I keep coming back to a plain point: academic databases are built to gather research that has been screened for scholarly use. In practice, that means they do not act like the open web. They are shaped collections, with records, journal titles, and search tools arranged for study, not for casual browsing.
That word curated matters. It means someone has chosen what goes in, and what stays out. A database may cover journals, books, conference papers, or other source types, but its value comes from selection and description. The records are not just stored there. They are described with details that help a search work well, like author names, titles, subjects, and dates.
The phrase peer-reviewed research also needs care. Peer review is the process where experts in the same field read a paper before publication. They check the work for quality, method, and fit. That does not make every item in a database peer reviewed. It means the database often points to, or includes, material that has already passed that step. Some databases mix peer-reviewed articles with other kinds of content, such as reviews, news items, or book records. So the database is a container, but the contents can vary.
That is the part I think readers need most. An academic database is not the same as a search engine. It is usually more selective. It is also more structured. Search terms can be matched against the text of an article, but also against the record that describes it. This is why a search in a database can feel sharper than a general web search. The system is working with a planned body of scholarly material, not with every page it can find.
There is a second fact worth saying plainly. A database can be useful even when it does not store the full text of everything it lists. Sometimes it gives an abstract, which is a short summary. Sometimes it gives a citation, which is the key information needed to find the item elsewhere. Sometimes it gives the full text. The level of access depends on the database and on the item. That is one reason I do not like loose talk that treats all databases as if they work the same way.
I also want to keep one limit in view. The label “academic database” does not guarantee the same quality, coverage, or search behavior across products. Some are broad and large. Some are narrow and subject based. Some index only certain journals. Some include older material, while others focus on current work. The term sounds tidy, but the reality is not tidy at all. That uncertainty is not a flaw in the idea. It is part of how these tools are made.
When I teach this, I try to keep the distinction simple. A database is a managed collection of records and often full texts. A peer-reviewed journal is a publication where experts review articles before they appear. The database may point to that journal, hold its contents, or index its articles, but it is not the journal itself. That difference matters because it changes what the user is actually searching.
This is also where many people get tripped up. They hear “database” and expect one uniform thing. Instead, they get a mix of abstracting services, full-text collections, and citation indexes. Each one serves a different task. One helps with access. One helps with discovery. One helps with tracing who cited whom. The label alone does not tell the full story.
For Beyond Citation, that is the core of the work. I want the words to stay plain enough that a new user can use them, and exact enough that an experienced user does not feel misled. If a database says it contains peer-reviewed research, I still want to know what that claim covers, what it leaves out, and where the record is thin. That is the useful question, and it is the one that keeps this project honest.
The honest limit here is simple. No single definition can cover every academic database without blur. Some are broad. Some are small. Some are only indexes. So the safest answer is still the clearest one: academic databases are curated collections of scholarly material, and much of that material is peer reviewed, but not all of it is the same.
That is the kind of plain, workable detail I want The Source List to keep in view too: one digital source worth knowing, one search tip, and one honest limitation.