Research › Choosing a topic and searching the literature

Choosing a topic and searching the literature

A narrow, answerable question makes everything else easier. Then a systematic search finds what has already been said, so you can add something new.

Last updated 2 November 2026

From interest to question

"Social media" is a topic. "How do 16 to 18-year-olds in UK sixth forms say social media affects their sleep?" is a question. The second is something you can actually answer. A good research question is:

• Focused: specific enough to answer in your word limit.
• Answerable: you can find evidence or collect data.
• Interesting: to you, to your supervisor and ideally to others in the field.
• Feasible: within your time, access and skills.
• Original enough: not simply repeating what has been done, but adding a new context, group, time or angle.

Narrowing a broad topic

Start with a general interest and narrow it by adding limits.

Talk to your tutor before you commit. They can tell you whether the question is too big, too small or has already been answered.

Building a search strategy

1. Identify the main concepts. For the question above: social media, sleep, adolescents.
2. List alternative words. Social media, social networking, Instagram, TikTok; sleep, sleep quality, insomnia; adolescents, teenagers, young people.
3. Combine with Boolean operators. Use AND to narrow ("sleep AND teenagers"), OR to widen ("teenagers OR adolescents"), and NOT to exclude.
4. Use wildcards and truncation. An asterisk finds word endings, so adolescen* finds adolescent and adolescence.
5. Use quotation marks for exact phrases, such as "sleep quality".
6. Refine. Use database filters for date, peer-reviewed articles and subject.

Where to search

• Your library's search tool and databases. Your university pays for access to journals that are not free on the open web. Ask your subject librarian which databases suit your subject.
• Google Scholar. A useful starting point, especially when connected to your library for full-text access.
• Books and textbooks. Good for background and theory.
• Government and official reports. For UK data, GOV.UK and the Office for National Statistics are reliable starting points.
• Reference lists. Look at the bibliography of a good article and follow it backwards; look at who cited it and follow forwards.

The UCL library guide advises searching widely beyond well-known, established resources, including grey literature (reports and unpublished documents), and starting with subject guides and subject librarians, who can advise on selection.

Keep a search log

Record the date, database, search terms and number of results. If your supervisor asks how you found a source, you can answer. If a search yields little, try different words or ask a librarian.

Reading efficiently

1. Read the abstract, then the conclusion, to decide if it is relevant.
2. Skim the headings and tables.
3. Read the method if you need to judge the quality.
4. Make notes in your own words, with the page number.
5. Write one sentence on how it helps your question.

How many sources do I need?

There is no magic number. Your module handbook may give a guide. Quality matters more than quantity: ten relevant, well-judged sources beat forty irrelevant ones. As a rule, cover the key theory, the main studies on your topic and any directly relevant UK context.

Once you have a pile of sources, move to evaluating sources to decide which to trust. Keep track of them using the format in Harvard and APA referencing.

Spotted a mistake or have a question? Get in touch.