Why is "Writing an aim when asked for a hypothesis" a common mistake in Methodological Concepts?

Why it happens: The two sound similar. How to avoid it: An aim states a purpose ('to investigate…'); a hypothesis is a PREDICTION ('participants will…') and must be operationalised. Source: 9990 Paper 2 mark schemes

Why is "Writing a vague, non-operationalised hypothesis" a common mistake in Methodological Concepts?

Why it happens: Students keep it general. How to avoid it: Operationalise both variables: define exactly what is manipulated and how the DV is measured (e.g. 'words recalled out of 20').

Why is "Treating extraneous and confounding variables as the same" a common mistake in Methodological Concepts?

Why it happens: Both are 'nuisance' variables. How to avoid it: Extraneous = COULD affect the DV; confounding = systematically varies WITH the IV and so DOES distort the result.

Why is "Calling 'whoever was available' a random sample" a common mistake in Methodological Concepts?

Why it happens: 'Random' is used loosely in everyday speech. How to avoid it: Random = every member of the population has an EQUAL chance of selection. 'Whoever was around' is an OPPORTUNITY sample.

Why is "Always recommending the mean" a common mistake in Methodological Concepts?

Why it happens: The mean is the most familiar average. How to avoid it: With extreme outliers the mean is distorted — the MEDIAN is more representative. Match the statistic to the data.

Research Methods (AS Level)Cambridge International A Level Psychology 9990

Methodological Concepts

Work through the notes, try the practice questions, then take the quiz. The report tells you exactly what to revise next.

PreviousNext

Learn this Topic

Start with the slides for the quick version, then go deeper with the full study notes.

Short Study Notes in the form of Slides

Read the notes first. If the method in a worked example clicks, you're ready for the questions.

Short Study Notes — Methodological Concepts

Start with these resources to cover the key concepts, then work through the practice questions.

Page 1 / 0

Detailed Notes

Full prose, callouts and a recap — built for A* mastery, not just a quick scan.

Take these study notes with you

Download a branded PDF — full prose, callouts, recap and memorise list for Methodological Concepts, ready to print or save offline.

Methodological Concepts — Cambridge AS & A Level Psychology 9990 Study Notes (2024-2026)

The toolkit that lets you plan, run and judge a study for Paper 2: aims and hypotheses, variables and controls, experimental designs, sampling, data types, reliability and validity, ethics, and descriptive statistics. This note builds each concept with worked phrasing and the exact wording examiners reward.

At a glance

Aim = the purpose; hypothesis = a testable prediction (alternative/experimental vs null; directional vs non-directional).
Operationalisation = defining variables so they can be precisely measured (the IV and DV in particular).
Variables: independent (IV), dependent (DV), and extraneous/confounding variables you must control.
Experimental designs: independent measures, repeated measures, matched pairs (each with a key strength and weakness).
Controls: standardisation, randomisation and counterbalancing (to reduce bias and order effects).
Sampling: target population → sample; opportunity, volunteer (self-selected), random; watch for bias and generalisability.
Reliability = consistency; validity = measuring what you intend (incl. ecological validity).
Descriptive statistics: mean, median, mode (central tendency); range and standard deviation (spread); plus graphs.

What you’ll learn

Mapped to the Cambridge International A Level 9990 syllabus (2024-2026).

Write aims and hypotheses, distinguishing null vs alternative and directional vs non-directional, and operationalise variables.
Identify and control independent, dependent, extraneous and confounding variables (standardisation, randomisation, counterbalancing).
Describe and evaluate the three experimental designs (independent measures, repeated measures, matched pairs).
Describe and evaluate sampling techniques (opportunity, volunteer, random) and discuss generalisability and bias.
Explain reliability, validity and data types, apply ethical guidelines, and calculate/interpret descriptive statistics.

Aims and hypotheses (background substance)

The aim states the purpose; the hypothesis is a testable, operationalised prediction (null vs alternative; directional vs non-directional).

Getting hypotheses right is a guaranteed Paper 2 skill.

Aim = a general statement of the purpose of the study (what it is trying to find out). Example: 'to investigate whether caffeine affects reaction time.'

Hypothesis = a precise, testable prediction. Two kinds:

Alternative (experimental) hypothesis — predicts there is an effect/relationship. Example: 'Participants who drink caffeine will have faster reaction times than those who do not.'
Null hypothesis — predicts there is no effect/relationship; any difference is due to chance. Example: 'There will be no difference in reaction time between participants who drink caffeine and those who do not.'

Directional vs non-directional:

Directional (one-tailed) — states the direction of the effect ('faster', 'more', 'higher'). Use when previous research suggests a direction.
Non-directional (two-tailed) — predicts a difference but not the direction ('will differ'). Use when there is no clear prior evidence.

Operationalisation. A good hypothesis is operationalised — variables are defined so they can be measured. 'Caffeine' becomes '200mg of caffeine'; 'reaction time' becomes 'time in milliseconds to press a key'.

Why this matters. Marks are lost for vague hypotheses. Write them as testable, operationalised statements naming both the IV and the DV.

Aim = purpose; hypothesis = testable, operationalised prediction.
Alternative/experimental = there IS an effect; null = NO effect (chance).
Directional (one-tailed) names the direction; non-directional (two-tailed) just predicts a difference.
Operationalise: define the IV and DV so they can be measured.

Variables and controls

IV is changed, DV is measured; extraneous/confounding variables must be controlled by standardisation, randomisation and counterbalancing.

Independent variable (IV) = the variable the researcher manipulates/changes. Dependent variable (DV) = the variable that is measured (the outcome).

Extraneous variables (EVs) = 'nuisance' variables that could affect the DV if not controlled (e.g. noise, time of day, individual differences). Confounding variables = an extraneous variable that systematically varies with the IV and so does spoil the result — you can no longer tell whether the IV or the confound caused the change.

Controls — how to deal with them:

Standardisation — keep the procedure identical for all participants (same instructions, materials, timing) → reduces EVs and aids replication.
Randomisation — use chance to decide order/allocation (e.g. randomly ordering word lists) → reduces researcher bias.
Counterbalancing — in repeated measures, vary the order of conditions across participants (half do A→B, half B→A) → controls order effects (practice/fatigue).

Why this matters. Controlling variables is what lets an experiment claim cause and effect. In a 'suggest an improvement' question, naming a specific control (e.g. 'standardise the instructions') in context scores well.

IV = manipulated; DV = measured.
Extraneous variable = could affect the DV; confounding variable = systematically varies with the IV (does spoil it).
Standardisation = identical procedure for all (aids replication).
Randomisation = use chance (reduces bias); counterbalancing = vary condition order (controls order effects).

Experimental designs

Independent measures, repeated measures and matched pairs — each with a defining strength and weakness.

There are three ways to allocate participants to conditions.

1. Independent measures — different participants in each condition.

+ No order effects; participants can't guess the aim from doing both conditions.
− Participant variables differ between groups (individual differences); needs more participants.

2. Repeated measures — the same participants do all conditions.

+ Controls participant variables (same people); needs fewer participants.
− Order effects (practice/fatigue); participants may guess the aim (demand characteristics). Fix: counterbalancing.

3. Matched pairs — different participants, but matched on key variables (e.g. age, IQ), one of each pair per condition.

+ Reduces participant variables without order effects.
− Matching is difficult/time-consuming; can never match perfectly.

Example: Bandura used a matched-pairs design (children matched on pre-existing aggression).

Choosing: match the design to the study — e.g. use repeated measures (with counterbalancing) when individual differences would otherwise swamp a small effect.

Independent measures: different people per condition (+ no order effects; − participant variables).
Repeated measures: same people in all conditions (+ controls participant variables; − order effects).
Matched pairs: different but matched people (+ reduces participant variables, no order effects; − hard to match).
Counterbalancing fixes order effects in repeated measures.

Sampling and generalisability

A sample is drawn from a target population; opportunity, volunteer and random sampling differ in bias and representativeness.

The target population is the whole group the researcher is interested in; a sample is the smaller group actually studied. We want a sample that is representative so findings generalise.

Sampling techniques:

Opportunity sample — use whoever is available.
- + Quick and easy. − Often biased/unrepresentative (e.g. all from one place).
Volunteer (self-selected) sample — people respond to an advert.
- + Easy; gets motivated participants. − Volunteer bias (volunteers may differ from the general population). Example: Milgram's newspaper advert.
Random sample — everyone in the population has an equal chance of selection (e.g. names drawn/numbered).
- + Least biased, most representative. − Time-consuming; need a full list of the population; chosen people may decline.

Bias and generalisability. A biased sample (e.g. all male, all one culture, all students) limits how far results generalise. Watch for androcentric (male-only), ethnocentric (one-culture) and age-limited samples — all common evaluation points.

Target population → representative sample → generalisable findings.
Opportunity: + quick; − biased.
Volunteer (self-selected): + easy/motivated; − volunteer bias (e.g. Milgram).
Random: + least biased; − time-consuming, needs full list.
Watch for androcentric / ethnocentric / age-limited samples (generalisability).

Data types, reliability and validity

Quantitative vs qualitative (and primary/secondary, objective/subjective); reliability = consistency; validity = measuring what you intend.

Types of data:

Quantitative (numbers) vs qualitative (words/descriptions).
Primary (collected first-hand by the researcher) vs secondary (already collected by someone else).
Objective (bias-free, e.g. a timer) vs subjective (opinion-based, e.g. a rating).

Reliability = consistency. A reliable measure gives the same result if repeated.

Internal reliability — consistency within a test (e.g. all items measure the same thing).
External reliability — consistency over time (test–retest) or between observers (inter-observer reliability).
Improve with standardisation and clear behavioural categories.

Validity = measuring what you intend to measure (truth/accuracy).

Internal validity — the effect is genuinely due to the IV, not a confound.
Ecological validity — findings generalise to real-life settings/behaviour.
Threats: demand characteristics, social desirability, confounding variables.

Key contrast: a study can be reliable (consistent) but not valid (consistently measuring the wrong thing) — so you need both.

Quantitative vs qualitative; primary vs secondary; objective vs subjective.
Reliability = consistency (internal; external = test–retest / inter-observer).
Validity = measuring what you intend (internal; ecological = real life).
Threats to validity: demand characteristics, social desirability, confounds.
Reliable ≠ valid: you can be consistently wrong.

Ethics and descriptive statistics

Ethical guidelines (consent, deception, withdrawal, protection, confidentiality, debrief); central tendency, spread and graphs.

Ethical guidelines protect participants. Learn the core set:

Informed consent — agree to take part knowing what's involved (special care with children).
Deception — should be avoided; if used, must be justified and debriefed.
Right to withdraw — participants can leave at any time (and withdraw their data).
Protection from harm — no more risk than everyday life (physical or psychological).
Confidentiality & privacy — keep data anonymous; don't observe people where they expect privacy.
Debriefing — after the study, explain the true aim and restore wellbeing. (Milgram and Piliavin are classic ethics discussion studies.)

Descriptive statistics summarise data.

Measures of central tendency: mean (average — uses all data, but distorted by outliers), median (middle value — resistant to outliers), mode (most frequent — good for categories).
Measures of spread: range (highest − lowest — simple, but affected by outliers) and standard deviation (average distance from the mean — uses all data, more informative).
Graphs: bar charts (categories/discrete data), histograms (continuous data), scatter graphs (correlations). Always label axes and title the graph.

Why this matters. Paper 2 routinely asks you to choose and justify a statistic, calculate one, or interpret/sketch a graph — quick, reliable marks.

Ethics: informed consent, (avoid) deception, right to withdraw, protection from harm, confidentiality/privacy, debrief.
Central tendency: mean (uses all data, outlier-sensitive), median (middle, robust), mode (most frequent).
Spread: range (simple, outlier-sensitive), standard deviation (uses all data, informative).
Graphs: bar chart (categories), histogram (continuous), scatter graph (correlation) — label axes.

How it’s examined

Examined mainly on Paper 2 (Research Methods), including the planning/design question. Common tasks: write an operationalised (and correctly typed) hypothesis; identify the IV/DV and a confounding variable; name and evaluate the experimental design; identify/evaluate the sampling technique; suggest controls and ethical safeguards; choose, justify, calculate or interpret a descriptive statistic or graph. Examiner guidance: operationalise variables, justify methodological choices, and apply ethics to the specific scenario.

Sources: Cambridge International AS & A Level Psychology 9990 syllabus (2024-2026); 9990 Learner Guide (for examination from 2024); 9990 Scheme of Work (for examination from 2024); 9990 Specimen / past Paper 2 and mark schemes. Last reviewed 2026-05-21.

Master this Topic

Worked examples, formulae, definitions and the mistakes examiners flag — everything you need to push from a pass to an A*.

Take this whole topic with you

Download a branded revision sheet — worked examples, formulae, definitions and common mistakes for Methodological Concepts, ready to print or save as PDF.

Step-by-step worked examples — Methodological Concepts

Step-by-step solutions to past-paper-style questions on methodological concepts, written exactly the way a tutor would explain them at the board.

1Write a null hypothesis (2 marks)
Getting started• hypotheses, AO2
Question
A study compares recall of words in silence vs with music. Write a suitable null hypothesis. (2 marks)
Step-by-step solution
1. Step 1
  A null hypothesis predicts NO difference (any difference is due to chance).
2. Step 2
  Operationalise: condition (silence vs music) and the DV (number of words recalled).
Answer
There will be no difference in the number of words recalled between the silence and music conditions; any difference is due to chance. (2)
Examiner tip
A null hypothesis must say there is NO effect and be operationalised. 'There is no difference' alone is too vague.
2Identify the experimental design (2 marks)
Getting started• designs, AO1
Question
The same participants complete a memory test once after coffee and once after water. Identify the experimental design used. (2 marks)
Step-by-step solution
1. Step 1
  The SAME participants do both conditions.
2. Step 2
  That is a repeated-measures design.
Answer
Repeated-measures design (same participants in both conditions). (2)
Examiner tip
Same people = repeated measures; different people = independent measures; matched people = matched pairs.
3Identify sampling + one weakness (4 marks)
Building confidence• sampling, AO2
Question
A researcher puts up a poster asking for participants. Identify the sampling technique and explain one weakness. (4 marks)
Step-by-step solution
1. Step 1
  Technique: volunteer (self-selected) sampling — people respond to an advert.
2. Step 2
  Weakness: volunteer bias.
3. Step 3
  Volunteers may share characteristics (e.g. more motivated/cooperative) that differ from the general population.
4. Step 4
  So the sample is unrepresentative, limiting how far results generalise.
Answer
Volunteer (self-selected) sampling; weakness = volunteer bias → unrepresentative sample → poor generalisability. (4)
Examiner tip
Name the technique precisely AND explain the weakness's effect (generalisability), not just 'it is biased'.
4Choose a measure of central tendency (4 marks)
Building confidence• statistics, AO2
Question
A data set is: 4, 5, 5, 6, 40. State the most appropriate measure of central tendency and justify your choice. (4 marks)
Step-by-step solution
1. Step 1
  There is an extreme outlier (40).
2. Step 2
  The mean would be distorted upwards by the outlier (it would be 12).
3. Step 3
  The median is resistant to outliers.
4. Step 4
  The median here is 5, which better represents the typical value.
Answer
Median (= 5). The outlier (40) would distort the mean (=12), so the median is more representative. (4)
Examiner tip
Justify by the data: with extreme outliers the MEDIAN beats the mean. Showing the actual values (mean 12 vs median 5) strengthens it.
5Identify a confound and control it (6 marks)
Stretch• variables, controls, AO2, Paper 2
Question
In a study, the 'caffeine' group is tested in the morning and the 'no caffeine' group in the evening. Identify a confounding variable and explain how to control it. (6 marks)
Step-by-step solution
1. Step 1
  Time of day is a confounding variable: it varies systematically with the IV (caffeine = morning, no caffeine = evening).
2. Step 2
  So any difference in performance could be due to time of day, not caffeine.
3. Step 3
  This makes the cause-and-effect conclusion invalid.
4. Step 4
  Control: test both groups at the SAME time of day (standardisation).
5. Step 5
  Alternatively, counterbalance/randomise testing times across both groups.
6. Step 6
  Then time of day no longer varies with the IV, so any difference can be attributed to caffeine.
Answer
Confound = time of day (varies with the IV). Control by testing both groups at the same time (standardisation) or randomising/counterbalancing times. (6)
Examiner tip
Top band: explain WHY it is confounding (it varies WITH the IV) and give a specific control that removes the systematic link.
6Apply an ethical guideline (4 marks)
Stretch• ethics, AO2, Paper 2
Question
A study involves mild deception about its true aim. Explain how the researcher could deal with this ethically. (4 marks)
Step-by-step solution
1. Step 1
  Deception means participants cannot give fully informed consent.
2. Step 2
  If deception is necessary, it must be justified and kept to a minimum.
3. Step 3
  The researcher should fully debrief participants afterwards, explaining the true aim.
4. Step 4
  Participants should then be offered the right to withdraw their data, restoring their wellbeing.
Answer
Justify minimal deception, then fully debrief (explain the true aim) and offer the right to withdraw data. (4)
Examiner tip
Apply ethics to the SCENARIO: name the issue (deception/consent) and the safeguard (debrief, withdraw data).

Model Answers — Methodological Concepts

High-scoring sample answers for methodological concepts on the Cambridge International A Level 9990 paper, with examiner-style notes mapping each response to the mark scheme and assessment objectives.

Question 1
9990 Paper 2 (describe) style6 marks
Q (6 marks). Describe the three experimental designs and give one strength of each.
Model answer
In an independent-measures design, different participants take part in each condition. A strength is that there are no order effects, and participants cannot guess the aim from experiencing both conditions.

In a repeated-measures design, the same participants take part in all conditions. A strength is that participant variables are controlled, because the same people are compared with themselves, and fewer participants are needed.

In a matched-pairs design, different participants are used but they are matched on key variables (such as age or IQ), with one of each pair in each condition. A strength is that it reduces participant variables (like repeated measures) without order effects (like independent measures) — combining the benefits of both. For example, Bandura matched children on pre-existing aggression.
Why this scores
Why this scores well. All three designs are correctly defined with a genuine, distinct strength each, and the matched-pairs point shows it combines the other two designs' advantages — precise AO1.
Question 2
9990 Paper 2 (evaluate) style8 marks
Q (8 marks). Evaluate sampling techniques in psychological research, with reference to generalisability.
Model answer
Opportunity sampling uses whoever is available. Its strength is that it is quick and convenient, but its weakness is that it is often biased and unrepresentative (for example, all participants from one place or class), which limits generalisability to the wider population.

Volunteer (self-selected) sampling recruits people who respond to an advert (as in Milgram's study). It is easy and gathers motivated participants, but it suffers from volunteer bias: volunteers may be more cooperative or curious than average, so again the sample may not represent the target population.

Random sampling gives everyone in the population an equal chance of selection. This is the least biased and most representative technique, supporting generalisability, but it is time-consuming, requires a complete list of the population, and selected people may still decline to take part.

Conclusion. No technique is perfect: opportunity and volunteer sampling are practical but risk bias, while random sampling is fairer but harder to carry out. Because representativeness drives generalisability, researchers should use random sampling where feasible and otherwise acknowledge the bias in their sample when interpreting results.
Why this scores
Why this reaches the top band. Each technique is evaluated for bias AND generalisability with an example, and the conclusion weighs practicality against representativeness (AO3).
Question 3
9990 Paper 2 (describe) style6 marks
Q (6 marks). Describe the main ethical guidelines that protect participants in psychological research.
Model answer
Informed consent means participants agree to take part knowing what the study involves; with children, consent is obtained from parents/guardians. Deception should be avoided; if it is necessary for valid results, it must be justified and followed by a debrief. The right to withdraw means participants can leave the study at any time, and withdraw their data, without penalty.

Protection from harm requires that participants are exposed to no more physical or psychological risk than in everyday life. Confidentiality and privacy mean personal data are kept anonymous and people are not observed where they expect privacy. Finally, debriefing at the end restores participants' wellbeing and explains the true aim, especially where deception was used. Studies such as Milgram's are criticised on several of these guidelines (deception, consent, protection from harm and the right to withdraw).
Why this scores
Why this scores well. It covers the full set of guidelines accurately and ties them to a real study (Milgram), showing applied understanding rather than a bare list.
Question 4
9990 Paper 2 (discussion) style8 marks
Q (8 marks). 'Quantitative data are better than qualitative data in psychology.' Discuss how far you agree.
Model answer
Support for the claim. Quantitative data (numbers) are objective and easy to analyse and compare — they can be summarised with descriptive statistics, displayed in graphs and tested for patterns. This makes results reliable and replicable, which is why controlled experiments favour them.

Against the claim. Quantitative data can lack depth and miss the reasons behind behaviour. Qualitative data (words/descriptions) capture rich, detailed accounts of why people think or act as they do — for example, the detail gained from an unstructured interview or a case study such as Saavedra & Silverman. Reducing complex experiences to numbers can lose meaning, so qualitative data can be more valid for some questions, even if they are harder to analyse objectively.

Conclusion. Neither type is simply 'better': quantitative data are stronger for objective comparison and reliability, while qualitative data are stronger for depth and understanding. The most useful approach depends on the research question, and many studies combine both (for example, via triangulation) to gain comparison and depth.
Why this scores
Why this reaches the top band. It argues both sides with specific reasoning (objectivity/reliability vs depth/validity), uses examples, and reaches a measured 'depends on the question' conclusion (AO3).

Key Definitions and Keywords — Methodological Concepts

Definitions to memorise and the exact keywords mark schemes credit for methodological concepts answers — sharpened from recent examiner reports for the 2026 Cambridge International A Level 9990 sitting.

Aim
Examiner keyword
A general statement of the purpose of a study — what it is trying to find out.
Null hypothesis
Examiner keyword
A testable prediction that there is no effect or relationship; any difference is due to chance.
Alternative (experimental) hypothesis
Examiner keyword
A testable prediction that there is an effect or relationship between the variables.
Directional vs non-directional hypothesis
Examiner keyword
Directional (one-tailed) states the direction of the effect; non-directional (two-tailed) predicts a difference but not the direction.
Operationalisation
Examiner keyword
Defining variables precisely so they can be measured (e.g. 'memory' as 'number of words recalled').
Confounding variable
Examiner keyword
An extraneous variable that varies systematically with the IV, so it — not the IV — may explain the results.
Counterbalancing
Examiner keyword
Varying the order of conditions across participants (half A→B, half B→A) to control order effects in repeated measures.
Reliability
Examiner keyword
The consistency of a measure — whether it gives the same result if repeated (e.g. test–retest, inter-observer).
Validity
Examiner keyword
Whether a study measures what it intends to measure (includes internal and ecological validity).
Measures of central tendency
Examiner keyword
Mean (average), median (middle value) and mode (most frequent) — ways of summarising the 'typical' score.

Common Mistakes and Misconceptions — Methodological Concepts

The traps other students keep falling into on methodological concepts questions — taken from recent Cambridge International A Level 9990 examiner reports and mark schemes — and how to avoid them.

✕Writing an aim when asked for a hypothesis
9990 Paper 2 mark schemes
Why it happens
The two sound similar.
How to avoid it
An aim states a purpose ('to investigate…'); a hypothesis is a PREDICTION ('participants will…') and must be operationalised.
✕Writing a vague, non-operationalised hypothesis
Why it happens
Students keep it general.
How to avoid it
Operationalise both variables: define exactly what is manipulated and how the DV is measured (e.g. 'words recalled out of 20').
✕Treating extraneous and confounding variables as the same
Why it happens
Both are 'nuisance' variables.
How to avoid it
Extraneous = COULD affect the DV; confounding = systematically varies WITH the IV and so DOES distort the result.
✕Calling 'whoever was available' a random sample
Why it happens
'Random' is used loosely in everyday speech.
How to avoid it
Random = every member of the population has an EQUAL chance of selection. 'Whoever was around' is an OPPORTUNITY sample.
✕Always recommending the mean
Why it happens
The mean is the most familiar average.
How to avoid it
With extreme outliers the mean is distorted — the MEDIAN is more representative. Match the statistic to the data.

Practice questions

Exam-style questions with step-by-step worked solutions. Try one before checking the method.

Past paper style quiz

Get a report showing which sub-topics you've nailed and which ones still need work.

Methodological Concepts

Short Study Notes in the form of Slides

Short Study Notes — Methodological Concepts

Detailed Notes

What you’ll learn

How it’s examined

1Write a null hypothesis (2 marks)

2Identify the experimental design (2 marks)

3Identify sampling + one weakness (4 marks)

4Choose a measure of central tendency (4 marks)

5Identify a confound and control it (6 marks)

6Apply an ethical guideline (4 marks)

Aim

Null hypothesis

Alternative (experimental) hypothesis

Directional vs non-directional hypothesis

Operationalisation

Confounding variable

Counterbalancing

Reliability

Validity

Measures of central tendency

✕Writing an aim when asked for a hypothesis

✕Writing a vague, non-operationalised hypothesis

✕Treating extraneous and confounding variables as the same

✕Calling 'whoever was available' a random sample

✕Always recommending the mean

Practice questions

Past paper style quiz

Assess your understanding