Refining Your Topic Before the Proposal
Four decisions you made loosely at the start have to be made again, tightly, before a proposal can be written. The supplied source is unusually blunt about what makes this hard, and about how fast you should be willing to drop an idea.
A second pass over four decisions you have already made
- Select a topic
- Review literature
- Conceptualise
- Define the research question
The word doing the work is revisit. These four elements were worked through when the idea first arrived. What is asked for now is a second pass under a harder standard — that of a document naming a method, a site, a dataset and a boundary.
That standard changes the answers. A topic that survived the first pass on interest and plausibility can fail the second on access alone — which is where an area you would like to know more about becomes a project with a delivery risk attached.
THE SAME FOUR ELEMENTS, TWO PASSES APART
| Element | First pass — is this worth doing? | Second pass, before the proposal — can this be done? |
|---|---|---|
| Select a topic | An area you are interested in, competent in, and can sustain for the duration | One idea chosen out of several, with the others written down and formally set aside |
| Review literature | Enough reading to establish that the area is live and that others have written on it | What is known, what is settled, what authors themselves nominate for further research |
| Conceptualise | The key ideas that keep appearing when you describe the topic | A statement containing all the key concepts, in the form a reader could search on |
| Define the research question | A working question that gets you into the literature | One statement or question, fixed, from which aims and objectives can be written |
The four element names are the source's. The two-pass reading is this page's; the source instructs you to revisit the elements without saying what changes between passes.
The first pass, with the source's separate screening questions, is at Step 2: Reviewing and Testing Your Research Topic. This page is the second, run immediately before drafting.
Too many ideas is the diagnosis, not too few
Most practitioners arrive at a master's project with a surplus. Years of delivery generate more unanswered questions than one study can carry, and the instinct is to keep the good ones in reserve rather than kill them.
The source's framing makes that instinct the problem. If the difficulty is deciding which idea to focus on, refinement is subtraction — choosing between several and disposing of the rest properly.
Five questions to put to the literature
The first five of the twelve address what you have read. They are not a literature review — they check whether the reading has told you enough to commit.
THE FIVE LITERATURE-CHECK QUESTIONS
| The source's question | What it settles | The answer that means you are not ready |
|---|---|---|
| What are the authors of those journal articles saying about the topic? | Whether you can characterise positions in the field, not just findings | A list of what individual papers found, with no sense of where they agree or differ |
| What is known about the topic? | The current state of knowledge, which is what your justification will be measured against | "Quite a lot" or "not much". Both are assertions; neither can be pointed at in a proposal |
| What issues and aspects have been thoroughly explored? | Which parts of the area are closed, so you do not propose to reopen them | Silence. If nothing is thoroughly explored, the reading is not yet deep enough to tell |
| What issues, aspects and questions do the authors recommend for further research? | A gap nominated by the field rather than one you inferred | You have not read the closing sections. This is the cheapest gap-finding move available and it is routinely skipped |
| What issues are yet to be explored, and is there one among them you could or should explore? | Whether an unexplored issue is also yours — within your access, competence and time | An issue that is genuinely unexplored and that you have no route into. Interesting is not the same as available |
Column one close-paraphrases the source's questions. Columns two and three are this page's reading; the source poses the questions without discussing answers.
Seven questions about method and access
The second set moves from what is known to what you could actually do. The order matters: the source starts by asking what other researchers did, before asking what you would do.
THE SEVEN METHOD-AND-ACCESS QUESTIONS
| # | The source's question | What it is testing |
|---|---|---|
| 1 | What methodologies did these researchers use in their study? | Whether the field has an established way of approaching this kind of question — and whether you are proposing to depart from it deliberately or by accident |
| 2 | What kind of data collection methods did they use? | The practical instruments the topic tends to require, and whether you can run them |
| 3 | Where would you conduct the research? Do you have access to a site or an organisation? | Whether the study has a location. A topic with no site is not yet a project |
| 4 | What kinds of data would you look for to produce some new knowledge in this field? | The link between evidence and contribution — what data would have to exist for the answer to be new |
| 5 | How would you go about gathering the data? | The procedure, in enough detail that a timetable and an ethics application could be built from it |
| 6 | Where is that data? Is it in records and reports? Is it in the testimony of people? Do you have access? | Location, form and permission, in one question. Treated separately below |
| 7 | Why would you gather the data in that manner? | Justification. This is the question that separates a chosen method from a default one |
Close-paraphrased from the supplied source, which poses all seven and answers none of them.
Question seven is the one that gets assessed. The component table asks the research design section for a rationale for every selection, not just a description, so your answer here becomes text in the document — see The Research Design Section.
Where is that data, and do you have access?
Three questions are compressed into one, and each has a different consequence. Where is a location problem. Records or testimony is a design decision that determines your whole method. Access is a permission problem, and it is the one that ends projects.
If the data is in records and reports
- You need permission to use organisational records, which is a different conversation from permission to speak to people
- The records were created for another purpose, so their definitions are not yours — check what a field actually contains before you design around it
- Coverage and consistency over the period you need are the first things to verify, not the last
- Two datasets you intend to join must share a key — confirm that before the proposal, not during analysis
- Records held by a third party may need a data agreement your timetable has not allowed for
If the data is in the testimony of people
- You need named participants, a recruitment route and a plan for approaching them
- Ethics clearance becomes unavoidable and sits on the critical path of the whole project
- If the participants are colleagues or report to you, the design has to deal with an unequal power relationship
- Consent, confidentiality and what happens to recordings all become proposal content, not afterthoughts
- Availability is a schedule risk: interviews get postponed, and yours are not the interviewee's priority
Accept and reject quickly
Speed is being asked for as a skill, not a shortcut. The two instructions only look opposed. Refinement is what makes a fast rejection possible, because a vague idea cannot be rejected on any specific ground.
Grounds for a fast rejection
The end state: one statement or question
The output of the whole pass is a single sentence. Not a paragraph, not a title, not a list of sub-questions — one statement or one question that contains the project.
Tests on the one statement
- It is one sentence, and you can say it aloud without notes
- Every key concept in the project appears in it
- A population or a context is visible in it, not implied
- It names or clearly implies where the evidence would have to come from
- You can describe what an answer would look like, including one you would not like
- Nothing in it is joined by and where the two halves need different data
- It is narrow enough that aims and objectives can be written directly from it
The source is explicit about what happens next: once you have the properly conceptualised research statement or questions, the next step is to develop your aims and objectives. That step is covered at Aims, Objectives and Questions in a Proposal, and the fuller interrogation of the question itself at Step 4: Deciding Your Research Question.
The conceptual framework, defined a second way
One usable point survives the confusion. On either definition the concepts and the statement must contain each other: if a key concept does not appear in your one sentence, either the sentence is incomplete or the concept is not central.
What to carry forward
- Refinement is a second pass over four decisions already made — topic, literature, conceptualisation, question — under the harder standard a proposal imposes.
- The problem is choosing between ideas, not having them. Broadening the topic to keep three ideas alive is the standard failure.
- Five questions test the reading; the most productive one asks what the authors themselves nominate for further research.
- Seven questions test method and access, and the last of them asks for the rationale that will end up in your research design section.
- "Where is that data — records, or testimony — and do you have access?" is the question that ends topics. Test it in an afternoon, across your whole shortlist.
- Working in an organisation is knowledge of where the data sits, not permission to use it.
- The end state is one statement or question containing every key concept. Aims and objectives are written from it, not before it.
Frequently asked questions
How is this different from testing the topic at the start of the project?
The first pass asks whether a topic is worth doing — interest, competence, boundaries, timeframe. This pass asks whether it can be done, under the standard a proposal imposes: a named site, an identified dataset, a method with a rationale and a question narrow enough to write aims from. A topic can pass the first test and fail this one on access alone.
What if I genuinely cannot choose between two topics?
Run the access question across both at once rather than working through either in depth. Ask where the data sits, what form it takes and whether you have permission, and one of the two usually eliminates itself quickly. The source's instruction is to be able to accept and reject a topic quickly, and access is the fastest discriminator available.
Is one statement really enough for a whole research project?
It is the end point of refining, not the whole proposal. The source says you should be able to state your idea in one statement or question, and that this properly conceptualised statement is what you then develop aims and objectives from. Sub-questions, aims and objectives all come afterwards and all have to trace back to that sentence.
Which definition of conceptual framework should I use?
Whichever you use, define it where you use it. The supplied material defines the term two incompatible ways — as the research statement itself in one place, and as the structure of concepts from which the theoretical framework emerges in another. Neither is corrected by the other. Stating your definition in your own text costs a sentence and prevents a reader marking you against the other one.
Do I have to have confirmed access before submitting the proposal?
The supplied material does not set that as a requirement, but it asks the access question twice in twelve, which is a strong signal. In practice a proposal that names a site and a dataset without any confirmation carries a risk that surfaces later, when it is expensive. One conversation with a data custodian or manager before drafting either removes the risk or removes the topic.
References and source attribution
- Naoum, S. G. 2013, Dissertation Research & Writing for Construction Students, 3rd ed., Routledge — cited in the supplied source as a reading on topic selection.
- O'Leary, Z. 2017, The Essential Guide to Doing Your Research Project, 3rd ed., Sage Publications, London — one of the general research methods texts cited across the supplied source.
- Quinlan, C. 2011, Business Research Methods, 1st ed., Cengage Publishing — cited in the supplied source, and the origin of its framework for conceptualising a research project.
- Bryman, A. 2016, Social Research Methods, 5th ed., Oxford University Press, Oxford — cited in the supplied source.
- The supplied teaching source: consolidated weekly teaching notes and slide material on developing a research proposal, which supplies the four elements to revisit, the twelve review questions, the accept-and-reject instruction and the restated definition of the conceptual framework quoted on this page.
Suggested questions for Ask KEVOS
- Run the twelve review questions across my three candidate topics and tell me which one survives.
- My topic needs data from two systems. Help me work out whether they can actually be joined.
- Reduce this paragraph describing my research interest to one statement containing all the key concepts.
- Which of my key concepts appear in my research statement, and which appear nowhere else in the proposal?
- Draft the questions I should ask a data custodian before I commit to this topic.