Quick answer: Scopus supports two proximity operators the basic search box never advertises — `W/n` (within n words, any order) and `PRE/n` (precedes by n words, fixed order). They sit between the bluntness of `AND` and the rigidity of an exact phrase, and they can collapse thousands of irrelevant results into a few hundred relevant ones.
Here’s the problem they solve. Search `hydrogen AND subsidy` in Scopus and you get everything where both words appear anywhere in the record — often thousands of results, most of them noise. `AND` “works,” it just works badly.
How to use them
- Open Advanced Search.
- Scope your terms with a field code — `TITLE-ABS-KEY` keeps the search to titles, abstracts, and keywords: `TITLE-ABS-KEY(climate W/5 migration)`.
- That returns only records where the two terms appear within five words of each other.
- Use `PRE/n` when word order matters — `PRE/3` requires the first term to precede the second by no more than three words.
- Tune `n`: start at 5, tighten to 2–3 if results are still noisy.
A real example. A policy analyst replaced `hydrogen AND subsidy` (4,800 results) with `hydrogen W/4 (subsid* OR incentiv*)` — and got 312 relevant records. A 93% cut in noise, from one operator.
Proximity searching isn’t unique to Scopus, either: Web of Science uses `NEAR/n`, and it works inside citation reports too.
Try it: take your most-used saved search, replace one `AND` with a proximity operator, and compare the result counts. See how much noise you just eliminated.
Want the full profile of what Scopus does well and badly? It’s in the Database Directory. And every week we break down one technique like this — subscribe to get the next one.
[Subscribe CTA]