A third-year doctoral candidate looks back: the first year went on reading “to be thorough”, data collection in year two ran four months over, and writing only began in year three, when it turned out the literature review drafted in year one was out of date. No single week was a disaster. There was simply never a clear milestone that would have shown he was falling behind.
A thesis is the longest project most people ever carry largely alone. The undergraduate approach, pushing hard in the last few weeks, does not scale to several years. You need project management tools, kept as simple as possible.
Plan backwards from the submission date
Start with your official submission deadline under your programme's regulations, then work backwards:
- Subtract post-submission administration: internal checks, corrections, final formatting.
- Subtract time for your supervisor to read the full draft and for you to revise, at least two rounds.
- Subtract time to write the remaining chapters.
- Subtract time for data analysis.
- Subtract time for data collection, plus a buffer.
The result is often a shock: the date you must start collecting data is much earlier than it feels. That is why this exercise belongs in your first year.
Break the thesis into output milestones
A good milestone produces something checkable, not an activity. “Read about feedback in learning” is not a milestone. “A summary table of forty key studies plus a 3,000-word review draft” is.
| Phase | Output milestone | Who sees it |
|---|---|---|
| Orientation | Approved proposal; timeline | Supervisor, panel |
| Foundations | Literature review draft; one-page conceptual framework | Supervisor |
| Preparing data | Piloted instruments; ethics approval | Supervisor, ethics committee |
| Collection | Clean dataset with a codebook | Self-check, supervisor |
| Analysis | Key result tables and figures; results chapter draft | Supervisor, lab group |
| Writing | Full first draft; revised draft | Supervisor, outside reader |
Every milestone needs a date and someone who will see the output. Milestones nobody sees slip quietly.
A weekly research log
The simplest and most effective tool: every Friday, spend fifteen minutes writing five lines:
- Done: concrete outputs this week.
- Not done: and why.
- Stuck: questions to put to whom.
- Next week: the three most important tasks.
- Decisions made: for example, dropping variable X because of Y.
The last line is gold when you write your methods chapter a year or two later: you will not remember why you chose that coding scheme, but the log will. Sending the log to your supervisor before meetings also lets meetings get straight to the point.
Buffers: time only drifts one way
Almost nobody finishes a research phase early. Data collection depends on other people, software breaks, rejected papers need resubmitting. A practical rule:
- Add roughly a third more time to any phase that depends on others: data collection, permissions, approvals.
- Do not overlap two externally dependent phases without a parallel plan.
- Keep a few unassigned weeks at the end. If you do not need them, you submit early.
Write early, write continuously
A common trap is waiting for results before writing. Yet much can be written straight away: the methods chapter while you design the study, the context sections while you collect data. Early writing is also thinking. Putting design decisions into prose exposes vague spots while they can still be fixed.
A steady writing target, such as a few hundred words per working day or a few fixed writing sessions each week, usually beats long bursts of binge writing.
Signs your progress is drifting
- Three weeks in a row your log says “more reading” with no written output.
- You avoid supervision meetings because you have “nothing to show”.
- A milestone has moved twice with no revised plan for those after it.
- You keep expanding the project with exciting new questions.
- You cannot say in two sentences where your thesis stands.
One sign: review your plan. Two or more: have an honest conversation with your supervisor. Supervisors can help most when they hear early.
When you need to cut
Sometimes the original plan is no longer feasible. Cutting is not failure; it is a research decision. Common cuts include fewer sub-studies, a narrower scope of analysis, or moving a secondary question into “future research”. A thesis is judged on the quality of its answer to the question it asks, not on page count or the number of experiments.
Files and data: the most underrated risk
Losing a chapter to a dead drive or a version mix-up happens, and it costs months. Three minimum habits: automatic backups in at least two separate places; file names that start with the date in year-month-day form so they sort correctly; and raw data kept read-only, with all processing done on copies and the steps written down.
Tools: the simpler, the longer they last
A spreadsheet of milestones and dates, a log file, a clearly structured folder and automatic backups are enough for most theses. Many people spend weeks building elaborate systems and abandon them two months later. What matters is the weekly review habit, not the software.
This week, do the backward-planning exercise from your deadline, write down the date by which data collection must start, and begin your five-line log on Friday. Take both to your next supervision meeting.
Câu hỏi thường gặp
How do I avoid missing my thesis deadline?
Plan backwards from the deadline, break the thesis into milestones with concrete outputs that someone reviews, add buffer to phases that depend on others, and check progress every week.
What should a PhD research log include?
Each week, note what you completed, what you did not and why, where you are stuck, your three priorities for next week and any decisions made. The decisions section is invaluable when you later write your methods chapter.
Should I wait for results before writing my thesis?
No. The methods chapter and context sections can be written while you design the study and collect data. Early writing exposes unclear thinking while it can still be fixed.
How much buffer time should I build into a thesis plan?
A practical rule is adding about a third more time to phases that depend on other people, such as data collection or approvals, and keeping a few unassigned weeks at the end.
What should I do if my thesis is behind schedule?
Talk to your supervisor early, replan the remaining milestones and consider narrowing the scope. Cutting part of a project is a legitimate research decision, not a failure.
