Descriptive Research Methodology
The source's shortest methodology definition carries its only explicit warning. Observation recorded for later analysis, dependent on human responses, and vulnerable to two named distortions you have to design against.
What the source defines
Each clause does distinct work, and it is worth separating them before you build anything. The first tells you what the methodology consists of. The second tells you where its evidence comes from and therefore what kind of claim it can support. The third tells you what will go wrong.
THE DEFINITION, CLAUSE BY CLAUSE
| Clause | What it establishes | What it obliges you to do |
|---|---|---|
| Relies on observations, recorded to be subsequently analysed | The methodology is a two-stage operation: capture now, interpret later | Design a recording protocol before you observe anything, because analysis can only work on what was recorded |
| Depends on human responses and observations | The evidence is generated by people — those observed and the person observing | Account for both parties in your method chapter, including yourself as an instrument |
| Can be distorted by selective observation and biased interviews | Two named failure modes, presented as inherent rather than incidental | Show what you did to mitigate each, since the source raises them and offers no defence |
Descriptive methodology is not a descriptive purpose
The supplied source uses the word descriptive in two places to mean two different things, and never flags the difference. If you use the word loosely in a proposal, a reviewer will ask which one you mean.
Descriptive as a purpose
- Defined in the terminology table as research that seeks to describe what is
- Its associated term is exploratory
- Sits alongside explanatory, which seeks to explain relationships, and evaluative, which tests outcomes against benchmarks
- Answers the question: what kind of knowledge is this study trying to produce
- Chosen at the point where you settle your research question
Descriptive as a methodology
- Defined in the methodologies table as reliance on recorded observation
- Labelled qualitative, alongside historical, comparative, evaluation and action
- Sits alongside correlation and experimental, which are labelled quantitative
- Answers the question: what strategy will this study be carried out by
- Chosen after the question, before the instruments
The two are related but not equivalent. A descriptive purpose can be pursued by other means — a correlation study describes a pattern of association without observing anybody. And a descriptive methodology can serve an evaluative purpose, which is precisely what the source's definition of evaluation research does. The purpose-level distinction is set out in Descriptive, Explanatory and Evaluative Research, and the methodology's neighbours in Research Methodologies: An Overview.
The observe, record, analyse chain
The phrase recorded to be subsequently analysed describes a separation in time between two activities that project managers habitually run together. In delivery work you observe and interpret simultaneously — that is what situational judgement is. In descriptive research the two are deliberately pulled apart.
- Decide what would count as an observation
- Observe
- Record faithfully
- Set aside
- Analyse the record
- Interpret against the literature
The separation is not bureaucratic. It exists because interpretation contaminates observation: once you have decided what a behaviour means, you stop noticing the parts that do not fit the meaning. Recording first, and analysing a record rather than a memory, is the only defence the design offers.
Why the source calls it qualitative
The label follows from the second clause. Because the evidence depends on human responses and observations, it arrives as words, accounts and descriptions rather than as measurements on an agreed scale. The source's definition of qualitative research is exactly that: research in which words, and possibly images and sounds, are the medium.
Descriptive studies frequently produce counts as well, and the source anticipates this. It describes classifying things into categories — the example it gives is categories such as gender, nationality and age — and calls the result attribute or categorical data. It notes that we are often interested in the number of people in each category, and that these counts and proportions become the basis for analysis.
That last point has a practical consequence. If your descriptive study produces category counts, do not describe them as qualitative in a proposal without explaining that you are following the teaching source's framing. The distinction between the two research designs is worked through in Qualitative and Quantitative Research Compared.
The two distortions the source names
These are the only failure modes the source attaches to any of its seven methodologies. That makes them worth taking seriously as a signal about where this methodology characteristically fails, even though the source supplies nothing beyond the two phrases.
Selective observation
The source names selective observation and stops. In plain terms the phrase describes noticing the instances that fit what you already expect and passing over the ones that do not — and the danger is that it feels like nothing at all from the inside. You do not experience yourself as ignoring evidence; you experience yourself as recognising a pattern.
For a project manager researching their own organisation the exposure is high. You already have a view about why the delay happened. You already know which teams are difficult. Every one of those priors is a filter operating on what you record.
How selective observation gets into a study
Choosing where to look
Observing the projects, meetings or sites that are convenient or that you expect to be informative. The sample is then a sample of your expectations, and no amount of care during observation recovers it.
Choosing what counts as an event
Recording the exchanges that seem significant and letting the rest pass. Significance is a judgement, and making it in the moment is the interpretation you were supposed to defer.
Choosing what to write down
Summarising what happened rather than describing it, so that your interpretation is baked into the record and cannot be separated from it later.
Choosing what to report
Returning to the record for confirming instances and treating the rest as noise. This is the one stage where a colleague reading your raw record can catch you.
Biased interviews
The second distortion is the interview itself. The source names it without elaboration, but its own framing of social science research explains why it is unavoidable rather than merely possible: people are aware of research being conducted on them, and they are less predictable than non-human phenomena.
For an internal researcher the problem sharpens. When you interview a colleague, they are not answering a researcher; they are answering someone with a role, a history and possibly influence over their work. What they choose to tell you is shaped by that, whatever your intentions.
WHERE BIAS ENTERS AN INTERVIEW
| Point of entry | What it looks like | A practical response |
|---|---|---|
| Who you interview | The willing, the senior and the articulate are over-represented; the people the question is actually about are not in the room | Define the population from the question, then record and report who declined or was unavailable |
| How the question is worded | The question contains the expected answer, or a premise the respondent has to accept in order to reply | Test the wording on someone outside the study and ask them what answer they think you want |
| Who is asking | Your role in the organisation makes some answers costly and others safe | State your position and its likely effect in the method chapter rather than claiming neutrality |
| How you respond in the moment | Agreement, follow-up and body language reward some answers and close down others | Use a consistent probe, and note in the record where you departed from it |
| What gets transcribed | The quotable lines survive and the hesitations, contradictions and qualifications do not | Record and transcribe fully, and keep the qualifications — they often carry the finding |
The two distortions are the source's. The entry points and responses in this table are a practical reading, not source content.
Descriptive designs in project management research
Descriptive work is the mainstay of practitioner research, and the exemplar volume of master's project management research bears that out. Across its seventeen designs, survey and interview dominate; almost every design pairs a literature review with one or two primary collection methods. That is an observation about the supplied volume rather than a published statistic, but it tells you what your examiners are used to reading.
Is descriptive the right methodology for your question?
Pre-flight before you observe anything
Settle these before the first observation or interview
- You can state what counts as an observation in your study, precisely enough that someone else could apply the definition.
- Your recording template separates description from interpretation, and you have tested it on one session.
- Your observation schedule and interview list were fixed in advance and you can show why those sites and those people.
- You have written down what you expect to find, dated it, and put it away to check against later.
- You have a plan for recording disconfirming instances and reporting how many there were.
- You have described your own position relative to the people you are studying, rather than assuming it away.
- Ethical approval covers what you intend to observe and record, including anything incidental you may capture.
What to carry forward
- The source's definition is three clauses: recorded observation analysed later, dependence on human responses, and vulnerability to two named distortions.
- Descriptive methodology and descriptive purpose are different things sharing a word. Say which one you mean.
- Recording and analysing are deliberately separated in time. Design the recording protocol before you observe anything.
- Selective observation is invisible from the inside — it feels like recognising a pattern. Fix your schedule in advance and hunt for disconfirming cases.
- Interview bias is structural for an internal researcher. State your position rather than claiming neutrality.
- Evaluation research is defined as a kind of descriptive research, so both distortions apply to evaluations as well.
Frequently asked questions
How does descriptive methodology differ from descriptive research purpose?
A descriptive purpose is a kind of question — research that seeks to describe what is, sitting alongside explanatory and evaluative purposes. A descriptive methodology is a strategy for carrying out a study, defined in the source as reliance on recorded observation analysed subsequently. You can pursue a descriptive purpose using a different methodology, and you can use a descriptive methodology to serve an evaluative purpose.
What are the two distortions the source names?
Selective observation and biased interviews. The source names both in a single clause of the methodology definition and does not define either, does not explain how they arise, and offers no mitigation. Any defence you build must therefore be described as your own design decision and justified accordingly.
Why is descriptive research labelled qualitative when I am counting things?
The source labels the methodology qualitative because its evidence depends on human responses and observations, arriving as words rather than measurements on an agreed scale. It also treats category counts as attribute or categorical data and describes that as a form of qualitative data. Note that this is a teaching simplification, since many texts treat categorical and ordinal data as levels of measurement that can be handled quantitatively.
Can I use descriptive methodology inside my own organisation?
Yes, and it is the common case for practitioner research. The two distortions are sharper there: you already hold views about what is happening, and colleagues answer you as a person with a role rather than as a neutral researcher. Fix your observation schedule and interview list in advance, and state your own position in the method chapter instead of claiming neutrality.
Does the source give an observation protocol?
No. It states only that observations are recorded to be subsequently analysed. There is no guidance on what to record, when to record it, how to sample observations, or what makes a record faithful. Those decisions are yours to make and to justify, and they should be described explicitly rather than left implicit.
Is evaluation research the same as descriptive research?
The source defines evaluation research as descriptive research designed to deal with complex social issues, with the added condition that participants are equal partners in the design, implementation, interpretation and results. So it is a species of descriptive research rather than a separate family, and both named distortions apply to it.
References and source attribution
- Clarke, R. J. 2005, Research methods and methodologies, pp. 35–45 (the work the supplied source adapts its methodologies table from).
- Bryman, A. 2016, Social Research Methods, 5th ed., Oxford University Press, Oxford.
- O'Leary, Z. 2017, The Essential Guide to Doing Your Research Project, 3rd ed., Sage Publications, London.
- Veal, A. J. 2005, Business Research Methods: A Managerial Approach, Longman (the source of the social science research characteristics cited in the teaching material).
- The supplied teaching source: Week 2 study notes and slide deck on introduction to research methods, and Week 3 study notes on qualitative and quantitative data.
- An edited volume of master's degree research projects in project management, 2018, academic publisher, 378 pp. (the exemplar volume of seventeen research designs).
Suggested questions for Ask KEVOS
- Is my research question descriptive as a purpose or does it need a descriptive methodology?
- How do I design an observation protocol that separates description from interpretation?
- What can I do about selective observation when researching my own team?
- How should I word interview questions to reduce bias in an internal study?
- Does the descriptive label mean I cannot count anything in my study?
- How do I write up the limitations of a descriptive study without weakening it?