Literature Search Help for UK Dissertations
Quick answer: Literature search help for UK dissertations covers the search stage only — the planned, documented hunt for the studies your review will later discuss — not the writing of the review itself. That distinction matters, because a review can only be as good as the search behind it. If the search was unsystematic, sources were found by chance, or the process was never recorded, the review that follows will inherit those weaknesses no matter how well it is written. A defensible search gives you a set of sources you can justify, a record you can show your supervisor, and a foundation your methodology chapter can describe honestly.
Many students begin by typing a few keywords into a search engine, downloading whatever looks relevant, and calling that a literature search. The problem is not effort; it is that nobody can tell what was missed. A planned search answers that objection in advance: it starts from a clear question, breaks it into concepts, builds search strings, runs them across the right databases, screens the results against stated criteria, and records every step. It is slower at the start and much faster later, because you stop re-searching the same ground and stop discovering, halfway through writing, that a key body of literature was never looked for.
This page walks through that process as it applies to UK dissertations, including systematic and scoping approaches, and explains what help with the search stage looks like. It deliberately stops where the writing stage begins: choosing themes, synthesising findings and drafting the review chapter are a different skill, covered by literature review support, and are not duplicated here.
Start from the question, not from the database
Every search decision flows from the research question, so the question has to be workable before searching starts. Frameworks help to make it workable. In health and intervention studies, PICO — Population, Intervention, Comparison, Outcome — breaks a question into searchable parts. In qualitative research, SPIDER — Sample, Phenomenon of Interest, Design, Evaluation, Research type — does similar work for questions about experiences and views. Neither framework is compulsory, but both force the same useful discipline: naming each concept separately, so that each can be searched separately and then combined.
Take a question about whether peer support improves the wellbeing of first-year nursing students. Its concepts are peer support, wellbeing, and first-year nursing students. Each needs synonyms: peer support might also appear as peer mentoring or buddy schemes; wellbeing as mental health or stress; first-year nursing students as student nurses or pre-registration nurses. Building that list before opening a database stops you missing relevant studies simply because they used different words.
At the same time, set your inclusion and exclusion criteria: which study designs, populations, settings and date ranges count, and whether you will include grey literature such as theses and reports or journal articles only. Supervisors accept different choices — what they do not accept is choices made silently, halfway through, because a search returned too much or too little. Write the criteria down before searching, and change them only deliberately, with a note of why.
Choosing databases and building search strings
No single database covers a field, so a dissertation search normally runs across several. Which ones depends on the subject: health and nursing students commonly use CINAHL, MEDLINE or PubMed and PsycINFO alongside their library catalogue; business and management students use Business Source and ABI/INFORM; education students use ERIC and the British Education Index; social science students use Scopus or Web of Science. Your university library’s subject guide is the quickest route to the right list, and your librarian — a resource students underuse — can often spot a missing database in minutes.
Within each database, the search is built from Boolean logic. OR gathers synonyms for one concept: (peer support OR peer mentoring OR buddy scheme). AND combines concepts: the peer support set AND the wellbeing set AND the student nurse set. NOT excludes a term, but should be used sparingly, since it can silently remove relevant studies that happen to mention the excluded word. Quotation marks keep a phrase together; truncation — peer* — catches peer, peers and peer-reviewed in one stroke; wildcards catch spelling variants, such as wom?n for woman and women. These operators are simple, but using them deliberately, concept by concept, is what separates a search from a keyword guess.
Most databases also have their own controlled vocabulary — subject headings assigned by indexers, such as MeSH terms in MEDLINE or CINAHL headings. Searching only free text misses studies whose authors used different wording; searching only subject headings misses very recent studies not yet indexed. A thorough search uses both. When you move to a second database, the string must be translated, not copied, because each database indexes differently — and that translation step is exactly where shortcuts show up in a methodology chapter.
Running, recording and refining the search
Run the search one database at a time and record it as you go. A search log needs, for each database: the date searched, the exact string used, any limits applied, and the number of results returned. That log lets you reproduce the search later, lets your supervisor spot gaps, and saves you from repeating work weeks later when you cannot remember what you already tried.
Refine before you commit to screening thousands of records. A first run returning far too many results usually means a concept is too broad or a synonym too loose; one returning almost nothing usually means a term is misspelled or the AND combination too strict. Check known relevant papers against your search: if your string does not retrieve them, it is not yet good enough. Two or three rounds of refinement now prevent days of screening later.
Export the results to a reference manager such as Zotero or EndNote, and deduplicate before screening. The same study will appear in several databases, and screening duplicates wastes time and inflates your counts. Most reference managers deduplicate automatically, with a manual check for near-duplicates that differ in punctuation or author initials. Only after deduplication do you have a meaningful total to report.
Screening: from hundreds of records to a defensible set
Screening happens in two stages. First, titles and abstracts are screened against your inclusion and exclusion criteria: clearly irrelevant records go, clearly relevant ones go through, and borderline ones are kept for the next stage rather than rejected on a guess. Second, the full texts of the survivors are read and screened again, more strictly. The reason for two stages is practical — full texts take far longer to obtain and read — and methodological: it keeps the process consistent and auditable.
Two habits make screening defensible: recording a reason for exclusion at the full-text stage (wrong population, wrong design, duplicate, and so on), and having a documented check — a second screener or a supervisor discussion — for borderline decisions. What weakens a methodology chapter is a set of included studies with no visible route from search to selection.
Supplementary searching catches what databases miss: check the reference lists of included studies and relevant reviews — citation chaining, backwards and forwards — and consider grey literature where your field values it. Note its limits honestly: hand-searching a few reference lists is not the same as a second database search.
The PRISMA flow: showing your working
The PRISMA flow diagram is the standard way to report a systematic search, and many dissertation markers now expect something like it even for non-systematic reviews. It is a simple diagram with four stages: records identified through database searching and other sources; records after duplicates removed; records screened, then full texts assessed for eligibility; and studies finally included, with reasons for exclusion at each stage.
Building the diagram is straightforward once the search log and screening records exist — the numbers come directly from them. Without those records it is guesswork, which is why the diagram is the output of the process, not a decoration added at the end. For a narrative review, a short paragraph reporting databases, strings, numbers screened and numbers included serves the same purpose, and is far stronger than a vague statement that “the literature was searched”.
One caution: the diagram reports the search, it does not evaluate the studies. Quality appraisal comes after selection and belongs to the review and methodology chapters; keeping the stages separate prevents the common muddle where screening criteria and quality judgements are mixed together.
Common search pitfalls
The most common pitfall is searching too narrowly: one database, a handful of obvious keywords, and a conclusion that little has been published. Usually a great deal has been published, under synonyms the search never tried. The opposite pitfall is a search so broad that screening becomes unmanageable, followed by silent narrowing — keeping only the first hundred results — that is never reported. Both have the same remedy: a planned, recorded search refined openly.
Other pitfalls are quieter: unjustified date limits can exclude seminal studies, language limits should be stated since they shape what the review can claim, and relying on a general search engine alone produces results that cannot be reproduced — a serious weakness in a dissertation methodology.
What literature search help includes
Help with the search stage is practical and specific. It typically begins with your research question and draft criteria, and works through concept breakdown, synonym and subject-heading lists, and a search string built for your main database and translated for the others. You receive the search log in a form you can paste into an appendix or methodology chapter, the screening approach set out in advance, and a PRISMA-style flow with numbers that trace back to the log — so that everything you report can be checked.
What it does not do is write your review. The themes, the critical comparison of findings, and the argument about the gap are yours to develop from the sources the search finds. That separation protects you: a supervisor who can see a clean, documented search is far more likely to accept the review built on it.
How we would work through your search together
The starting point is a conversation, not a form: your research question as it stands, any criteria your supervisor has set, and the databases your university provides. From that discussion we sketch the concept map — the main ideas, their synonyms, and the subject headings worth testing — and run a trial search in one database, so you can see early whether the net catches the studies you expect. Only once that trial looks right do we translate the string across your remaining databases and log every run. You stay in the loop at the trial stage, because a search you have seen working is one you can explain.
Frequently asked questions
How is literature search help different from literature review help?
Literature search help covers finding and selecting the sources: question frameworks, search strings, databases, screening and the PRISMA flow. Literature review help covers what happens next: organising those sources into themes, synthesising and critically evaluating them, and writing the chapter. This page is about the first stage only, so that the second stage starts from a defensible base.
How many databases should I search?
There is no universal number. Most UK dissertations search three or more subject-appropriate databases plus their library catalogue, rather than relying on one. The right set depends on your subject: a health dissertation and a business dissertation need different databases. Your library’s subject guide, and a quick check with your librarian, usually settles the list.
Do I need a PRISMA diagram for a non-systematic review?
Not necessarily in full, but the transparency behind it is still expected. Even for a narrative review, a short account of databases searched, search terms, numbers screened and numbers included — with exclusion reasons at the full-text stage — gives your methodology chapter credibility that a general statement about searching cannot.
What if my search finds almost nothing?
First, check the search before concluding the topic is unresearched: test synonyms, check subject headings, loosen an over-strict combination, and confirm that known relevant papers are retrieved. If the gap is genuine after that, it is a finding in itself — but it needs to be demonstrated by a search that can be shown, not asserted from a single database query.
Should I include grey literature?
It depends on your question and your supervisor’s expectations. Grey literature — theses, reports, conference papers — can fill gaps where journal publication is slow or partial, but it varies in quality and must be appraised like any other source. Decide in your criteria before searching, and report the decision, rather than adding grey sources ad hoc at the end.
Can you help if I have already started searching?
Yes. An existing search can be audited: its strings checked for missing synonyms and subject headings, its databases checked for gaps, and its screening records brought into a form that supports a PRISMA flow. Starting from a partial search is common, and it is usually quicker to repair than to restart from scratch.
If your dissertation is waiting on a search you can defend — strings, databases, screening and flow all documented — send your research question and we will map the search with you, starting with a trial run you can see and question before the full search begins.