How to Search Google Scholar and Library Databases

# How to Search Google Scholar and Library Databases Properly

A first independent study lives or dies on the quality of its source list. Supervisors rarely ask whether you found enough sources. They ask how you found them, and whether the same search would produce the same list again. That is what separates a proper search from a lucky one. Three levers do most of the work: Boolean operators, database filters, and citation chaining. Google Scholar and subscription databases also do different jobs, and the choice between them is itself a methodological decision. Search systems do not perform equally, and the differences are measurable [S1][S4].

Start with a question, not a search box

A search string is a translation of a research question. Narrow the question before you open a database. Students who type their whole topic into a search box are asking the database to guess which part of the topic matters. Do the narrowing on paper first.

Break the topic into two or three concept blocks

A concept block is one noun or noun phrase that a relevant source must contain. A topic like "the effect of mobile money on smallholder maize farmers in Uganda" has three blocks: mobile money, smallholder farmers, and Uganda. Three blocks is usually right. One block returns too much. Five returns nothing.

List synonyms and spelling variants for each block

Inside each block, list every word an author might have used. For the mobile money block that could be mobile money, mobile banking, mobile payments, digital financial services, and mobile wallet. Include British and American spellings. Databases index the spelling the author used, not the one you prefer. For Ugandan topics, include local terminology alongside the international term. The same practice is often named differently in national policy documents and in international journals. A Ugandan ministry report may write "digital financial services" where a journal article writes "mobile money". A district-level study may name a specific provider rather than the general category.

Decide what would make a source irrelevant

Before searching, write down your exclusion criteria: a year range, a country, a study design, a population. This list becomes your filters later, and it stops you from making ad hoc decisions in the middle of a result set. A source you exclude on a stated criterion is a defensible exclusion. A source you exclude because it looked dull is not.

Boolean operators: the three you will actually use

Boolean operators are the grammar of a search string. Almost every database and Google Scholar accepts the same three, and the concept blocks from the previous section are the building material.

AND narrows

AND tells the database that both terms must appear. Use it between concept blocks, never inside one. Every additional AND removes results, so each one should be doing real work.

OR widens

OR tells the database that either term is acceptable. Use it between synonyms inside a single block. Bracket the group so the database reads it as a unit: (mobile money OR mobile banking OR mobile payments). Without brackets, the database may apply the OR to the wrong pair of terms and return nonsense.

NOT removes

NOT excludes a term entirely. Use it sparingly, and only after you have seen the term polluting your results. A NOT applied too early can silently delete the most relevant paper in the set, because the excluded word appears somewhere in its record.

Phrase searching and truncation

Quotation marks fix a phrase: "mobile money" will not return records where the two words appear apart. An asterisk truncates a word stem: farm* returns farmer, farmers, farming, farmed. The exact symbols vary by platform. Check the database's own help page rather than assuming that what worked in one place will work in another.

A finished string for the example topic might look like this:

("mobile money" OR "mobile banking" OR "mobile payments")
AND (smallholder* OR "small-scale farm*" OR farmer*)
AND (Uganda OR "East Africa")

Filters: turning hundreds of results into a usable list

Filters are the second half of the search, applied after the string has run. Each filter is a decision you will have to justify, so set each one deliberately rather than conveniently.

The filters worth setting first

Publication date, peer review, language, document type, and subject area cover most needs. Date is usually the first to set, because it is the easiest to defend: a study of mobile money published before mobile money existed cannot help you. Peer review is the second, if your supervisor expects only peer-reviewed work. Language matters if you can read only one or two.

Filters that quietly lose good sources

Two filters cause most of the damage. Full-text-only hides records whose full text sits behind a different subscription or in print. "Available in my library" hides everything you could obtain through an interlibrary request or by writing to the author. Use both only when you are short of time and have said so in your methods.

Record what you filtered

Keep the filter settings next to the search string, with the date you ran it. A search you cannot rerun is not a method; it is an anecdote. This record becomes your methods paragraph later.

Google Scholar: what it is good at, and what it hides

Set your expectations before you start clicking. Scholar is a different kind of tool from a subject database, and the comparison studies are not flattering on every measure.

Strength: breadth and citation links

Scholar reaches across publisher sites and repositories rather than indexing one curated collection, which makes it useful for a first sweep of an unfamiliar topic [S6][S7]. Its citation links are the feature students underuse most.

Weakness: recall and transparency

Comparative studies report uneven recall for Scholar against curated databases [S2][S3][S4], and limited reporting of what was actually searched [S2][S4]. The second point matters more than the first. If you cannot say what a search covered, you cannot defend it. Two of those comparisons are biomedical studies, so their recall findings should be read as evidence about biomedical searching and may not transfer to every discipline. The broader evaluation is the safer basis for a general statement [S4].

Practical consequence for a student

Scholar is a reasonable starting point for orientation and a weak sole basis for a literature review [S1][S7]. Use it to map a field and to chain citations. Do not let it be the only place you looked.

A note on access from Uganda

Google Scholar often links to a publisher page that asks for payment. Do not pay for an article before checking three things: whether your university library already subscribes to it, whether the author has deposited a free copy in an institutional or subject repository, or whether the library can request it through interlibrary loan. What your library can obtain depends on its own subscriptions and on any consortium arrangements it belongs to. Those arrangements differ between Ugandan universities, and this guide cannot state what any particular institution holds. Ask your subject librarian directly.

Library databases: what the subscription buys you

Logging in through your university library instead of staying in Scholar trades convenience for three things that matter at write-up stage.

Discipline-specific indexing

Subject databases index controlled terms and structured metadata that a general web index does not [S4]. That structure is why a well-built string behaves predictably in a subject database and unpredictably in a general search engine.

Reproducibility

Saved searches, search history, and exportable result lists let a supervisor check your work [S1]. This is the practical difference between "I searched for sources" and "here is my search."

Access

Subscription content and interlibrary requests are the practical reason to start in the library when your topic is specialised. Which databases you can reach, and whether you can reach them from off campus, depends on your own library's subscriptions and on its remote-access setup. Availability varies by institution, so check what your own library subscribes to rather than assuming from a friend's experience at another university. If a database you need is not available, ask a librarian whether it can be requested or whether an open-access alternative covers the same ground.

Citation chaining: using one good source to find the next ten

Chaining compensates for weak keyword guessing, and it is the step most students skip. It works because authors cite their intellectual neighbours.

Backward chaining

Take a source you know is relevant and read its reference list. Those are the foundations of the field, and an expert selected them rather than a keyword match.

Forward chaining

Use the "cited by" link to find later work that builds on the source. This is where Scholar's citation graph is genuinely useful [S7]. Forward chaining is how you find the 2023 paper that answers the 2015 paper you started with.

Stop when the same names and papers recur

Saturation, not a target number, is the signal to stop. When a new search or a new reference list keeps returning papers you already have, you have covered that part of the field. A count of thirty sources means nothing if the thirty-first would have changed your argument.

A working sequence for one sitting

The steps above fit into one ordered procedure. Question, concept blocks, string, database, filters, chaining, export, repeat.

Run the same string in at least two places

Run it in one subject database and in Scholar, then compare what each returned. The two systems do not produce the same results, which is precisely why both should be tried [S1][S4]. The differences tell you something about your topic's literature, not just about the tools.

Keep a search log

Record the string, the database, the date, the filters, and the number of results. That log becomes your methods paragraph, and it is the single most useful document you will produce during the search phase.

Common mistakes that waste an afternoon

Four failure modes account for most lost time. Searching a full sentence, which no database parses well. Stacking AND between synonyms, which returns zero results and looks like a broken database. Filtering to full text in the first five minutes, which hides the best sources. And stopping at the first page of results, which in most databases is sorted by relevance and therefore by a formula you did not choose.

Where this fits in the rest of your project

The search log is not a private note. It feeds directly into your methodology section, where you will be asked to justify your source selection. From there, the source list becomes a literature review, and the literature review becomes the foundation of your proposal.

For the next steps, see How to Find Credible Academic Sources for a Literature Review and How to Write a Research Proposal Methodology Section: Justify Every Choice. When you start recording what you found, APA 7 Referencing Explained and How to Paraphrase Without Plagiarising cover the mechanics. If you use AI tools alongside a real search, read Using AI in Academic Research Responsibly first — a tool cannot verify a source it has not retrieved.

A note on the evidence behind this guide. The comparison studies cited here (S1–S4, S6, S7) evaluate how search systems perform. They do not document Boolean syntax, filter mechanics, or citation chaining as procedures, and none of them is Uganda-specific. Sections 2, 3, 4, 7, 8 and 9 are therefore written as method and platform guidance rather than as cited findings. Where you need a citation for those steps — for example, in a methodology chapter — use your database's own help documentation, an information-literacy handbook, or a systematic review searching guide, and cite those directly. The same applies to the access notes above: what your library holds and can request is a local fact, and this guide does not state it.

Sources used

• S1 — Asher, Andrew; Duke, Lynda M.; Wilson, Suzanne (2013). "Paths of Discovery: Comparing the Search Effectiveness of EBSCO Discovery Service, Summon, Google Scholar, and Conventional Library Resources." College & Research Libraries. doi:10.5860/crl-374
• S2 — Boeker, Martin; Vach, Werner; Motschall, Edith (2013). "Google Scholar as replacement for systematic literature searches: good relative recall and precision are not enough." BMC Medical Research Methodology. doi:10.1186/1471-2288-13-131
• S3 — Bramer, Wichor Matthijs; Giustini, Dean; Kramer, Bianca M. R.; Anderson, Patricia F (2013). "The comparative recall of Google Scholar versus PubMed in identical searches for biomedical systematic reviews: a review of searches used in systematic reviews." Systematic Reviews. doi:10.1186/2046-4053-2-115
• S4 — Gusenbauer, Michael; Haddaway, Neal (2019). "Which academic search systems are suitable for systematic reviews or meta-analyses? Evaluating retrieval qualities of Google Scholar, PubMed, and 26 other resources." Research Synthesis Methods. doi:10.1002/jrsm.1378
• S6 — Pomerantz, Jeffrey (2006). "Google Scholar and 100 Percent Availability of Information." Information Technology and Libraries. doi:10.6017/ital.v25i2.3331
• S7 — Mikki, Susanne (2009). "Google Scholar compared to Web of Science. A Literature Review." Nordic Journal of Information Literacy in Higher Education. doi:10.15845/noril.v1i1.10

Not cited: S5 (Parlette-Stewart, Babineau & Owen, 2010) and S8 (Grimmelmann, 2009) were supplied but are not cited in this draft. S5 concerns Google Book Search and library futures; S8 concerns the Google Book Search settlement. Neither bears on Boolean searching, database filters, or citation chaining, so no claim in the text was attached to them.

Related reading

• How to Find Credible Academic Sources for a Literature Review
• How to Write a Research Proposal Methodology Section: Justify Every Choice
• APA 7 Referencing Explained: In-Text Citations, Reference Lists, and Common Errors
• How to Paraphrase Without Plagiarising: A Checklist