Start where the credible sources live

When you have an essay to write, the instinct is to type your question into a general search engine and grab whatever looks reasonable. That approach works well for finding opening hours or a recipe, but it falls short for academic work. Search engines rank by popularity and engagement, not by scholarly rigour, and much of the best research sits behind paywalls that general crawlers never reach. Your university library subscribes to that material on your behalf, so the database is effectively a key you have already been given.

Library databases are curated collections of journals, conference papers, book chapters, reports and theses. Someone has already checked that these items were published by a reputable press or peer-reviewed journal, which saves you an enormous amount of judging and second-guessing. You still need to evaluate what you find, but you start from a much stronger position.

You will usually have two routes in: a broad discovery service that searches across many collections at once, and subject-specific databases for your discipline. Start broad to see what exists, then move to a specialist database when you need depth. Both are reached through your library's website, and both work best when you sign in first, so that full-text access follows you around.

Turn your question into search terms

The single biggest improvement to your searching comes from how you build the query. Databases do not understand meaning the way a person does; they match the words you type. That makes precision your friend.

• Break your question into concepts. A question about the effect of social media on teenage sleep has three: social media, teenagers, sleep.
• List synonyms for each concept. Teenagers might be adolescents, young people, youth or pupils. Try each one, because authors vary.
• Join synonyms with OR and wrap them in brackets: (teenager* OR adolescent* OR youth). This widens the net for that concept.
• Join different concepts with AND. Combining the concept groups narrows the results to items that cover all of them.
• Use quotation marks for phrases so the database looks for "social media" as a unit rather than two separate words.
• Use an asterisk to catch word endings. Searching adolescen* finds adolescent, adolescents and adolescence in one go. Behaviour becomes behaviour* to catch behavioural too.

Once you have a set of results, look at the subject terms or keywords attached to the most relevant articles. These are the database's own controlled vocabulary, and swapping your everyday words for them often transforms a mediocre search into an excellent one.

Let the filters do the heavy lifting

A well-built search can still return thousands of items. Filters, usually listed down the side of the results page, are how you turn that pile into a shortlist.

• Peer review. Tick this when your tutor expects scholarly sources. It removes editorials, news items and other material that has not been through formal review.
• Publication date. Set a range based on your topic. Fast-moving fields may need the last five years, while historical or theoretical work can go back much further.
• Document type. Choose journal articles for current research, book chapters for overviews, and dissertations when you want detailed local studies.
• Language and geography. Useful when you are focusing on the UK or on a particular policy context.

Apply filters one at a time and note which ones helped. If a filter leaves you with almost nothing, remove it and adjust your search terms instead. Filters are a refinement tool, not a rescue tool.

Read the abstract before you commit

The abstract is the short summary at the top of most journal articles, and it is the most efficient reading you will do all week. In a well-written abstract you can usually spot the aim, the method, the sample and the main finding in under a minute.

Use it as a triage system. Ask yourself three questions: does this address my question, is the method appropriate for the claims being made, and is it recent enough for my purposes? If the answer to any of these is no, move on without guilt. Downloading dozens of PDFs you never open is a common student habit, and it costs you time later.

When you find a genuinely useful article, use it as a hub. Check its reference list for earlier work that shaped the field, and look for later articles that cite it. This technique, sometimes called citation chaining, is one of the fastest ways to map the key debates on a topic.

Keep a record as you go

Research becomes painful when you cannot remember where a quotation came from. Build the habit of recording details the moment you decide an item is useful.

• Save the full record, not just the PDF. You need author, year, title, journal, volume, issue and page numbers for your reference list.
• Note the database and the date you searched. Some tutors ask for this, and it helps you retrace your steps.
• Write one line of summary in your own words. Two sentences now will save you half an hour when you are drafting.
• Use the citation tools in the database or a reference manager, but always check the output against your department's style guide. Machine-generated references routinely contain small errors.

When a search comes up empty

Fewer results than expected is not a failure; it is information. It usually means your terms are too narrow or use vocabulary the database does not recognise. Try removing one concept, swapping in a synonym, or spelling a British term the American way, since many databases index US spellings. If your topic is genuinely new, broaden it slightly and look for related debates instead.

And do ask for help. Subject librarians know their collections intimately and can suggest databases you have never heard of, troubleshoot a stubborn search, or point you towards a relevant reading list. Booking a short session with them early in a project is one of the best investments you can make in your own work.