Step 1: Generating and Refining a Research Idea
The first numbered step in the thesis process, and the one with the worst ratio of effort to consequence. An idea that survives the screen buys you months of productive work; one that does not should be abandoned inside an afternoon.
Why the first step is the expensive one
The source then sets two constraints that pull against each other. The topic must be "narrow and focused enough to be interesting, yet broad enough to find adequate information". Neither end is negotiable, and most failed topics fail at one end or the other rather than somewhere in the middle.
There is also a prerequisite that is easy to skip: "Before selecting your topic, make sure you meet the requirements of what your final project should look like." Read the assessment or deliverable specification before you generate ideas, not after you have become attached to one. A topic that cannot produce the artefact you are required to submit is not a topic.
If we knew what it was we were doing, it would not be called research, would it?
Keep the hedge. The source does not claim the attribution is verified, and neither should you. What the line is doing in the material is licensing a degree of uncertainty at the start — you are not expected to know the answer, only to have picked a question worth not knowing the answer to. The source balances it with a plainer proverb: well begun is half done.
The three places the source says ideas come from
The source is explicit that ideas can come from any source, then names three. For a working project manager all three are usually available at once, which is a luxury and a trap — the ideas that arrive from all three directions simultaneously feel obvious and are often too big.
Three generators, in the source's own terms
A subject that held your attention
"Is there a particular subject that interested you and any specific area in the subject that you could research?" The advantage is that you already have the vocabulary and some of the reading. The risk is choosing a subject you enjoyed rather than a problem that exists.
The work actually in front of you
Your work in your industry. This is the richest generator for a practitioner and the one most likely to produce something your organisation will act on. It is also the one where access, confidentiality and your own position need thinking about early.
What the trade and the journals are arguing about
Readings in the media and journals that relate to your industry — "maybe a theme or topic that stimulates debate and is topical". A live argument is a gift, because a topic already being contested is a topic with literature attached.
A different part of the same teaching material reproduces a longer account of where research problems originate, including requests for proposals and the researcher's own thinking. That treatment is separate and fuller; see where research topics come from. If you are short of candidates rather than short of a screen, start there. If you are stuck for a domain, the material also lists broad research themes in project management.
Triage: which of the three kinds is this?
Before screening the idea for viability, the source asks you to classify it. Its statement is short — research can generally be one of three types — and the classification matters because it determines what your idea must eventually produce.
THE THREE KINDS, AND WHAT EACH COMMITS AN IDEA TO
| Kind | Source gloss | The idea has to be phrased as | And it will need |
|---|---|---|---|
| Descriptive | Finding out, describing what is | A situation nobody has yet characterised in your context | Access to the situation, and a way of recording it systematically |
| Explanatory | Explaining how or why things are as they are, and using this to predict | A relationship between two things you suspect is real | Enough cases or depth to support a claim about why, not just what |
| Evaluative | Evaluation of policies and programmes | Something already in place that can be judged against a stated intent | A benchmark or objective to evaluate against, and usually a sponsor |
Column two is the source's wording. Columns three and four are our reading of what each kind demands at idea stage.
Most raw practitioner ideas arrive as evaluative — something is not working and you would like to know whether it should be replaced. That is a legitimate kind, but it is the one most dependent on somebody else's cooperation. The three kinds are set out properly on descriptive, explanatory and evaluative research; at this stage you only need to know which one you are holding.
The eight screening questions
The source then supplies a list of things to keep in mind when deciding on a topic. Read as a screen, it is unusually direct about wanting you to abandon ideas quickly.
Two features of that list are worth noticing before you use it. First, six of the eight are about the literature; only the opening two are about the practicalities of the idea itself. The source is telling you where it thinks ideas die.
Second, two of the eight are not questions at all. Items four and eight carry rules — one says what to do when literature is thin, the other says what it means when journal articles exist. Treat those two as the outputs of the screen rather than as further inputs to it.
Running the screen on one candidate idea
The following pass is constructed for this page as an illustration of the screen in use. The candidate is drawn from the broad themes the material lists — managing transition from construction to operations — and the answers below are the kind of reasoning the screen is looking for, not findings about that field.
AN ILLUSTRATIVE PASS THROUGH THE EIGHT
| Screen item | What you are actually looking for | How the candidate fares |
|---|---|---|
| Sensitivities that stand out | Anything that touches confidentiality, contracts, live disputes or people's performance | Handover disputes are commercially sensitive. Flagged, not fatal — it narrows which organisations will talk to you |
| Enough to stop or delay it | A blocker you cannot design around inside your timeframe | Only if the study depends on a single organisation's contract documents. Design around it by making the unit of analysis the practitioner, not the contract |
| Anything written at all | Evidence the topic exists as a topic, in any form | Yes — enough to proceed to the finer questions |
| The thin-literature rule | Whether you are about to enter territory nobody has mapped | Not triggered here. If a scan had produced almost nothing, the source's instruction is to seek advice and expect to be told to stop |
| Who wrote it, and when | Named, findable authors, and dates recent enough to still describe your industry | Record the names and the dates. A topic written about once, a decade ago, is a different proposition from one written about last year |
| Where it appears | The balance between journals, trade press and online commentary | If it is all trade press and no journals, you have an industry concern without a theoretical base yet |
| What is being said | Whether the existing work argues, or merely describes | Locate the journal articles themselves rather than working from abstracts and summaries |
| The journal-article rule | Whether the topic has been developed theoretically | If yes, the source treats this as a strong positive — the theoretical development is already done and you can build on it |
Constructed illustration. The candidate topic is taken from the material's list of broad project management themes; nothing in the right-hand column is a claim about that field.
The two ways ideas fail this screen
In the source's own framing, an idea fails either because something about it will stop or delay the work, or because nobody has written about it. Those two failures behave very differently.
Keep, reshape or drop
What the screen result should trigger
Before you take the idea to step two
- You have read the requirements for the final deliverable, and this idea can produce it
- You can say whether the idea is descriptive, explanatory or evaluative, and you have not hedged
- You have run all eight screen items and written the answers down, including the ones that came back weak
- You have located actual journal articles, not just references to them
- Any sensitivity you found is recorded, with the design change that works around it
- You still have at least one alternative idea alive, in case this one does not survive step two
That last item is the one practitioners skip. Generating a second and third candidate costs an hour and makes it psychologically possible to drop the first — which the screen may well tell you to do. The next step, testing the topic, is where an idea becomes something you can defend; the full sequence it belongs to is on the ten-step thesis model.
What to carry forward
- The source calls topic selection often the hardest part and ties the quality of the whole project to it. Budget real time for it.
- Three generators: your studies, your industry work, your industry reading. Check the deliverable requirements before you commit to any of them.
- Classify the idea as descriptive, explanatory or evaluative at the outset — the kind determines what the idea must eventually produce.
- Six of the eight screen items are about literature. If nobody has written on your topic, the source expects you to be advised against it.
- Move on quickly from ideas with blocking sensitivities, and keep a spare candidate alive so that moving on stays easy.
Frequently asked questions
How much literature is enough for a topic to pass the screen?
The supplied teaching source does not say. It contrasts "little literature", which triggers a warning, with the topic being addressed in journal articles, which it treats as strongly positive. In the absence of a threshold, the practical test is whether you can find several journal articles that argue with each other about your topic.
Can I research something my organisation has asked me to look at?
The source acknowledges that a supervisor or host organisation may assign a specific topic. It gives no separate guidance for that case, so run the same eight-item screen on the assigned topic and raise any blocking sensitivity early — an assigned topic with no literature is still a topic with no literature.
What if my idea is interesting but nobody has written about it?
The source's instruction is explicit: seek advice as to whether to continue, and expect to be advised against it. The reasoning is that without literature you have no justification for the study and nothing to attach your findings to at the end.
Do I have to decide between descriptive, explanatory and evaluative this early?
The source presents the three types at the idea stage, which suggests an early classification. It is not irreversible — an idea often shifts from evaluative to descriptive once you discover no benchmark exists to evaluate against — but you should always be able to say which one you are currently working to.
How many candidate ideas should I generate?
The source sets no number. It does tell you to move on quickly from ideas with blocking problems, which is only realistic if something else is waiting. Keeping two or three live is our practical suggestion rather than a source requirement.
References and source attribution
- Naoum, S. G. 2013, Dissertation Research & Writing for Construction Students, 3rd ed., Routledge (set as reading on topic selection, pp. 9–11).
- O'Leary, Z. 2017, The Essential Guide to Doing Your Research Project, 3rd ed., Sage Publications, London (listed among the source's references).
- Trochim, W. M. K. 2006, Research Methods Knowledge Base (http://www.socialresearchmethods.net/kb/probform.php) — excerpted in the same teaching material for its fuller account of where research problems come from.
- The supplied teaching source: master's-level teaching material on developing a research topic, supplied as study notes and accompanying slides. The three generators, the three kinds and the eight screen items are quoted from that material.
Suggested questions for Ask KEVOS
- What are the eight questions to screen a research idea before committing to it?
- Where does the source say research ideas come from?
- How do I tell whether my idea is descriptive, explanatory or evaluative?
- What should I do if there is almost no literature on my research topic?
- How do I screen a research topic that my organisation assigned to me?