Boolean Search String Builder

A Boolean search string is a precise description of what you are looking for. Recruiters use them to find candidates on LinkedIn and job boards; students and researchers use them to search databases for a literature review. The logic is the same everywhere, but each site writes it slightly differently, and one stray bracket can silently change the results. This builder handles the syntax so you can concentrate on the words.

Break the question into concepts

Start with the question in plain words — “senior Python developers” or “does exercise reduce depression in teenagers?” — and pull out its separate ideas. Each idea becomes a concept box: role, skill and seniority for the first; exercise, depression and adolescents for the second. Two to four concepts is typical. More makes the search too narrow.

List the synonyms

Inside each box, list every way people write that idea, one per line: “software engineer”, “developer”, “programmer”. Phrases are put in quotation marks automatically. Where a database allows it, use an asterisk for word endings: adolescen* finds adolescent, adolescents and adolescence. LinkedIn and Google do not support it, so the asterisk is dropped there and the word forms have to be listed instead.

How the parts combine

Synonyms within a box are joined with OR and wrapped in brackets; boxes are joined with AND. Words in the exclude box are removed with NOT, or with a minus sign on Google. Use exclusions sparingly: NOT student also removes profiles that mention mentoring students.

Choosing the platform

Choose the site before copying. LinkedIn and EBSCO use plain AND, OR and NOT in capitals. Google needs no AND and excludes with a minus sign. PubMed adds a field tag such as [tiab] to every term. Scopus wraps each group in TITLE-ABS-KEY(…) and uses AND NOT, and Web of Science uses TS=(…) for topic searches.

Checking a string

The checker reads either the string you built or any string you paste. It reports unclosed phrases and brackets, operators in lower case (searched as ordinary words on most sites), curly quotes copied from a word processor, operators with nothing on one side, and platform limits such as Google’s 32-word cap.

Worked examples

The recruiting example produces ("software engineer" OR developer OR programmer) AND (Python OR Django) AND (senior OR lead OR principal) NOT (intern OR student). The literature example in PubMed becomes (exercise[tiab] OR "physical activity"[tiab] OR sport*[tiab]) AND (depression[tiab] OR "depressive symptoms"[tiab]) AND (adolescen*[tiab] OR teen*[tiab] OR youth[tiab]). Pasting (nurse OR "registered nurse" AND icu into the checker flags the bracket that is never closed.

Recording your search

For a systematic review, save the final string, the database and the date you ran it; reviewers and the PRISMA checklist expect the full search to be reported, and it lets you rerun it later.

Testing and refining

A good search is built in rounds. Run it, read the first twenty results and ask two questions: are relevant items missing, and are irrelevant ones crowding in? Missing items point to a synonym you have not listed; irrelevant ones point to a word with a second meaning that needs a more specific phrase or a careful exclusion. Change one thing at a time so you can see its effect.

Related tools

Build web searches with operators like site: and filetype: in the Google Search Operator Builder, organise what you find in the Literature Matrix Builder, format sources with the Bibliography Formatter, or pull links from results with Extract URLs.