PSYC10006 Chap.13 Research Design, Sampling and Ethics
Research Design, Sampling and Ethics
The research methods stream runs underneath every week two modules at a time and is examined in the same quiz as the topic modules, which is why it is worth revising as a whole rather than in the short slices the weeks deliver it in.
Its first seven modules move in a fixed order, and each step constrains the next: a question is refined into something testable, a construct is given an operational definition, a design is chosen that licenses the claim you want to make, and a sample is recruited that decides who the finding is about.
The module pages themselves are short, and nearly all the examinable content sits in their stated outcomes, which name precisely which distinctions you should be able to make.
What this chapter covers
- 01
Psychology as an evidence-based discipline, and what the scientific method contributes
- 02
Why the stream is examined alongside the topic modules rather than separately
- 03
Constructs as intangible concepts, and operational definitions as the bridge to measurement
- 04
Why the narrowing is where the danger lies, and how two studies can disagree honestly
- 05
Refining a broad question into something narrow, specific and testable
- 06
Why a statement nothing could contradict is empty rather than cautious
- 07
Four scales of measurement, and what each does and does not license
- 08
Validity and reliability as independent properties of a measure
- 09
Five study designs, ordered by the strength of claim each supports
- 10
Where quasi-experimental designs sit, and what they buy and pay for
- 11
Populations and samples, and the internal against external split in a critique
- 12
Five named ethical issues, and the national framework behind them
Refine a question until it can be tested
- 1Name the constructs hiding in the question. Music covers everything from quiet ambience to lyrics at volume, and helps could mean faster, more accurate, more enjoyable or better retained. Until both are pinned the question cannot be answered wrongly, which is the same as saying it cannot be answered.
- 1Operationalise each one: music becomes a specific condition such as instrumental music at a fixed volume, and helps becomes a specific measure such as items correctly recalled on a test a day later.
- 1Name the comparison, because a hypothesis is always implicitly one. Silence is the obvious baseline, and choosing it shapes what the result can mean.
- 1State a direction. A statement that music affects recall is compatible with any outcome; a statement that participants studying with music will recall fewer items than participants studying in silence commits to something.
- 1Check it could be wrong by describing the refuting result, which here is the music group recalling at least as many items as the silence group. If no such result can be described, the refinement is not finished.
Key terms
- Construct
- A concept that is real enough to reason about and not directly observable, such as motivation or attention. Measuring one always requires committing to a narrower definition than the concept itself.
- Operational definition
- A statement of the specific observable procedure by which a construct will be measured in a particular study. It is a choice, and forgetting that a choice was made is what turns a result about a measure into a claim about a concept.
- Interval scale
- A measurement scale with order and equal steps but no true zero, so differences can be compared while ratios cannot. Zero does not mean an absence of the thing being measured.
- Quasi-experimental design
- A comparison of groups that differ on the variable of interest without the researcher having assigned anyone to them. It buys access to variables that cannot be assigned and pays for it in interpretive strength.
- External validity
- The question of who a finding applies to beyond the study, which is decided by sampling. It is separate from whether the comparison within the study was sound.
- Beneficence
- One of the named ethical issues: whether the likely benefits of research justify its risks and whether harm has been minimised. It sits alongside merit, integrity, justice and respect in the subject's list.
Research Design, Sampling and Ethics FAQ
Why does the methods stream sit in the same quiz as the topic modules?
Because every claim in the topic modules is a result, and a result is only as strong as the design that produced it. The stream supplies the vocabulary for saying how strong that is, which is also the vocabulary the written assignment is marked in. Two of the four modules each quiz covers come from it.
How do I tell the four measurement scales apart quickly?
Two questions settle all four. Are the gaps between adjacent values equal? If not you are in categorical or ordinal, and order decides between them. If they are, does zero mean none of the thing? If not it is interval, and if it does it is ratio. That works on unfamiliar measures.
Does a large sample fix a correlational design?
No, and this is the most common way a critique goes wrong. More data narrows the uncertainty around an estimate and does nothing about the direction of causation or an unmeasured third factor, because those are properties of the design. Reaching for sample size when the objection is structural misses the point.
What is the difference between validity and reliability?
Validity asks whether a measure captures the construct it claims to; reliability asks whether it gives consistent results on repetition. They are independent: a scale reading four kilograms heavy every time is perfectly reliable and not valid, which is the example worth keeping.
How many ethical issues should I list?
The subject names five: merit, integrity, justice, beneficence and respect. The national statement itself presents these as four values, treating research merit and integrity as one. Follow the subject's five-item list if asked to enumerate, and knowing the grouping protects you if the phrasing comes from the statement.
Assessment move
Revise this stream as one sequence rather than as seven modules, because the steps constrain each other and the order is what makes them memorable. For each concept the outcomes name, write the distinction it is asking you to make rather than a definition, since the items are discriminations. Practise the scales rule on measures you invent until it is automatic.
For design, learn the ladder by what each rung licenses and rehearse saying why sample size does not move you up it. Keep the internal and external questions separate when critiquing anything, because answering the wrong one produces a plausible paragraph that does not address the item. And learn the five ethical issues as a checklist you can apply to a described study, which is how they are most likely to be tested.
Working through Research Design, Sampling and Ethics in PSYC10006? Sia is AskSia’s AI Psychology tutor — ask any PSYC10006 Research Design, Sampling and Ethics question and get a clear, step-by-step explanation grounded in how PSYC10006 is taught and assessed. Read this chapter free, then take your hardest questions to Sia.
