The Research Project Lifecycle End to End
Fourteen weeks of material assembled into one arc, from first idea to submission. What the subject teaches at each stage, what you must get elsewhere, and where the gates that control your schedule actually sit.
Why an end-to-end page exists
A subject taught week by week teaches stages in isolation. That is a reasonable way to teach and a poor way to plan, because almost everything that goes wrong in a research project is a relationship between two stages rather than a failure inside one — a method chosen before the question settled, data collected before clearance granted, an analysis planned after the instrument was already in the field.
This page reads the whole fourteen weeks as one arc. It is a map rather than a replacement: each stage links to the pages that teach it, and the value here is in the joins, the gates and the honest account of where the material runs out.
- Idea
- Literature
- Question
- Design
- Approval
- Collection
- Analysis
- Write-up
The eight stages, and how well each is served
COVERAGE BY STAGE
| Stage | What the material gives you | Depth |
|---|---|---|
| 1. Idea and topic | Four sources of ideas, screening questions, feasibility tests, fourteen suggested themes, and a narrowing procedure | Good |
| 2. Literature | Search technique, source credibility, reading questions, the gap, and how to structure and write the review | Strong — eighteen pages |
| 3. Question and framework | Question interrogation, the one-sentence rule, conceptual and theoretical frameworks, hypotheses | Good |
| 4. Design and method | Philosophy, methodology, method selection questions, validity and reliability, mixed methods | Good on frameworks, thin on options — ten methods named, none compared |
| 5. Approval | Ethics principles, application structure, consent, confidentiality, data retention, and real failure cases | Strong — seventeen pages |
| 6. Collection | Questionnaires, surveys, question wording, response formats, four interview types, interview conduct | Good on instruments; thin on sampling; observation and focus groups barely present |
| 7. Analysis | Qualitative: steps, coding, a complete worked categorisation, ten questions. Quantitative: screening, coding, tables, graphs | Good qualitative; weak quantitative — no dispersion measure, no inferential procedure |
| 8. Write-up | The proposal, drafting stages, two report structures, results, discussion, conclusions, the abstract | Strong — the best-delivered run in the subject |
Depth ratings are this library's judgement, formed from the objective audit and from what each stream could actually be written from.
The gates: where a research schedule actually breaks
Tasks take the time you give them. Gates take the time they take, and you do not control them. That asymmetry is the single most useful thing a project manager brings to a research project, and the subject does not state it.
THE APPROVAL GATES
| Gate | What is approved | What the material says about duration |
|---|---|---|
| Supervisor agreement on topic | That the study is worth doing and can be supervised | Nothing |
| Proposal or prospectus approval | The problem, the method, the plan, the bibliography | Nothing. One external document records three separate prospectus approvals — chair, committee, graduate director |
| Ethics clearance | Consent, risk, data handling, participant rights | One worked case: five rounds over more than five months. That is one case, not a norm |
| Supervisor sign-off on drafts | That the work is ready to examine | Nothing |
| Final examination | The thesis | Nothing |
The five-month ethics case is one recorded instance from the supplied material and is not a planning figure. What it establishes is that clearance is not a formality.
The joins — where two stages meet and projects fail
These are the relationships the weekly structure cannot show, and each has a page behind it.
Six joins worth checking
The method follows the question
Not preference, not familiarity, not what you already own a licence for. The material is consistent on this across several weeks and it is the join most often reversed in practice.
Decide the analysis before you collect
Your response formats decide what analysis is possible. A five-point satisfaction scale cannot become a schedule variance afterwards. This join is where quantitative projects quietly fail.
You cannot recover a variable you did not measure
Once collection closes, anything you failed to record about your participants is gone. Restrict the sample or measure the characteristic — there is no third option and no retrospective fix.
Clearance precedes contact
The hard gate. Data gathered before clearance cannot be used.
Record how, while you are doing it
The Methods chapter is written from the record of your process, not from your data. If you did not keep the record, you cannot write the chapter — which is precisely the library's central case study.
Results follow the order of the methods
The single most useful sentence in the write-up material, stated once and in passing. If you cannot derive a results order from your methodology chapter, the problem is in the methodology chapter.
The half of the lifecycle nobody plans for
Roughly half of a research project is not doing research. It is recording that you did it — and this is the part the weekly structure hides most effectively, because recording is never a stage.
What to record while you work, not after
- How access was negotiated, with whom, and what was agreed
- How participants were selected, and who declined or dropped out
- Every change to the instrument or the protocol, with the date and the reason
- Bibliographic details at the moment you find a source, not at write-up
- The decisions you rejected and why — this is what a limitations section is written from
- Draft fragments, however rough, dated and kept
- Every promise made to a participant, with a date attached in your schedule
What you must get from elsewhere
Being clear about this is more useful than pretending the material is complete. Four things are needed by most project research studies and are not supplied.
THE FOUR EXTERNAL DEPENDENCIES
| What you need | Why the material cannot give it | When you need it |
|---|---|---|
| Statistics | One rule is given, expressed in a unit never defined. No measure of dispersion, no inferential procedure, no formula but one | Before you collect. This gap can invalidate data already gathered |
| Sampling procedure | Six types are named with usable distinctions and no procedure for executing any of them; no sampling frame appears anywhere | At design |
| Instrument models | No interview guide, observation schedule or coding frame appears in fourteen weeks | At design. Examined theses in the exemplar stream are the nearest substitute |
| Your institution's requirements | Format, length, submission and approval rules are institution-specific and the material is not yours | Immediately, and again before submission |
The first is the urgent one. The others cost you time; that one can cost you the study.
A realistic sequence
The subject gives a seven-stage model and a ten-step model that do not enumerate identically, and neither attaches durations. What follows is this library's practical ordering, built from both plus the one worked schedule in the dataset. It is a shape to adapt, not a template.
Idea to submission
Settle the question before anything else
One sentence. Everything downstream is unanswerable until this is fixed, and every week you spend here saves several later.
Read, and record as you read
The literature review is the longest single task in the one worked schedule, by a wide margin. That matches the subject's own emphasis and it matches practice.
Design, and decide your analysis at the same time
Not afterwards. The response formats you choose decide what analysis is available to you.
Write the proposal and get it approved
Three gates in the one documented procedure. Approval is a boundary that protects you as well as a hurdle.
Get ethics clearance, then collect
In that order, with float in front of the gate. Never overlap them.
Analyse as you collect, not after
The final week's checklist says this explicitly, twice — once as a Do and once as a belief to correct.
Draft early and show it
Both the checklist and the write-up week say so. A first draft is a milestone, not an achievement; the work is what turns it into a final draft.
Edit, then assemble, then submit
The one worked schedule allocates no time at all to editing, and the write-up material says to take as much as is needed. Do not repeat the schedule's omission.
A closing word on the material
This library has been candid about where fourteen weeks of teaching material falls short, and that record stands. But a lifecycle page is the right place to say the other thing plainly.
What to carry forward
- Eight stages. Three strongly served — literature, ethics, write-up. One barely present — statistics, and it is the one that can invalidate work already done.
- Plan from the gates outward. Tasks take the time you give them; gates take the time they take.
- Ethics clearance precedes data collection, always. Data gathered before clearance cannot be used.
- Half the lifecycle is recording rather than doing, and the Methods chapter is written from the record, not from the data.
- Decide your analysis before you collect. Response formats you choose now decide what is possible later, and nothing recovers it afterwards.
Frequently asked questions
Where do most research projects actually go wrong?
At the joins between stages rather than inside them — a method chosen before the question settled, an analysis planned after the instrument was in the field, or data collected before ethics clearance. The weekly teaching structure cannot show these, which is why this page exists.
How long does ethics clearance take?
The material gives one worked case: five rounds over more than five months. That is a single recorded instance, not a norm or a planning figure. What it establishes is that clearance is not a formality and needs float in front of it.
What is the one thing I should start doing today?
Keep a record of how you are doing the research, not just what you find. The Methods chapter is written from that record, and the library's central case study is a researcher whose research went well and who could not write it up because the record did not exist.
Which gap do I have to close before I start?
Statistics, if your project involves quantitative data. The material gives one rule expressed in a unit it never defines, no measure of dispersion and no inferential procedure. Every other gap costs you time; this one can cost you the study.
Can I analyse data before collection finishes?
Yes, and the final week's checklist says so twice — once as something to do and once as a belief to correct. Overlapping analysis with collection is good practice; overlapping collection with ethics clearance is not.
Is this material enough on its own?
For the proposal, the literature review, ethics and writing up, largely yes. For sampling, instrument design and statistics, no — and the material does not claim otherwise. It describes itself as a reference set to return to during a thesis rather than a self-contained course.
References and source attribution
- The supplied teaching source: weekly study notes, slide decks and assessment activities for a master's-level research methods subject in project management, covering fourteen topics. Author, institution and year not stated in the supplied files.
- A worked thesis schedule supplied as a project file with its printed Gantt chart, treated throughout this library as anonymised student work rather than as teaching material.
- Anonymised student submissions supplied with the teaching material, used as observed practice rather than as model answers.
Suggested questions for Ask KEVOS
- What are the stages of a research project from idea to submission?
- Where should I put float in a research schedule?
- What should I be recording while I collect data?
- Which parts of a research project does this material not cover?
- How do the approval gates in a research project affect my timeline?