Sampling and Selecting Participants
The week that named sampling in its objective taught it in one sentence and withdrew the sentence. A later week supplies six sampling types, non-response bias and a real argument for sampling at all. This page states what is now taught and what is still only named.
What a later week supplied
This page was written to record an absence: the data-gathering week named sampling in its stated objective and gave it one sentence, recommending random selection and immediately withdrawing it as usually impossible. That is no longer the whole of the material.
THE POSITION NOW
| Supplied | Still absent | |
|---|---|---|
| Vocabulary | Population and sample defined; six types named — purposive, random, simple random, stratified, systematic, cluster | Quota, judgement and self-selected sampling remain bare labels |
| Method | Nothing | No procedure for executing any of the six. No sampling frame is mentioned anywhere in six weeks |
| Size | The question is posed on its own slide | No rule in the teaching material's own voice. The only figures come from a set reading in another discipline |
| Choosing between types | Nothing | No criterion of any kind is given |
| Bias | Non-response, with a sound causal chain, and a cost-and-error argument | No remedy for non-response once it has happened |
Full treatment at the linked pages below.
The four sentences that carry the whole of it
The week's study notes open the topic by asking two questions — where do you get the data, and from whom should the data be collected — and then organise the answer under three sub-headings: "Who?", "How?" and "Approaches". The "Who?" subsection is short enough to quote in full, and it is quoted in full below because nothing else in the week adds to it.
Two criteria are stated, and the word "both" makes them cumulative rather than alternative. A participant who is reachable but cannot speak to your question fails the test, and so does a participant who could answer it perfectly but is out of reach.
Available to you
- Stated as a requirement, not a preference.
- No definition of availability is given: access negotiation, gatekeepers and consent pathways are not described in the week.
- The source's only elaboration is the concession that random selection "is not always possible, even if your population was quite small".
Relevant to the question you are studying
- Ties participant choice back to the research question rather than to convenience.
- No test of relevance is supplied — the criterion is stated and the passage moves on.
- The phrase "the population of people of interest" is the only population language anywhere in the week. It is never defined.
Random selection: recommended, over-claimed, then withdrawn
The middle two sentences of the passage do two different jobs. The first recommends random selection as the ideal; the second states what random selection buys you, in absolute terms.
By randomly picking people, you know that there is no way any of the results could be attributed to the biases in your selection of research participants.
Preserve the source's own hedges around that claim, because they are load-bearing: "Ideally", "you'd", and then "Of course, this is not always possible". The passage recommends a procedure, states an unqualified benefit for it, and then concedes that you will usually not be able to use it. No alternative procedure follows.
Every place sampling is touched in the week
The inventory below is complete. It is worth reading as an inventory, because the gap is easier to underestimate than to see: sampling is mentioned often enough to feel covered and is never operationalised.
EVERY APPEARANCE OF SAMPLING IN THE WEEK'S MATERIAL
| Where it appears | What it says | What it supplies you with |
|---|---|---|
| Study notes, "Who?" | Participants must be available and relevant; ideally pick randomly; often not possible | No method, no frame, no size |
| Slide on collecting primary data | Surveys "Involve entire populations or samples of populations" | A phrase only — no proportion, no procedure |
| Slide on questionnaire returns | Returned questionnaires "are often of little value as a sample" | A warning about the sample you end up with, not a way of drawing one |
| Slide on postal questionnaires | Postal use is permitted where "an appropriate sample of the non-responders is interviewed" | "Appropriate" is neither quantified nor defined |
| The six qualitative approaches | "Sample sizes for grounded theory are more limited than for ITA" | A comparison with neither term quantified |
Four of the five are incidental mentions inside guidance about something else. The fifth is a comparative claim about size that gives no size.
Do not read this page as a sampling method. It is an accurate account of teaching material that does not contain one, and inventing the missing content here would make the material look more complete than it is.
The objective that names sampling, and the audit that records it undelivered
The week states exactly one learning objective: to evaluate methods of data collection, "including: sampling, interviewing structures and techniques" and "questionnaires, observations, focus groups, surveys". Sampling is the first of the six items named.
The delivery audit for this week records sampling as not delivered, and records the same for observations and focus groups. It also records that the objective's verb is undelivered across all six items: to evaluate methods requires criteria and a comparison, and no two collection methods are ever compared with each other on any criterion in the week. The consolidated account of that pattern across the whole supplied source is on What This Material Does Not Teach.
How far the library's data collection gap now closes
This stream exists because of a recorded gap. The material behind Data Collection Methods: Data as Evidence supplied one paragraph of prose on data collection and thirteen method labels on a diagram with nothing attached to them — the library's second-largest source gap since its first release.
The material behind this stream narrows that gap substantially. It adds a ten-name method catalogue, six approaches to collecting qualitative data, survey research and survey construction, questionnaire design and wording, structured and unstructured questions, levels of measurement, four interview types, interview conduct, and recording and transcription. That is a great deal more than one paragraph.
Writing a defensible participant-selection section anyway
The procedure below is this page's construction, built on the source's two criteria and its own concessions. The source does not prescribe it. It is written so that a reader can produce an honest participant-selection section without claiming sampling properties the design does not have.
Five steps to a participant-selection section you can defend
Name the population of interest in one sentence
The source's own phrase is "the population of people of interest". Write yours as a bounded description — role, organisation type, time period, project stage — so a reader can tell who is inside it and who is not.
Apply both criteria and record both answers
For each candidate group, state its relevance to the research question and its availability to you. The source requires both; most weak sections satisfy availability silently and never argue relevance.
Say what you did, in the language of what you did
If you approached everyone you could reach, say that. If you selected on a characteristic, name the characteristic. Reserve the word random for a procedure you actually executed and can describe.
Record who you could not reach, and who did not reply
The week's own slides treat non-response as a defect that survives good selection. A section that reports only the participants obtained cannot be assessed for the bias the source warns about.
Take the sampling design itself from a cited text
Attribute the frame, the technique and any size argument to a methods text you have read, and cite it there. Do not attribute a sampling rule to this teaching material, because it contains none.
What your second source has to supply that this one does not
- A definition of the sampling frame and how to construct one from your population of interest
- The probability and non-probability families, and the conditions under which each is appropriate
- At least one named technique you can execute and describe in a methodology chapter
- A basis for the size of your sample — a rule, a convention, or an argued stopping point
- A treatment of non-response and self-selection that goes beyond a warning
- For qualitative designs, whatever your chosen approach requires in place of a size — the material here offers only an unquantified comparison
Common failures, and what to do instead
Participant selection: five situations you will actually meet
What to carry forward
- Two criteria, both required: participants must be available to you and relevant to the question you are studying.
- The source recommends random selection, over-claims what it removes, and concedes it is usually impossible — with nothing offered in its place.
- The notes and the slides disagree about whether randomisation eliminates selection bias. The disagreement is real, it is unresolved in the source, and it is left standing here.
- Sampling is named first in the week's only learning objective and the delivery audit records it as not delivered.
- This batch narrows the library's data collection gap a long way, but sampling, observations and focus groups remain named without being taught.
- Get your frame, technique and size from a cited methods text, and describe in your write-up only the procedure you actually performed.
Frequently asked questions
Does the supplied teaching source contain a sampling method I can follow?
No. It gives two selection criteria — available and relevant — and one recommendation, random selection, which it immediately withdraws as usually impracticable. No sampling frame, technique, size rule or procedure appears anywhere in the week, and sampling is recorded as undelivered in the week's own objective audit.
How large should my sample be?
The material states no number, minimum or proportion. Its only size statement is that grounded theory sample sizes are more limited than for inductive thematic analysis, with neither term quantified. Take a size and its justification from a methods text you have read, and cite it there rather than attributing it to this material.
The notes say random selection removes selection bias and the slides say returns are biased. Which is right?
The source never reconciles them and this page does not choose. The two statements are about different moments — how people are approached, and who then replies — but the material does not draw that distinction. Report both if you are asked to summarise the position, and design against the more cautious one.
Can I just use colleagues and contacts from my own projects?
The availability criterion permits it and the relevance criterion is what you have to argue. Write down which part of your research question only that group can answer, and record what the group's composition excludes. The source requires both criteria and gives no test for either, so the argument has to be yours and it has to be visible.
Why does this page not explain probability and non-probability sampling?
Because the supplied teaching material does not contain either term, and this library does not attribute definitions to a source that lacks them. Naming the gap tells you exactly what to look up; a plausible paragraph invented here would tell you that you had already covered it.
What does this stream actually add to the library's data collection coverage?
A great deal on instruments — survey research, questionnaire design and wording, question types, levels of measurement, interview types and conduct, recording and transcription. What it does not add is sampling, observations or focus groups, all three of which remain named without being taught.
References and source attribution
- 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.
- Trochim, W. M. K. 2006, Research Methods Knowledge Base — cited in the supplied source as the week's only listed reference, as a bare address with no content attached to it.
- The supplied teaching source: the week 10 slide deck on collecting data and both copies of the week 10 study notes on gathering your data, used here as the basis for every statement attributed to "the source".
Suggested questions for Ask KEVOS
- Draft a participant-selection section for my project using only the two criteria this material supplies.
- What exactly does the supplied teaching source say about sampling, and what does it leave out?
- Help me argue the relevance of each participant group to my research question.
- How should I describe a selection procedure that was not random without weakening my methodology chapter?
- What do I need to take from a second source before I can defend my sample?