TL;DR:
Start from a narrow question, search with several wordings, and keep a log of what you tried.
Take the same notes on every paper using one template, then compare papers in a grid.
Synthesis means saying what the papers together suggest, not summarizing each one in turn.
Your team wants to improve how new hires start when everyone works remotely. Someone says, "There must be research on this." There is, but a handful of papers read one after another will leave you with five summaries and no answer. A short, structured literature review gives you an answer you can defend, in a day or two rather than a month.
Start with a question narrow enough to search
"Remote onboarding" is a topic. "What practices in the first month help new remote employees feel able to do their job?" is a question. Write the question, the kind of evidence you will accept (studies with data, reviews, or practitioner reports) and a cut-off, such as publications of the last ten years. Decide this before searching, so the search decides nothing for you.
The workflow
Choose search terms: List the main ideas in your question and several wordings for each: remote, distributed, virtual; onboarding, induction, orientation, new hire. Combine them in different ways in a scholarly search tool or library database.
Keep a search log: Record the date, the tool, the exact terms and how many results you scanned. If someone asks how you found the papers, you can answer.
Screen by title and abstract: Keep papers that match your question and evidence rule. Note why you dropped others in a few words.
Read selectively: Read the abstract, the method and the conclusions closely; skim the rest. Check how large and what kind the group studied was.
Take notes in one template: Fill the same fields for every paper so they can be compared. See below.
Synthesize: Build a grid with papers as rows and themes as columns, then write what the grid shows.
A notes template that makes comparison easy
Reference: authors, year, title, link.
Question: what the paper tried to find out.
Who or what was studied, and how many.
Method: survey, interview, experiment, review.
Main findings in one or two sentences, in your own words.
Limits: what the study cannot tell you.
Relevance: how it bears on your question, from high to low.
Quotes worth keeping, with page numbers.
Worked example: reviewing five papers on remote onboarding
The papers below are invented placeholders, labeled A to E, to show the method. Findings and sizes are made up for illustration; in a real review each cell comes from the paper itself.
Paper | Method | Sample | Finding (example) | Limit |
|---|---|---|---|---|
A | Survey | Single company | New hires with a named buddy reported more confidence | Self-reported |
B | Interviews | Small group | Unclear first-week goals caused confusion | Few people |
C | Review of studies | Many studies | Regular check-ins appear in most successful programs | Mixed quality |
D | Case study | One firm | Short daily video check-ins in week one helped | One site |
E | Survey | Multiple firms | Confidence rose when equipment arrived before day one | Self-reported |
Now compare across the rows. Papers A and D both point to human contact in the first days, B points to clear goals, C supports regular check-ins in general, and E points to practical readiness. Count the support: contact and check-ins appear in A, C and D, which is three of the five papers; goals appear in B alone; equipment in E alone. Count it again column by column to be sure it is 3, 1, 1.
That count shapes the synthesis: "The evidence reviewed most consistently supports early, regular human contact. Clear goals and equipment readiness each have support from one source only." It also tells you the limits: most sources are self-reported and several are single sites, so treat the results as a starting point to test in your own team.
Writing the synthesis
Group by theme, not by paper. Each paragraph should make a point and name the papers that support it, say where they disagree, and state how strong the evidence is. Finish with what is missing: questions the papers do not answer and what you would test locally.
Mistakes to avoid
Searching with only one wording and missing papers that use a different term.
Summarizing each paper in turn and calling it a review.
Treating one small study as settled.
Citing a paper you only read the abstract of for details it does not state.
Ignoring papers that disagree.
Doing it faster
A table is the fastest way to compare, and it follows the same pattern as in comparing prices before a purchase. Before trusting any paper's claim, run through the habits in checking facts before you share them. If you want company while you read a dense methods section, DuckyHelper sees your screen and talks it through when you press fn on a Mac, but the notes should be in your words.
Frequently asked questions
How many papers are enough?
Enough to answer your question with some confidence and to see the same themes repeat. For a work decision, a handful of good sources is often more useful than a long list.
Do I need access to paid journals?
Often a library, a university alumni service or an author's own page offers a free copy. Abstracts alone are not enough to cite findings in detail.
Get started
