Build a search string for an assignment topic

Typing an assignment question into a database often produces an awkward mix of results. The wording is usually designed to tell you what to discuss, not to match the language that authors, indexers and databases use.

A useful search string translates the assignment into a small number of concept groups. It then connects alternative terms within each group with OR, and connects different groups with AND. The first version is only a test. You inspect what it retrieves, diagnose the problem and change one part deliberately.

This guide shows that process for an ordinary academic assignment. It does not turn a course essay into a systematic review, and it does not promise an exhaustive search.

Begin with the assignment’s central relationship

Read the brief before opening a database. Underline the topic, the relationship you are expected to discuss and any boundary that genuinely determines relevance. Do not automatically turn every word into a search term.

For example, consider this fictional assignment question:

Discuss how remote or hybrid study may affect students’ sense of belonging during their first year at university.

The question contains three useful concept groups:

1. the study mode: remote or hybrid study;
2. the experience or outcome: sense of belonging; and
3. the population or setting: first-year university students.

Words such as “discuss”, “how”, “may” and “affect” tell you what kind of reasoning the assignment expects. They are usually poor database terms. Searching for affect may also imply a causal relationship that the available studies do not test.

If the topic is still too broad to state clearly, first turn it into a focused question. A search can reveal useful vocabulary, but it cannot make an unfocused assignment coherent by itself.

Build a concept grid before writing the query

A concept grid keeps the logic visible. Put one idea in each column, then add terms that authors might use for the same or a closely related idea.

Concept group	Starting terms	Why each term belongs
Study mode	remote learning; online learning; hybrid learning	Different labels for study that is wholly or partly away from campus
Belonging	belonging; connectedness; isolation	Belonging and connectedness are close alternatives; isolation may retrieve a related but negatively framed experience
Student group	first-year student*; freshman; undergraduate*	The first two target the stage more directly; the last is broader and needs testing

This grid is a working hypothesis, not a vocabulary dump. A term belongs because it could help retrieve evidence relevant to the assignment. Adding every vaguely associated word usually reduces precision and makes the search harder to understand.

Start with your own language, then improve it by checking:

• terminology used in relevant titles and abstracts;
• subject headings or descriptors shown in the database record;
• wording used in your course readings; and
• spelling or naming differences that matter in the databases you use.

The Open University Library’s database-searching guidance likewise recommends identifying key concepts, listing synonyms and related terms, and then applying search techniques rather than pasting a whole question into the search box.

Use OR within a concept and AND between concepts

The basic pattern is:

(alternative A OR alternative B) AND (alternative C OR alternative D)

OR tells the database that either term can represent the same concept group. AND requires a result to contain something from both groups. Parentheses preserve those groups when the query becomes longer.

For the fictional assignment, a reasonable test query is:

("remote learning" OR "online learning" OR "hybrid learning") AND (belonging OR connectedness OR isolation) AND ("first-year student*" OR freshman OR undergraduate*)

This follows the same concept-block logic described by the University of Missouri Libraries’ search-strategy builder: combine synonyms with OR, then connect different concepts with AND.

Quotation marks ask for a phrase. They can be useful when the words form a stable expression, such as "hybrid learning". They can also be too restrictive when authors vary the wording. If a phrase search produces very few useful records, test the phrase without quotation marks or try a genuine alternative term.

An asterisk is commonly used for truncation. For example, student* may retrieve student and students. The exact symbol and behaviour can differ by database, so check its help page. Avoid shortening a word so aggressively that unrelated words enter the results.

Work through the example one revision at a time

The following log is fictional. It illustrates decisions rather than reporting real database result counts.

Version	Query or change	What the test suggests	Next decision
1	Paste the full assignment question	Function words and the causal-sounding verb make the search brittle	Extract the three concept groups
2	"remote learning" AND belonging AND "first-year students"	Every idea is represented, but each group has only one wording	Add defensible alternatives with OR
3	Full three-group query shown above	The structure is usable, but undergraduate* may admit later-year students and isolation may shift the outcome	Inspect relevant and irrelevant titles, abstracts and subject terms
4	Remove undergraduate*, while retaining first-year alternatives	Population precision may improve, although some useful records might be lost	Compare the character of the results and record the reason for keeping or reversing the change

The point is not that version 4 must be best. The point is that the change has a reason. If relevant studies use undergraduate without specifying the year in the searchable record, removing that term could hide useful evidence. You might instead search broadly and apply the first-year criterion while reading abstracts.

Change one feature at a time whenever practical. If you add three synonyms, remove a concept and apply five filters together, you will not know which action improved or damaged the search.

Read the results as feedback on the query

The first page of results is diagnostic evidence. Scan titles, abstracts and subject terms rather than judging the query only by the number of records.

What you observe	Likely cause	Purposeful response
Many results, mostly about unrelated student populations	A population term is too broad or missing	Restore the population group, or replace a broad term with a more precise alternative
Many results about technical online-learning platforms, not belonging	The outcome concept is weak	Strengthen the belonging group with terms found in relevant records
Very few results, although the topic seems well studied	Too many required concepts or an inflexible phrase	Temporarily remove the least important group or relax one phrase
A few highly relevant records use unfamiliar wording	Your vocabulary is incomplete	Add the new term to the correct concept group and note where it came from
Results are relevant but older than the assignment needs	The logic may be sound, but the date boundary matters	Apply a justified date filter after testing the concepts

The University of Nevada, Reno Libraries’ guidance recommends narrowing an over-broad search with a more specific term and broadening a narrow search by removing a term or adding alternatives. That is more useful than randomly making the query longer.

Do not assume that a large result count means success or that a small one means failure. A search is useful when it retrieves enough relevant material for the assignment and helps you understand the vocabulary of the field.

Treat NOT and filters with caution

NOT removes records containing a term. That can solve an obvious ambiguity, but it can also remove useful studies that mention the excluded word in passing. Before using it, ask whether the unwanted meaning can be controlled more safely by strengthening another concept group.

Filters also encode decisions. A date range may be required by the brief, justified by a policy change, or useful when current evidence is essential. A peer-reviewed filter may fit some assignments but exclude valuable reports, standards or government evidence. A language filter may be necessary for what you can assess, but it still limits the evidence you see.

Record important filters beside the query. Otherwise, the same text search can produce a different evidence set when you return later.

Use database features only when they solve a diagnosed problem

Some databases let you search particular fields, such as title, abstract or subject heading. Field searching can improve precision when a concept must be central to the record. It can also miss relevant work when the database has indexed the idea inconsistently.

A sensible sequence is:

1. test the concept groups in the default keyword search;
2. inspect the language of promising records;
3. identify the specific source of irrelevant results;
4. test a phrase, subject heading or field restriction that addresses that problem; and
5. compare what becomes more relevant and what disappears.

The George Mason University Libraries’ database-search guidance similarly treats keywords, synonyms, Boolean operators and repeated adjustment as parts of an iterative process.

Do not copy complex syntax from one platform into another without checking it. Phrase marks, truncation symbols, proximity operators and field codes can behave differently. Preserve the concept logic while adapting the syntax.

Keep a compact assignment-scale search log

You do not need a systematic-review protocol for an ordinary essay. You do need enough of a record to avoid repeating failed searches and to explain how you found your evidence.

Record:

• the database or library search platform;
• the date searched;
• the exact query that produced the useful set;
• important filters;
• terms added or removed and why;
• a brief note on relevance; and
• the next change you intend to test.

Date	Database	Query version	Change tested	Decision
10 Oct 2026	Example database	Version 3	Removed the broad undergraduate* alternative	Keep removed for the next test; review abstracts for missed first-year terminology

Label this as a fictional template row if you adapt it for teaching or demonstration. In your own work, record what you actually did. Do not invent search dates, result counts or databases after writing the assignment.

Once you begin saving sources, use a source-note system that keeps quotation, paraphrase and commentary separate. The search log explains discovery; the source note records what a specific text says and how you may use it.

Know the boundary between an assignment search and a systematic review

An assignment search aims to locate an appropriate body of evidence for a defined piece of academic work. You should search thoughtfully, document important decisions and test whether the evidence fits your claims. You should not describe the search as exhaustive if you have not followed a method designed to achieve and report comprehensive coverage.

Systematic reviews have a different evidentiary burden. The UCLA Library’s systematic-review guidance calls for reproducible documentation of the exact database and platform, complete search strategy, date, limits and result count, alongside searching designed for comprehensive capture. Those requirements are useful as a contrast, not as a label to borrow for a course essay.

If your assignment explicitly requires a systematic, scoping or other evidence review, follow the named methodology and your programme’s requirements. The compact workflow in this guide is not a substitute.

Decide when the search is good enough to support writing

Stop revising the query when you can answer four questions:

1. Does each required concept have a clear role in the search?
2. Do the results include multiple sources that are genuinely relevant to the assignment?
3. Can you explain why you kept the important alternatives, restrictions and filters?
4. Have you recorded the working query well enough to repeat it?

“Good enough” does not mean that every result is relevant. It means the search is transparent and productive enough for the scope of the task. You may return to it when reading reveals a new term or a gap in the evidence.

Finding a source is also not the same as deciding that it is suitable evidence. For a public web page, first trace its publisher, purpose, original evidence, currency and independent coverage. Then check the fit between the source evidence and the exact claim.

A reusable search-building sequence

If you need feedback on how your literature search supports an argument you have already designed, our academic writing development & revision service can help you review structure, evidence use and clarity while keeping the ideas and authorship yours.