Control procedures | AQA A-Level Psychology Revision

For 7182 specification, first teach in September 2025

AQA A-Level Psychology | Free Revision Notes

Estimated study time: 55 minutes

Control procedures help psychologists reduce unwanted influences and make fair comparisons between conditions. This Control procedures A-Level Psychology revision page explains random allocation, counterbalancing, randomisation, standardisation and control groups. You will learn what each procedure controls, when it should be used and how to apply it to unfamiliar research scenarios. These procedures form part of the scientific processes section of AQA A-Level Psychology research methods and are essential when designing valid, reliable and replicable investigations.

Learning Objectives 🎯

By the end of this revision page, you should be able to:

• Define random allocation, counterbalancing, randomisation and standardisation.

• Explain how random allocation may reduce systematic participant differences.

• Explain how counterbalancing may control order effects.

• Distinguish random allocation from random sampling and randomisation.

• Explain how standardisation improves consistency across participants and conditions.

• Explain and apply the use of a control group in psychological research.

Revision Notes 📚

What are control procedures?

Control procedures are methods used to reduce the influence of unwanted variables and make an investigation more systematic.

In an experiment, researchers aim to determine whether changes in the independent variable produce changes in the dependent variable. Other differences between participants, conditions or procedures could affect the results and provide alternative explanations.

Control procedures help researchers:

• Reduce the influence of extraneous variables.

• Make conditions more comparable.

• Apply procedures consistently.

• Reduce systematic bias.

• Improve the reliability of measurements.

• Make stronger judgements about the effect of the independent variable.

Before selecting controls, researchers must first identify what is being changed, measured and potentially left uncontrolled. This builds on identifying and operationalising research variables.

The control procedures required for this lesson are:

1. Random allocation

2. Counterbalancing

3. Randomisation

4. Standardisation

5. Control groups

Control does not mean complete control

Psychologists cannot remove every possible influence on human behaviour. Participants differ in many ways, and unexpected events may occur.

The aim of control is therefore to:

• Identify likely unwanted influences.

• Reduce their effect where practical.

• Prevent one condition from being systematically advantaged.

• Make the procedure clear enough to repeat.

A well-controlled investigation is not one in which every person or situation is identical. It is one in which important differences have been considered and managed appropriately.

Random allocation

Random allocation is the use of a chance procedure to place participants into the different conditions of an experiment.

Each participant should have an equal chance of being allocated to any condition.

For example, in an experiment comparing revision in silence with revision while listening to music, participants could be randomly allocated to:

• Condition A: revision in silence.

• Condition B: revision with music.

Random allocation is mainly associated with an independent groups design, where different participants take part in each condition.

How random allocation can be carried out

A researcher could use the following procedure:

1. Give each participant a unique number.

2. Use a random number generator to produce an allocation sequence.

3. Allocate participants to condition A or condition B according to that sequence.

4. Continue until the required number of participants has been placed in each condition.

Another simple method would be to place equal numbers of condition labels into a container and allow each participant to select one without seeing it.

The method must genuinely depend on chance. Allocating the first ten participants to one condition and the next ten to another would not be random.

Why is random allocation used?

Participants differ in ways that may affect the dependent variable. These participant variables may include:

• Relevant ability.

• Motivation.

• Previous experience.

• Tiredness.

• Familiarity with the task.

If participants choose their own condition, or if the researcher chooses for them, one group could become systematically different from the other.

For example, participants who enjoy music might volunteer for the music condition. Their interest in the activity could influence their performance.

Random allocation reduces the likelihood that participant variables will be distributed systematically between conditions.

Worked example: random allocation

A psychologist recruits 30 participants for an experiment comparing two instructions:

• Encouraging instructions.

• Neutral instructions.

The researcher randomly allocates 15 participants to each condition.

This helps reduce the chance that one condition contains, for example, more highly motivated participants than the other.

However, random allocation does not guarantee that the two groups will be identical. Differences may still occur by chance, particularly when the groups are small.

Strengths of random allocation

It reduces researcher choice

The researcher does not decide which participant enters each condition. This reduces the possibility that expectations influence allocation.

It may distribute participant variables more evenly

Individual differences are less likely to be systematically concentrated in one condition.

It supports a fairer comparison

Where groups are broadly comparable before the manipulation, differences in the dependent variable are more likely to be connected to the independent variable.

Limitations of random allocation

The groups may still differ by chance

Random allocation reduces systematic differences but does not eliminate participant variables.

A random process could still place several high-performing participants in the same condition.

It requires more than one group

Random allocation is not normally needed in a repeated measures design because every participant completes every condition.

It may not be possible in every investigation

In a quasi-experiment, participants belong to pre-existing groups based on a characteristic that the researcher cannot allocate.

Random allocation and experimental design

The control procedure chosen should suit the design.

Experimental design	Main issue	Relevant procedure
Independent groups	Different participants may vary between conditions	Random allocation
Repeated measures	Completing one condition may affect performance in another	Counterbalancing
Matched pairs	Members of each pair must be divided between conditions	Random allocation within each pair

This relationship is explored in more detail when studying how participants are organised across conditions.

Random allocation is not random sampling

Random sampling and random allocation are separate stages of a study.

Random sampling	Random allocation
Selects participants from a target population	Places recruited participants into conditions
Aims to reduce sampling bias	Aims to reduce systematic participant differences
Occurs during recruitment	Occurs after participants have been recruited
Concerns the representativeness of the sample	Concerns the comparability of experimental groups

A researcher could use one without the other.

For example, a researcher might obtain an opportunity sample of college students and then randomly allocate those students to experimental conditions.

Counterbalancing

Counterbalancing is a procedure used to vary the order in which participants complete experimental conditions.

It is used in a repeated measures design, where the same participants complete every condition.

Counterbalancing aims to distribute order effects across the conditions.

What are order effects?

Order effects occur when a participant’s performance is influenced by the order in which conditions are completed.

These may include:

• Practice effects: performance improves because the participant becomes familiar with the task.

• Fatigue effects: performance declines because the participant becomes tired or bored.

Participants might also work out the research aim after experiencing the first condition, which could influence later behaviour.

Simple counterbalancing

Consider an experiment with two conditions:

• Condition A: completing a task in silence.

• Condition B: completing the task while listening to music.

The researcher could divide participants into two groups:

Participant group	First condition	Second condition
Group 1	A	B
Group 2	B	A

Half of the participants complete the conditions in the order AB. The other half complete them in the order BA.

If participants improve through practice, the benefit is not consistently attached to one experimental condition because each condition is completed second by some participants.

An ABBA procedure

Another form of counterbalancing is an ABBA design:

1. Condition A

2. Condition B

3. Condition B

4. Condition A

The order is then reversed within the sequence.

However, asking participants to complete conditions multiple times may increase fatigue, make the aim more obvious or require additional materials.

The researcher should therefore choose a form of counterbalancing that is practical for the investigation.

What counterbalancing achieves

Counterbalancing does not stop participants becoming practised or tired.

Instead, it aims to prevent order effects from consistently favouring one condition.

For example, if the second task is completed faster because of practice:

• Some participants receive condition A second.

• Other participants receive condition B second.

The practice advantage is spread across both conditions rather than being attached only to one.

Strengths of counterbalancing

It reduces systematic order effects

Practice and fatigue are less likely to provide a consistent alternative explanation for differences between conditions.

It allows a repeated measures design to be retained

Researchers can control participant variables by using the same participants in every condition while also reducing the effect of condition order.

It is relatively straightforward with two conditions

An AB and BA arrangement can be organised easily where there are only two conditions.

Limitations of counterbalancing

Order effects still occur

Participants may still become tired, bored or practised. Counterbalancing distributes these effects rather than removing them.

It becomes more complicated with several conditions

As the number of conditions increases, the number of possible orders also increases.

Equivalent materials may still be needed

If participants complete similar tasks more than once, they may remember earlier materials or answers.

The aim may become obvious

Experiencing every condition can reveal what the researcher is comparing, increasing the possibility of demand characteristics.

These participant-related influences are examined further in how participants and investigators can affect results.

Counterbalancing is not random allocation

The two procedures address different problems.

Counterbalancing	Random allocation
Changes the order of conditions	Assigns participants to conditions
Used mainly with repeated measures	Used mainly with independent groups
Controls systematic order effects	Reduces systematic participant differences
Each participant completes every condition	Each participant normally completes one condition

An exam question may describe participants completing conditions in different orders. This indicates counterbalancing, not random allocation.

Randomisation

Randomisation involves using chance to determine aspects of a research procedure.

It reduces the possibility that a fixed or predictable order systematically influences the results.

Randomisation is a broad control procedure. It may be used to decide:

• The order in which stimuli are presented.

• The order of questions.

• The arrangement of items within a task.

• The order of trials.

• The order in which participants are tested.

• The position of correct answers in a test.

Example of randomising stimuli

A psychologist investigates memory for a list of words.

If every participant sees the words in the same fixed order, the position of a word could affect whether it is recalled.

The researcher could randomise the order of the words for each participant. This reduces the risk that a particular word is always advantaged because it appears in the same position.

Example of randomising questions

A researcher uses a questionnaire to measure attitudes.

If sensitive questions always appear first, participants’ responses to later questions could be affected. The researcher might randomise the order of selected questions, provided the questionnaire still makes sense.

This reduces the possibility that one fixed sequence consistently influences responses.

Randomisation and random allocation

Random allocation is one specific use of a chance procedure, but the terms should not be used as though they always mean exactly the same thing.

• Random allocation refers specifically to assigning participants to conditions.

• Randomisation may apply to many parts of the procedure, such as the order of tasks, stimuli or questions.

Strengths of randomisation

It reduces systematic order bias

A particular stimulus or question is not always placed in the same position.

It reduces researcher influence

The researcher is less able to choose an order that favours an expected result.

It can strengthen procedural control

Different materials or trials are less likely to be linked consistently with one position or sequence.

Limitations of randomisation

Random sequences may still contain patterns

A random sequence can include clusters or repeated patterns by chance.

It may make procedures harder to organise

Researchers need an accurate method for generating and following the random order.

It may be unsuitable where order is meaningful

Some instructions, interview questions or tasks must follow a logical sequence. Fully randomising them could confuse participants or change the nature of the procedure.

Counterbalancing and randomisation compared

Both procedures may address order-related problems, but they work differently.

Counterbalancing	Randomisation
Deliberately varies condition order across participants	Uses chance to determine an order
Ensures different conditions appear in different positions	Does not guarantee equal use of every possible order
Commonly used in repeated measures experiments	Can be used with stimuli, questions, tasks or trials
Designed to distribute order effects	Designed to reduce systematic ordering bias

A researcher could use both. For example, the order of two experimental conditions could be counterbalanced, while the order of items within each task is randomised.

Standardisation

Standardisation means keeping the procedure consistent for all participants, apart from the planned manipulation of the independent variable.

A standardised procedure gives clear instructions about exactly how the investigation should be conducted.

This may include:

• The wording of instructions.

• The time allowed.

• The materials used.

• The testing environment.

• The order of procedural stages.

• The researcher’s responses to questions.

• The way the dependent variable is measured.

• The way responses are scored.

• The debriefing procedure.

Why is standardisation important?

Without standardisation, participants may receive different experiences.

For example, one researcher might explain a task in more detail, give extra encouragement or allow additional time. These differences could affect the dependent variable.

If only the independent variable is intended to vary, other procedural features should remain consistent.

Standardised instructions

Standardised instructions are instructions prepared in advance and delivered in the same way to each participant.

They should explain:

• What the participant must do.

• How long the task will last.

• How responses should be recorded.

• What to do if clarification is needed.

• Relevant ethical information, such as the right to withdraw.

The instructions should not provide one condition with more information or assistance than another.

Standardising the environment

Researchers may attempt to keep environmental conditions consistent, including:

• Room.

• Lighting.

• Temperature.

• Noise level.

• Seating position.

• Equipment.

• Time available.

This helps prevent environmental variables from providing alternative explanations for the findings.

Standardising measurement and scoring

A researcher must also decide how the dependent variable will be recorded.

For example, if memory is measured using word recall, the procedure should specify:

• Which words count as correct.

• Whether spelling errors are accepted.

• How repeated answers are treated.

• The maximum score.

• The recall period.

Without clear scoring criteria, researchers may judge the same response differently.

Standardisation in different research methods

Standardisation can be used across psychological research.

Experiments

Researchers use the same instructions, materials and measurement procedures across participants and conditions.

Observations

Observers use the same operationalised behavioural categories and recording schedule.

Questionnaires

All participants receive the same questions and response options in the same intended format.

Structured interviews

Interviewers use the same prepared questions, order and standardised prompts.

The degree of standardisation may be lower in an unstructured interview because the interviewer is able to respond more flexibly.

Strengths of standardisation

It improves consistency

Participants experience a similar procedure, reducing unintended differences.

It may improve reliability

A clearly specified procedure can be repeated in the same way by another researcher.

It supports replication

Other psychologists can follow the documented method and check whether similar results are obtained.

It reduces investigator effects

A fixed script and scoring system reduce opportunities for researchers to treat participants differently.

Limitations of standardisation

It cannot control every difference

Participants may still interpret the same instruction differently or respond differently to the setting.

Very rigid procedures may be less natural

A tightly controlled interaction may not reflect how people normally behave outside the study.

Flexibility may sometimes be useful

In an unstructured interview, for example, researchers may need to ask follow-up questions based on an individual participant’s answer.

Piloting a standardised procedure

A procedure may appear clear when written but fail when used with participants.

A pilot study can reveal that:

• Instructions are ambiguous.

• Equipment is difficult to operate.

• Participants require unexpected help.

• A task takes longer than planned.

• Scoring rules are incomplete.

• Researchers interpret parts of the procedure differently.

The researcher can then revise and standardise the procedure before the full investigation. This connects to testing research materials and procedures in advance.

Control groups

A control group is a comparison group that does not receive the experimental treatment or manipulation being investigated.

The control group provides a baseline against which the experimental group can be compared.

For example, a psychologist investigates whether a new revision activity improves test performance:

• Experimental group: completes the new revision activity.

• Control group: does not complete the new activity and follows the comparison procedure.

The researcher compares the performance of the two groups.

Why use a control group?

Suppose test scores improve after an intervention.

Without a control group, the researcher may not know whether improvement occurred because of:

• The intervention.

• Practice on the test.

• Increased familiarity with the procedure.

• The passage of time.

• Participants’ expectations.

• Another event occurring during the study.

A control group helps show what happens without the experimental manipulation.

A control group must be treated comparably

The control and experimental groups should experience the same general procedure, except for the independent variable.

For example, both groups should receive:

• The same amount of time.

• The same testing environment.

• The same instructions where appropriate.

• The same outcome measure.

• The same scoring procedure.

If the experimental group spends 30 minutes completing a revision activity while the control group immediately takes the test, time and attention may differ as well as the revision activity.

The researcher would then have more than one possible explanation for the result.

Worked example: control group

A psychologist investigates whether ten minutes of a breathing exercise reduces errors on an attention task.

Group	Procedure
Experimental group	Completes a ten-minute breathing exercise
Control group	Sits quietly for ten minutes
Both groups	Complete the same attention task under the same conditions

The quiet-sitting group provides a comparison for the effect of spending ten minutes before the task.

If the control group completed no waiting period, the groups would differ in both activity and timing.

Control groups and independent variables

A control group is often one condition of the independent variable.

For example:

• IV: type of pre-task activity.

• Experimental condition: breathing exercise.

• Control condition: quiet sitting.

• DV: number of errors on the attention task.

The control group does not mean that no controls are used elsewhere. The researcher must still standardise instructions, control extraneous variables and allocate participants appropriately.

Strengths of control groups

They provide a baseline

The researcher can compare the experimental outcome with what happens without the target manipulation.

They help identify alternative explanations

Changes that also appear in the control group may not be caused by the experimental treatment.

They support clearer interpretation

A difference between otherwise comparable groups strengthens the case that the independent variable contributed to the result.

Limitations of control groups

The groups may differ before the experiment

Participant variables may affect the results unless allocation and other procedures are appropriate.

A poorly designed control condition may create extra differences

The control and experimental groups must differ only in the intended manipulation as far as practical.

A control group does not prove causation by itself

The researcher must also control extraneous variables, measure the dependent variable accurately and use a suitable experimental method.

Control group and control condition

The terms are related but not always identical.

• A control group refers to the participants who receive the comparison procedure.

• A control condition refers to the comparison situation itself.

In an independent groups design, a separate group of participants completes the control condition.

In a repeated measures design, the same participants may complete both the experimental condition and the control condition. In that case, there is a control condition but not a separate control group.

Combining control procedures

Researchers often use several control procedures together.

Consider an experiment investigating whether background noise affects reading comprehension.

The researcher could:

1. Randomly allocate participants to the quiet or noise condition.

2. Randomise the order of comprehension questions.

3. Standardise the instructions, reading time and scoring.

4. Use the quiet condition as the control group.

5. Keep lighting, room temperature and materials consistent.

Each procedure addresses a different potential problem.

Control procedure and the problem it addresses

Control procedure	Main problem addressed	Example
Random allocation	Systematic participant differences	Using chance to place participants in two conditions
Counterbalancing	Order effects	Half complete AB and half complete BA
Randomisation	Fixed-order bias	Randomising the presentation order of words
Standardisation	Procedural inconsistency	Reading the same instructions to all participants
Control group	Lack of a baseline comparison	Comparing an intervention group with a group that does not receive it

Choosing a suitable control procedure

A strong research-methods answer should match the control to the problem.

Problem: groups may differ in ability

Use random allocation to reduce systematic participant differences.

Problem: all participants complete condition A before condition B

Use counterbalancing so that some complete AB and others complete BA.

Problem: the same stimulus always appears first

Use randomisation to vary stimulus order.

Problem: researchers explain the task differently

Use standardised instructions or a prepared script.

Problem: there is no comparison showing what would happen without the intervention

Include a control group or control condition.

Control procedures A-Level Psychology revision: applying knowledge

Application questions frequently describe a weakness in a research procedure and ask you to suggest an improvement.

A strong response should contain:

1. The problem identified in the scenario.

2. The appropriate control procedure.

3. A clear description of how it would be used.

4. An explanation of why it would improve the investigation.

Example

All participants completed the silent condition before the music condition.

A strong answer would be:

The researcher should counterbalance the order of conditions. Half of the participants should complete the silent condition followed by the music condition, while the other half complete the music condition followed by the silent condition. This would distribute practice or fatigue effects across both conditions.

A weaker answer would be:

The researcher should randomise it.

The weaker response does not clearly identify what would be varied or how the procedure controls the problem.

Developing a controlled investigation

When planning an investigation, a researcher should:

1. Identify the independent and dependent variables.

2. Identify likely extraneous variables.

3. Choose an appropriate experimental design.

4. Decide whether participants require random allocation.

5. Decide whether counterbalancing is needed.

6. Randomise relevant stimuli, trials or materials.

7. Prepare standardised instructions and scoring rules.

8. Include a suitable control group or control condition where needed.

9. Pilot the procedure.

10. Revise any controls that do not work as intended.

These decisions form part of developing a complete psychological investigation.

When the study is carried out, researchers must then follow the agreed procedure consistently.

Key Words 🔑

Key word	Student-friendly definition	How it may be used in an exam
Control procedure	A method used to reduce unwanted influences and make conditions more comparable.	Suggest an improvement to a research scenario.
Random allocation	Using chance to assign participants to experimental conditions.	Explain how participant differences may be reduced.
Participant variable	An individual difference that may affect the dependent variable.	Explain why random allocation may be needed.
Counterbalancing	Varying the order of conditions across participants.	Suggest a way of controlling order effects.
Order effect	A change in performance caused by the position of a condition in a sequence.	Explain a weakness of a repeated measures procedure.
Practice effect	Improved performance caused by completing an earlier task or condition.	Identify why later scores may be higher.
Fatigue effect	Reduced performance caused by tiredness, boredom or loss of motivation.	Identify why later scores may be lower.
Randomisation	Using chance to determine an aspect of the research procedure.	Explain how stimulus, trial or question order could be controlled.
Standardisation	Keeping the procedure consistent across participants and conditions.	Suggest the use of fixed instructions, timing or scoring.
Standardised instructions	Instructions prepared in advance and delivered consistently.	Explain how investigator effects may be reduced.
Control group	A comparison group that does not receive the experimental manipulation being investigated.	Explain how a baseline comparison could be created.
Control condition	The comparison situation against which the experimental condition is assessed.	Identify the baseline condition in an experiment.
Replication	Repeating a study using the same procedure.	Explain why a clearly standardised method is useful.

Common Mistakes ⚠️

Mistake: Confusing random allocation with random sampling.

Why this is incorrect:Random sampling selects participants from a population. Random allocation places recruited participants into experimental conditions.

How to improve:Identify the stage of the investigation. Recruitment concerns sampling, while assignment to conditions concerns allocation.

Mistake: Saying random allocation removes all participant variables.

Why this is incorrect:Groups may still differ by chance, particularly when the sample is small.

How to improve:State that random allocation reduces the likelihood of systematic participant differences.

Mistake: Suggesting counterbalancing for an independent groups design.

Why this is incorrect:Participants in an independent groups design complete only one condition, so there is no condition order for each participant.

How to improve:Use counterbalancing with repeated measures and random allocation with independent groups.

Mistake: Claiming that counterbalancing removes practice and fatigue effects.

Why this is incorrect:Participants may still experience these effects. Counterbalancing distributes them across conditions.

How to improve:Use phrases such as “reduces systematic order effects” rather than “eliminates order effects”.

Mistake: Using random allocation and randomisation as interchangeable terms in every situation.

Why this is incorrect:Random allocation specifically concerns assigning participants to conditions. Randomisation may concern the order of materials, trials or questions.

How to improve:State exactly what is being determined by chance.

Mistake: Describing standardisation only as using the same room.

Why this is incorrect:Standardisation can apply to instructions, materials, timing, researcher behaviour, measurement and scoring.

How to improve:Identify the exact aspect of the procedure that must remain consistent.

Mistake: Calling every comparison condition a separate control group.

Why this is incorrect:In a repeated measures design, the same participants may complete both the experimental and control conditions.

How to improve:Distinguish between a group of participants and the condition they complete.

Mistake: Creating a control group that differs in several ways from the experimental group.

Why this is incorrect:Any additional differences could affect the dependent variable and provide alternative explanations.

How to improve:Keep the procedure comparable and vary only the intended independent variable as far as practical.

Mistake: Naming a control procedure without explaining how it would be used.

Why this is incorrect:Application questions require a procedure tailored to the scenario.

How to improve:Describe the practical steps, such as placing participant numbers into a random number generator or using AB and BA condition orders.

Exam-Style Questions ✍️

Question 1

Define random allocation. (1 mark)

Question 2

State one difference between random allocation and random sampling. (2 marks)

Question 3

Explain how standardised instructions may improve a psychological investigation. (2 marks)

Question 4

A psychologist investigates performance in silence and while listening to music. Every participant completes the silent condition first.

Explain how the psychologist could use counterbalancing in this investigation. (3 marks)

Question 5

A researcher presents the same ten images to every participant in the same order. The researcher is concerned that the position of an image may affect whether it is remembered.

Explain how randomisation could be used to improve the procedure. (3 marks)

Question 6

A psychologist uses an independent groups design to compare two revision activities. Participants are allowed to choose which activity they complete.

Explain one problem with this allocation procedure and suggest an improvement. (4 marks)

Question 7

A researcher investigates whether a ten-minute breathing activity affects performance on an attention task.

The experimental group completes the breathing activity and then takes the test immediately. The control group waits for two minutes in a noisy corridor before taking the test.

Explain two problems with the control procedure and suggest one improvement for each problem. (6 marks)

Question 8

A psychologist investigates whether encouraging feedback affects persistence on a puzzle.

Different researchers test the two conditions. One researcher reads detailed instructions and allows participants 15 minutes. The other gives brief instructions and allows participants 10 minutes.

Explain how standardisation could be used to improve this investigation. (6 marks)

Question 9

A researcher investigates whether screen brightness affects reading accuracy. The same participants complete a low-brightness condition and a high-brightness condition.

Design suitable control procedures for this investigation. In your answer, refer to:

• Counterbalancing

• Randomisation

• Standardisation

• A control condition

(8 marks)

Question 10

A psychologist plans to investigate whether a new revision strategy improves test performance.

Explain how the psychologist could use random allocation, a control group and standardisation to conduct a controlled experiment. Justify the use of each procedure. (8 marks)