Common Mistakes in Project Management Research
None of these is exotic. Each is a failure the teaching material warns about in a single sentence, and each is recoverable if you catch it in the week it starts rather than the month it surfaces.
Where the source's warnings actually sit
None of this material is presented as a list of mistakes. Each warning arrives attached to a step, usually in a single clause, and is easy to read past. The warnings below are the source's; the assembly is this page's.
THE EIGHT WARNINGS AND WHERE THEY APPEAR
| Failure mode | Where the warning sits | Form it takes in the source |
|---|---|---|
| Taking on too big a project | Reviewing the research topic | A parenthesis: many people take on too big a project at first |
| A topic with little literature | Beginning with an idea | An instruction to seek advice, and a prediction that you will be advised against continuing |
| A topic you are not interested in | Reviewing the research topic | A question to ask yourself, with the consequence stated as a real slog |
| Methods that do not fit the aims | Reviewing the topic, and designing the research | A screening question, plus the rule that methods are only as good as the answers they give |
| Leaving ethics until late | Seeking ethics clearance | Placed at step six of ten, before conducting the research; also a named feasibility consideration |
| Selective observation and biased interviews | The definition of descriptive methodology | Stated as a property of the methodology itself: it can be distorted by these |
| Not backing up drafts | Checking the literature | An instruction to keep back-up copies in a separate location, illustrated with a story |
| Writing the report before concluding | Writing your report | A flat prohibition: you cannot write a report before you have collated, interpreted and concluded |
Mistakes made while choosing the topic
Three of the eight are set before you have written anything. They are also the three that respond best to intervention, because at this stage changing your mind costs nothing but pride.
Taking on too big a project
The source states it as an aside — many people take on too big a project at first — and pairs it with the requirement that a topic be narrow and focused enough to be interesting, yet broad enough to find adequate information.
It happens because scope inflation feels like ambition, and because a practising project manager is professionally accustomed to scope being negotiated during delivery. In a research project the scope is fixed by the question, and the question is examined.
TOO BIG A PROJECT — DETECTION AND RECOVERY
| Early signal | What to do about it |
|---|---|
| Your question needs two sentences | Delete the second sentence. If the first no longer describes a study, the second sentence was the study and the first was a theme. |
| You cannot name the population on one page | Narrow on one axis only: one setting, one role, or one time window. Changing all three at once usually breaks the question. |
| Your literature search returns results you could never read | The topic is a field. Find the sub-question inside it that the field has not settled. |
| Your title contains no exclusion | The source asks for a title that illustrates the boundaries of the topic. Rewrite it so it says what is out as well as what is in. |
Recovery is a cut, not a compression. Keep the discarded scope written down as further research — it costs a paragraph in the conclusion and it demonstrates you understood what you were setting aside. The narrowing mechanics are at Narrowing and Scoping a Research Topic.
A topic with little literature
This is the source's bluntest warning: if there is little literature on the topic, seek advice about whether to continue, and it is likely you will be advised against continuing. The reason it gives is positive rather than negative — a topic written about in journal articles has been addressed and developed theoretically, which is what a beginner researcher can build on.
It happens because practitioners choose problems from work, and the most infuriating problems at work are often the ones nobody has published on. Novelty feels like contribution. In a master's project it is more often a structural risk: no theoretical frame to position against, no instruments to adapt, and nothing to argue with in your conclusions.
Detect it early by running the source's own four questions — who has written on this, when, where, and what — and noticing when the answers come back thin. See Step 3: Checking the Literature.
A topic you are not interested in
The source puts this first among the questions to ask yourself, and states the consequence in one word: a real slog. It also acknowledges that a supervisor or host organisation may assign you a topic, which is precisely the situation in which this failure occurs.
The early signals are behavioural. Reading gets deferred, drafts do not appear, and you explain the project by describing the assessment requirement rather than the question. On a part-time project running over months, motivation is not a soft factor — it is the schedule.
Recovery is negotiation rather than abandonment: find the version of the assigned topic you would want the answer to, and propose that. The full set of self-directed questions is at Step 2: Reviewing and Testing Your Research Topic.
Mistakes made while designing
Methods that do not fit the aims
This is the most consequential of the eight, because it is the one that cannot be fixed by working harder. The source asks it directly as a screening question — how well do the research methods fit the aims of the research, and are they the best way of answering the research questions — and reinforces it twice: methods are only as good as the answers they give, and deciding on research methods makes no sense until you have decided on your research questions.
It happens through familiarity. A survey looks tractable, can be built in an afternoon and produces a satisfying quantity of data. It is then asked to answer a question about why people behave as they do, which it cannot do.
Diagnosing the misfit
Recovery almost never means collecting more data. It usually means making a smaller claim that the data you have can carry — and saying so in the limitations. The selection procedure is at Choosing a Research Methodology.
Leaving ethics until late
The source places ethics clearance at step six of ten, before conducting the research, and names ethical constraints as one of four practical feasibility considerations. Its screening questions are short: does the project require access to confidential or personal information, are the questions appropriate and respectful of cultural norms, and are the methods of collection unbiased and ethical.
Mistakes made while collecting
Selective observation and biased interviews
This warning is unusual because it is attached to a methodology rather than a step. Descriptive methodology relies on observations recorded for subsequent analysis, depends on human responses, and — the source says plainly — can be distorted by selective observation and biased interviews.
Practitioner research is exposed to both by construction: you are studying a setting you already have views about, recruiting through a network that shares them, and interviewing people who know what you hope to hear. The source's own characterisation of social science research includes that people are aware of research being conducted on them.
DISTORTION — SIGNALS AND SAFEGUARDS
| Early signal | Safeguard |
|---|---|
| Every interview so far has confirmed your expectation | Deliberately recruit someone you expect to disagree. If nobody disagrees, that is a finding worth reporting, not a relief. |
| All participants came through your own contacts | Record how each participant was reached, and report the recruitment route as a limitation. |
| Your notes record conclusions rather than what was said | Record verbatim where consent allows. Keep interpretation in a separate column from observation. |
| Nothing disconfirming appears in your summaries | Search transcripts specifically for counter-evidence before you write. The source requires you to establish where data contradicts the theory as well as where it supports it. |
| You are studying people who report to you | Treat this as an ethics question first and a bias question second. Both apply. |
Recovery here is honest reporting rather than repair. A stated, specific limitation costs you far less than an unstated one that an examiner finds. The methodology and its exposures are covered at Descriptive Research Methodology.
Mistakes made while writing
Not backing up drafts
The source's instruction is concrete: note and date all your drafts and references as you go along, keep back-up copies of your writings, and store copies in a separate location such as a portable hard drive. It illustrates the point with a story about a researcher who lost a nearly finished doctoral thesis in a house fire — an illustration, not evidence, but the instruction stands on its own.
This is the lowest-status task in the project and the only one whose failure is total. Part-time research spreads across several machines and versions multiply. The signals are simple: you cannot say where the current version is, or your reference list exists in exactly one file.
The five-minute fix
- One canonical location for the current draft, and you can name it without thinking
- A copy in a separate physical location, as the source specifies
- Drafts dated, so you can tell which is current without opening them
- References captured in a form that survives the loss of the draft
- Raw data stored separately from analysis, so a corrupted working file does not take the evidence with it
Writing the report before concluding
The source is categorical: you cannot write a report before you have collated your information, interpreted it and come to some rational conclusions. Only then do you summarise the project and communicate the results to your specific audience.
It happens because writing is the visible artefact and analysis is not, so under schedule pressure the temptation is to produce pages. A half-truth encourages it: the context and literature genuinely can be drafted early, which creates the impression that the rest can be too.
Legitimate to draft early
- The problem statement and why it matters
- The literature review and theoretical framework
- The methodology as designed, with justification
- Descriptions of instruments and how access was obtained
Not legitimate to draft early
- Findings, before the data is collated
- Interpretation, before the analysis is done
- Conclusions, before interpretation has produced anything
- The abstract, which describes a study that has happened
The tell is a conclusions section that restates the literature review. If your conclusions could have been written before you collected anything, the data did not change your mind — and either the analysis has not been done or the study was not needed. The write-up sequence is at Steps 9-10: Writing the Report and Storing the Data.
Six more warnings the source makes in passing
THE SECONDARY REGISTER
| Warning | What the source says | Consequence if ignored |
|---|---|---|
| Not planning time for writing | In allocating time, plan half for doing the research and half for writing and rewriting | A design that consumes more than half the schedule leaves the write-up unfinishable |
| Ignoring the feasibility factors | Time, ethical constraints, cooperation and cost; failure to consider any of these can mean disaster later | The project stalls at the point where the missing factor bites, usually data collection |
| Not checking what the final project must look like | Before selecting your topic, make sure you meet the requirements of what your final project should look like | A topic that cannot be presented in the required form, discovered at the worst moment |
| Treating the process as fixed | The steps should be viewed as an ever-changing process rather than a fixed set of steps; questions evolve as you gather evidence | A question defended past the point where the evidence has moved on |
| Not seeking advice | Seek advice on a thin literature; speak to your supervisor about the most appropriate methodology | Errors that a fifteen-minute conversation would have caught survive to examination |
| Not storing the data | Your research report and your data need to be stored appropriately; the requirements are institution-specific and not reproduced in the source | A completed project that fails a requirement nobody told you about |
The feasibility factors are the ones to take most seriously, because the source frames them as tradeoffs between rigour and practicality rather than as a checklist to pass. That framing is worked through at Assessing Research Topic Feasibility.
Catching them in the week they start
The source provides no detection schedule; this one is derived from the order in which its warnings appear. Each checkpoint asks about the failure mode being set at that moment, not the one currently visible.
- While the idea is still an ideaIs there literature? Am I interested in this? Do I know what the final project has to look like? All three are cheap now and expensive later.
- Before the topic is fixedCan I name the population on one page and state the question in one sentence? Does my title show a boundary?
- Before the instrument is draftedIs the question fixed? Does each question map to data that will answer it? Have I re-run the fit test on nature, aim, population and context?
- Before anything is sent to anyoneDo I have the approvals? This is the last checkpoint at which an ethics failure is still preventable.
- After the first few responsesHas anything disconfirmed my expectation? Where did my participants come from? Is my recruitment route reportable?
- Weekly, throughoutIs the current draft backed up in a separate location, dated, with references captured independently?
- Before the findings chapterIs the data collated, interpreted and concluded on? If not, write something else this week.
Recovering from one you are already in
Late-stage recovery follows one principle: shrink the claim to fit the evidence, rather than stretching the evidence to fit the claim. A smaller defensible finding passes; an overstated one does not.
RECOVERY BY FAILURE MODE, LATE IN THE PROJECT
| If you are already in it | The recovery that usually works |
|---|---|
| Scope is too big and time has gone | Report on the part you completed properly and define the rest as further research. Do not report thin coverage of everything. |
| The literature turned out to be thin | Reposition against an adjacent literature and make the absence itself part of the argument, explicitly and with evidence of the search. |
| The methods do not answer the question | Narrow the question to what the data answers, and state the original question in the limitations as unaddressed. |
| Ethics approval was not obtained | Stop. Do not analyse. Take it to your supervisor and your ethics office the same day — this one cannot be solved inside the project. |
| The evidence looks one-sided | Search deliberately for counter-evidence, report the recruitment route, and state the exposure to selective observation as a limitation. |
| Drafts are scattered or partly lost | Reconstruct from the most recent complete version, then fix the backup before writing another word. |
| Conclusions were written before analysis | Rewrite them from the analysis outputs only. If nothing changes, say so — a confirmed expectation is a finding when the design could have disconfirmed it. |
One study in the exemplar volume reports that the programme it was following was not successfully implemented within the research timeframe, and reports on why. It was examined and published. A project whose subject fails is not a failed project, provided the design could still produce an answer.
What to carry forward
- Eight warnings, scattered across three weeks of teaching material, each stated once and easy to miss. Most are set in the first third of the project.
- Too big a project is the most common and the cheapest to fix early: cut on one axis and keep the rest as further research.
- A thin literature is a structural risk, not a sign of originality. Reframe the lens or the level rather than abandoning the setting.
- Methods that do not fit the aims cannot be fixed by working harder. Fix the question first, then the instrument, then the claim.
- Ethics and backups are the two failures with no recovery path. Both are prevented by a task that takes under an hour.
- Insider research buys access and costs objectivity. Recruit dissent deliberately and report the recruitment route.
- Conclusions that could have been written before data collection mean the analysis has not been done.
- Shrink the claim to fit the evidence. That single move recovers most late-stage problems.
Frequently asked questions
Which of these mistakes is the most expensive?
Methods that do not fit the aims, because it is only visible once the data is in and it cannot be fixed by working harder. Ethics failures are more absolute but rarer and entirely preventable. Scope failures are the most common and the cheapest to fix if caught while the topic is still being reviewed.
My supervisor says my topic has too little literature. Do I have to abandon it?
The source says to seek advice and predicts you will be advised against continuing, so take the advice seriously. But you usually do not have to abandon the setting you care about. Keep the setting and change the lens to a well-published concept, or keep the question and look for the same question asked in another industry.
How do I stop my own views distorting interviews in my own organisation?
Recruit at least one participant you expect to disagree with you, record how every participant was reached, keep verbatim records separate from your interpretation, and search your transcripts specifically for counter-evidence before you write. The source names selective observation and biased interviews as distortions of descriptive work, so the exposure is expected rather than shameful — but it has to be reported.
Can I write my literature review before I have collected data?
Yes. The source's prohibition is on writing the report before you have collated, interpreted and concluded — that applies to findings, interpretation, conclusions and the abstract. The problem statement, literature review, theoretical framework and methodology as designed can all be drafted early, and drafting them early is how you protect the second half of the schedule.
What is the minimum I should do about backups?
What the source specifies: keep back-up copies of your writings in a separate location, and note and date all your drafts and references as you go. Add one thing to that — keep your references and your raw data in files that would survive the loss of the draft, since those are the two things you cannot reconstruct from memory.
I am halfway through and my question has changed. Is that a mistake?
No, provided you update the design to match. The source describes the research process as ever-changing rather than a fixed set of steps, and says questions evolve as you gather evidence. It becomes a mistake only when the question moves and the methods, the literature and the claim stay where they were.
References and source attribution
- The supplied teaching source: consolidated week 1, week 2 and week 3 teaching notes, study notes and slide material. Every warning in the primary and secondary registers is drawn from it.
- Trochim, W. M. K. 2006, Research Methods Knowledge Base — the excerpt reproduced in the supplied source on the feasibility of a study and the tradeoffs between rigour and practicality.
- Clarke, R. J. 2005, Research methods and methodologies, pp. 35–45 — the work adapted for the methodology definitions, including the statement that descriptive work can be distorted by selective observation and biased interviews.
- O'Leary, Z. 2017, The Essential Guide to Doing Your Research Project, 3rd ed., Sage Publications, London.
- Naoum, S. G. 2013, Dissertation Research & Writing for Construction Students, 3rd ed., Routledge.
- An edited volume of master's degree research projects in project management, 2018, academic publisher, 378 pp. — the exemplar volume; the study reporting an unimplemented programme is paraphrased from its published abstract.
Suggested questions for Ask KEVOS
- Run the ten-minute self-audit against my current research plan and tell me which failure mode I am in.
- My scope is too big — help me cut it on one axis without losing the question.
- What can I legitimately draft now, before my data collection is finished?
- Build me a detection schedule for my project timeline based on these warnings.
- My findings all confirm what I expected. How do I test whether that is a result or a bias?