate the article themselves.
For many open access journals, the author of an article (or the institution they work for) pays a fee for the article to be published, and that money goes to support the operations of the journal. The articles are then free for anyone to use, and the author is free to disseminate the article widely.
There is nothing inherent in an open access model that makes it lower quality, but it is a newer model of journal operations, and accordingly no open access journals have had the time to establish prestigious reputations. Because of the feedback loop between prestige and quality, OA journals are on average of lower quality than their traditional counterparts.
You can sometimes use the publisher of a journal or its institutional affiliation to make a rough judge of its quality, the way you would use a publisher to judge the quality and intended audience of a book. Important journals are often affiliated with an important professional organization or with a prestigious university. In sociology, the academic discipline I know best, one of the most prestigious journals is The American Sociological Review, a journal put out by the American Sociological Society, which is the national professional organization of sociologists in the U.S. By comparison, the smaller, less important Eastern Sociological Society publishes the less prestigious Sociological Forum. It's not that Sociological Forum is bad in the sense of being less scholarly or less authoritative. Its articles have gone through an equally rigorous peer review process. But because it is less prestigious, people don't send their most groundbreaking, most important work there, or they send it after it has been rejected by a more prestigious journal. The journals thus vary in the importance of the research they contain, not necessarily the quality of the research. This is a particularly important distinction when conducting a literature review.
In most disciplines, the most prestigious journals are the broadest. This makes intuitive sense, since the pool of potential articles is larger for a broader journal. The same article about coal geology published in Science would constitute a greater professional triumph than if it were published in Geology, which is in turn more prestigious than if it were published in The International Journal of Coal Geology. Nature publishes about coal only rarely, so you have to have a very important coal article to get published there. The flip side of this is that the broadest, most prestigious journals only have so much space. There simply isn't room for a lot of important detail in a given scholarly conversation in the broad journals. Once you are sufficiently deep in a given topic, you will find some of your best sources in the most prestigious sub-disciplinary journals for your topic.
Professional scholars use citation tracing as their primary method of exploring a literature. The works-cited list or bibliography of a scholarly article is a highly curated selection of sources that can allow you to arrive quickly at the most important, most useful sources on your topic, avoiding the clutter and noise of database or Google Scholar search results.
The easiest way to trace citations is backward in time, by using the bibliographies of articles or books you already have in hand. It can also be extremely useful to trace citations forward in time, by using data available in one of many citation network databases. This allows you to see the intellectual "descendants" of a given source, instead of just its "ancestors," allowing you to bring your literature review up to the present day. The PDF guide below explains one method of using this technique:
Search techniques are most useful when you know a source exists and you are trying to find a copy of it. When you are exploring a literature and trying to identify its most important sources, other methods are usually better (see all the above sections). That said, it is eventually necessary to conduct searches that are very specific to your topic, to make sure you haven't missed anything highly relevant via those other methods.
There are two tasks that call for searches in research: (1) when you are trying to discover the existence of useful sources and (2) when you are trying to retrieve a copy of a source you already know exists. If you are searching in a database that MMA has full-text access to, these two tasks can be accomplished almost simultaneously--you can learn about the existence of an article and download a PDF all on one website. However, in principle, these two tasks are not very compatible. When you are in discovery mode, you want to be able to discover everything that is important to your research, and not be artificially constrained by the limits of the MMA library budget. And when you are in retrieval mode, you want to use tools that most efficiently get your hands on the source in question, without unnecessary extra information or information about other sources.
For folks conducting introductory-level research projects (such as high-schoolers or people writing research papers for low-level humanities class), they often only require some sources that relate to their topic and meet their instructor's standards for credibility. In contrast, for folks conducting serious original research (that's you!), they require the right sources and an awareness of the complete landscape of sources in their topic area. This is one of the main differences between scholarly research and other kinds. For the person writing a quick course paper, whatever sources we happen to already have on campus are usually good enough. For the serious researcher, you need to be aware of the full literature, even if we don't have immediate access to it all. It is in this latter situation that it is most important to separate the discovery process from the retrieval process.
When you search any large database (such as Google Scholar, JSTOR, Academic Search Complete, etc.) with a relatively simple phrase ([scallop physiology] or [environmental dna sharks]) you will usually get a minimum of hundreds of thousands of results, sorted by some mysterious measure of "relevance." It is a rare situation in which this overwhelming number of unorganized results is useful. There are two primary ways to do better when using a large database: building very specific search terms in the first place or filtering/narrowing your search results after the fact using metadata.
Most scholarly databases allow you to use various logical operators to conduct very precise searches. In most cases there is both an "Advanced Search" page that allows you to construct that search using multiple fields and drop-down menus and functionality that allows you to use nested parentheses and Boolean operators to put more advanced searches directly into the basic search bar. Every site is different in terms of what it allows and how certain search tools work. Paid scholarly databases (such as those you access through the library website, as opposed to Google Scholar) always have instructions for how their search tools function. It is good practice to read these instructions and to play around with how the search functions.
One of the most basic but also most useful of these advance search tools is to use a combination of parentheses and AND, OR, and NOT (in most cases they need to be in all caps to function this way) to create logical structures to your search. This trio of words is often referred to as Boolean operators, after the English mathematician George Boole.
Imagine, for example, you are looking for articles about the sustainability of lobster fisheries, using the search phrase [lobster sustainability] but your results keep getting cluttered with things about Caribbean Spiny Lobsters, and you only care about American Lobsters. For most databases, if you input more than one word, the search tool assumes an AND between the words, thus you would get results for which both words appear. If you instead searched for [lobster OR sustainability] you would get articles about other aspects of lobsters and the sustainability of other industries. You can use parentheses just as you would in algebra to determine the logical order of operations. Thus the search string [(lobster AND sustainability) NOT spiny] would give you all the articles that contain lobster and sustainability, specifically excluding all those that contain the word spiny.
You can perform the same logical structuring of a search using the advanced search tool, which, in most cases, has the additional advantage of allowing you to specify fields to search. The screen shot below is from Academic Search Complete, and shows the same search adjusted so the words lobster and sustainability must appear in the title of the article, instead of anywhere in the complete text.
Using OR is particularly useful when you are trying to ensure you have found every possible treatment of a topic. You can list as many synonyms as you can think of, all joined by OR:
sorcery OR witch* OR magic OR occult* OR divination
This example also uses a search method called truncation, in which you tell the search tool that you would like any and all words that start with a particular string of characters. By searching for [witch*] and [occult*] you get results with the word witch, but also witches, witchcraft, occultists, and occultism.
One of the best features of scholarly databases is that they include a lot of metadata on every source they contain. Metadata is simply information about a given source. If the source is a journal article, the metadata includes the journal in which it was published, its subject keywords, its authors' names, its publication date, etc. It is often useful to start with a broad search, which will yield many more results than you want, and then use metadata filtering tools to experiment with narrowing those results.
These filtering tools are usually in a sidebar on the left of the search results (the are also sometimes in the upper right). Every database will have a slightly different array of filters available, but they usually always include some form of subject heading, source type, publication date, and a variety of other options. Google Scholar is again the exception to these general rules, as they have very limited filtering options. Subject headings are often the most useful mode of filtering, but it is important to pay attention to the details of subject metadata.
In Academic Search Complete, for example, there are two ways to filter your results by subject: "Subject" and "Subject: Thesaurus Term." "Subject" is based on subject headings provided by the sources themselves, which have been scraped from the sources by bots. When you click to expand this list, you tend to get a huge list of subject headings that each have a very small number of sources. This is only barely more useful than having no subject filtering at all. "Subject: Thesaurus Term", however, is based on a controlled vocabulary created by the editors of the database. A human being has taken the vast array of subject descriptions that came from the original sources and grouped them logically, mostly by combining synonyms.
Articles: If you have the citation information for an article, but can't find a copy of it, follow these steps:
If we have the title, go to the journal page within the relevant database and navigate to where the article should be using volume number or year and issue.
Books: If you are looking for a copy of a particular book, follow these steps:
Minerva books are delivered to campus twice a week.