To keep a research journal, make a dated record of what you worked on, what you noticed, why you made a decision and what you need to do next. Link each entry to the relevant paper, file or result so you can return to the evidence. Start with a few useful sentences after a research session, then review the entries regularly.
A PhD or dissertation rarely develops in a straight line. You revisit a definition, abandon a promising approach or realise that two papers mean different things by the same term. A journal gives those changes somewhere to live, instead of leaving you to reconstruct them months later.
This guide includes a copyable research journal template and a fictional worked example. You can use the structure in an approved digital notebook, a document or a paper notebook, adapting it to your discipline and institution.
What belongs in a research journal?
A research journal records the development of your thinking and work across a project. It can include questions, observations, interpretations, decisions, problems and reflections. Here, “journal” means a working research diary, rather than a scholarly publication.
The University of St Andrews recommends keeping a research journal during postgraduate research as a way to track development and keep writing throughout the degree. You can begin while you are still exploring the literature; you do not need a finished argument first.
Give different records clear jobs:
| Record | Main purpose | Example |
|---|---|---|
| Research-paper notes | Understand and retrieve a particular source. | What a study found and how it measured the outcome. |
| Research journal | Track how your project and thinking develop. | Why that measurement approach may not fit your question. |
| Task list or planner | Organise future work. | Compare two alternative measures before Friday. |
| Required research records | Document the work to the standards of your project. | Approved protocols, original data, analysis scripts and experiment records. |
These records can connect without repeating every detail. Keep your notes on individual research papers with their citations, then use the journal to explain what those sources change for your project.
Set up a journal you can retrieve and maintain
Choose one main location
Use a location permitted for your research. A digital document makes searching and linking straightforward; a paper notebook may suit sketches and handwritten thinking. A dedicated electronic research notebook may be appropriate when your team needs controlled sharing or a detailed edit history.
Cambridge’s guide to electronic research notebooks recommends considering access controls, edit history, export options and backups. Check these before committing important work to a tool. An attractive interface is less useful if you cannot retrieve your records when you leave the institution.
Give entries dates and descriptive headings
“2026-10-22: Defining feedback uptake” is easier to find than “Thursday thoughts”. Add a small number of consistent topic labels if they help, such as “methods”, “literature” or “supervision”. Avoid spending more time maintaining the labels than recording the work.
Link to evidence instead of copying everything
For a paper, record its citation or a stable reference-manager link. For analysis, identify the dataset version, script version and output location. For a paper notebook, use page numbers and a short index. Your research-paper organisation system should remain the home for the sources themselves.
A useful entry lets you answer: “What was I looking at when I wrote this?” A filename such as “results-final” may not be enough if several files share it.
A copyable research journal template
Copy the prompts below into your chosen system. They are an original suggested structure, rather than a required institutional form. Complete the fields that help with the session; expand them when a consequential decision needs more explanation.
Date and entry title: [When was this written, and what is it about?]
Question or purpose: [What was I trying to understand or accomplish?]
Work and evidence: [What did I examine or do? Which sources, files, versions or outputs support this account?]
Observation: [What did I actually notice?]
Interpretation and uncertainty: [What might it mean? What else could explain it? What remains unchecked?]
Decision and reason: [What am I choosing or proposing, why, and what would make me reconsider?]
Next action: [What specific action follows?]
Question for discussion: [What needs supervisor or collaborator input?]
If that feels too elaborate, begin with three lines: What changed? Where is the evidence? What happens next? Add detail where the shorter record leaves you unable to understand a decision later.
Worked example: from reading to a research decision
The following entry is fictional. The project, documents, observations and file references are invented to demonstrate the format. They are not findings from a real study or references to published papers.
| Field | Example entry |
|---|---|
| Date and title | 22 October 2026: What counts as feedback uptake? |
| Question | Can I use one definition of “uptake” when comparing the studies in my dissertation? |
| Work and evidence | Compared the definitions in practice documents A, B and C. Recorded exact passages and page references in comparison-table-v02, stored in the project’s methods folder. |
| Observation | A counts whether students open the feedback. B counts revisions to a later draft. C asks students whether they found the feedback useful. |
| Interpretation and uncertainty | These appear to describe access, action and perceived usefulness. I have not checked whether the full methods sections introduce additional measures. |
| Decision and reason | Provisionally keep the three measures separate. Combining them now could hide an important difference. Reconsider after checking the full methods and discussing the comparison. |
| Next action | Check each methods section and add the actual measure, timing and limitations to the comparison table. |
| Question for discussion | Does the dissertation question require a narrower definition, or should it explicitly compare these different meanings? |
The useful part is the chain connecting evidence, interpretation and action. “Read three papers; definitions are confusing” would record an activity, but give your future self very little to work with.
Notice that the entry preserves uncertainty. It does not turn an initial impression into an established finding, or describe a proposed choice as a supervisor-approved decision.
Adapt the entry to the work you are doing
After reading
Record a connection, disagreement or change in your question. You might write: “These authors use the same term differently; compare their operational definitions before treating the results as comparable.” Keep detailed summaries in the source notes.
After analysis or an unsuccessful attempt
Identify what you tried, which inputs and settings you used, what happened and what you have ruled out. “The output was empty” is an observation. “The hypothesis is wrong” is a much larger claim that the observation may not support. Link to the actual error, output or analysis record.
After supervision
Record the decision, its reason and any remaining disagreement or dependency. Distinguish your recollection from an agreed meeting record. If an important point is unclear, seek clarification rather than quietly treating your interpretation as settled.
During reflective or qualitative work
You may also need to consider how your assumptions, position or interactions shape your interpretation. Describe a specific instance and its possible effect, rather than adding a general statement that you might be biased. Follow the methodological and ethical requirements of your project.
Turn entries into useful questions and reports
Choose a review interval that fits your working pattern. Weekly is a reasonable starting point; the purpose is to use the record, rather than accumulate pages you never revisit.
- Find the changes. What do you understand differently from the previous review?
- Check open questions. Which can you answer yourself, and which need discussion?
- Extract decisions. Preserve their reasons and links to supporting work.
- Choose the next actions. Transfer current commitments to your planner.
- Carry forward only what remains relevant. Mark a question resolved, deferred or outside scope, with a reason.
Use unresolved issues to build your PhD supervisor meeting agenda. Over a longer period, dated entries and linked outputs can help you prepare a PhD progress report. Check the underlying evidence before using an early journal interpretation in a formal report.
Keep the record trustworthy and appropriate to share
The University of Iowa’s research-records guidance includes notebooks and logs among the materials documenting research activities. Treat a journal containing research information accordingly: follow your institution’s storage, access and retention requirements.
Avoid placing identifiable participant information, confidential material or restricted data into an unapproved personal app. Use permitted identifiers and links where appropriate. If you include personal reflection, consider who may legitimately access the record before sharing the whole journal.
Make later corrections visible and date additions written retrospectively. Do not silently rewrite an earlier entry to make it look as though you knew the outcome at the time. A journal supports your required records; it does not automatically replace a laboratory notebook, approved protocol, formal supervision log or data-management plan.
Common questions
Do I have to write every day?
No universal daily rule applies to this suggested template. Write when there is something worth preserving, such as a decision, a useful connection or a problem you may encounter again. Follow any separate record-keeping requirements set by your institution or research group.
How long should an entry be?
Long enough to reconstruct what happened and why it matters. A routine session may need a few sentences; a major methodological decision may need a fuller explanation and linked evidence. Word count is a poor measure of usefulness.
Can I start halfway through my PhD?
Yes. Write a clearly dated starting entry covering the current question, project status, evidence locations and open decisions. Label any reconstruction of earlier events as retrospective. Begin recording current work without trying to recreate every missing day.
What should I write when nothing worked?
Describe the attempt precisely, link to the record and state what remains uncertain. Finish with the smallest useful next check. An unsuccessful attempt can still be worth recording if it prevents repetition or helps someone diagnose the problem.
For your next research session, start with one entry and a descriptive heading. Record the question, point to the evidence and leave yourself a clear next action. Refine the system after you have used it.
Featured photograph: Nick Morrison on Unsplash.