Populations and samples | AQA A-Level Psychology Revision

For 7182 specification, first teach in September 2025

AQA A-Level Psychology | Free Revision Notes

Estimated study time: 45 minutes

These Populations and samples A-Level Psychology revision notes explain how researchers select participants and decide who their findings may apply to. You will learn to distinguish a target population from a sample, explain why psychologists rarely investigate an entire population and assess whether a sample is representative. This lesson follows turning research questions into testable predictions and introduces the principles needed to understand random, systematic, stratified, opportunity and volunteer sampling.

The AQA specification requires knowledge of the difference between a population and a sample, sampling bias and the implications of sampling for generalisation. These ideas form the basis of the sampling lessons in the research methods plan.

Learning Objectives 🎯

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

• Define a target population and a sample.

• Distinguish the population from the sample in an unfamiliar investigation.

• Explain why psychologists usually study samples rather than whole populations.

• Explain what makes a sample representative or biased.

• Analyse how the composition and selection of a sample affect generalisation.

• Evaluate conclusions drawn from samples in research scenarios.

Revision Notes 📚

Populations and samples A-Level Psychology revision overview

Psychologists usually want to draw conclusions about a broad group of people.

However, they rarely collect data from every person in that group.

Instead, the researcher selects a smaller group to take part.

The main terms are:

• Target population: the complete group about which the researcher wants to draw conclusions.

• Sample: the smaller group selected from the target population to participate in the research.

• Sampling: the process used to select the sample.

• Generalisability: the extent to which findings from the sample can be applied beyond the people who participated.

• Sampling bias: a systematic difference between the sample and the target population.

The basic process is:

Identify the target population → select a sample → collect data from the sample → decide whether the findings can be generalised

What is a target population?

The target population is the complete group of people the researcher is interested in.

It is the group to which the researcher hopes the findings will apply.

A target population should be defined clearly.

It may be based on characteristics such as:

• Age.

• Education.

• Occupation.

• Location.

• Psychological characteristic.

• Experience.

• Membership of an organisation.

• Time period.

For example, possible target populations include:

• All Year 12 students at one college.

• All A-Level Psychology students in England.

• All teachers employed by a particular school.

• All adults aged 18 to 25 living in one town.

• All employees working for a particular organisation.

• All people receiving support from one service.

The target population is determined by the research aim.

The word “target” matters

A researcher is not usually trying to generalise to every human being.

They are interested in a particular group.

For example, suppose the aim is:

To investigate the relationship between weekly revision time and assessment performance among Year 12 Psychology students at Westfield College.

The target population is:

All Year 12 Psychology students at Westfield College.

It is not:

• Every student in the UK.

• Every A-Level student.

• Every person who studies psychology.

• Everyone of the same age.

The conclusion should remain within the population identified by the aim.

Defining a population clearly

A vague population might be:

Teenagers.

This leaves several questions unanswered:

• Which ages count as teenagers?

• In which country or location?

• During which period?

• Are school and non-school populations included?

• Is any particular experience required?

A more precise population might be:

Students aged 16 to 18 attending state-funded sixth forms in one English county during the current academic year.

The clearer definition helps the researcher decide:

• Who is eligible.

• Who is not eligible.

• Where participants might be found.

• Which conclusions could be justified.

Inclusion and exclusion

Researchers must decide which people belong to the target population.

For example, a study of A-Level Psychology students might include:

• Students currently enrolled on AQA A-Level Psychology.

• Students in Year 12 and Year 13.

• Students attending selected sixth-form colleges.

It might exclude:

• Students studying a different qualification.

• Former students.

• Students studying Psychology at university.

• People who are not studying Psychology.

Clear boundaries prevent the researcher from selecting participants who do not match the aim.

Accessible population

A target population may be very broad, but the researcher may have access to only part of it.

For example:

• Target population: all A-Level Psychology students in the UK.

• People accessible to the researcher: students at three nearby colleges.

• Sample: 80 students selected from those colleges.

This creates an important limitation.

The sample may represent the accessible colleges reasonably well but may not represent students across the entire UK.

Researchers must distinguish between:

• The group they would like to study.

• The group they can realistically reach.

• The people who actually participate.

What is a sample?

A sample is the group of people selected from the target population to take part in the research.

The sample provides the data used in the study.

For example:

• Target population: all 600 students at a sixth-form college.

• Sample: 60 students selected to complete a questionnaire.

The sample is a subset of the population.

Every participant in the sample should meet the criteria used to define the target population.

Sample and participant

A participant is one individual who takes part in the research.

The sample is the whole group of participants.

For example:

Thirty students completed a memory experiment.

• Each student is a participant.

• The 30 students together form the sample.

Population and sample compared

Target population	Sample
The complete group of interest	The smaller group that participates
Defined by the research aim	Selected from the target population
May contain hundreds, thousands or millions of people	Must be manageable for the investigation
Researchers want conclusions to apply to this group	Researchers collect the actual data from this group
Usually not studied in full	Directly tested, questioned or observed

Worked example

A psychologist wants to investigate examination stress among all Year 13 students at a college.

The college has 480 Year 13 students.

The researcher selects 50 students to complete a stress questionnaire.

• Target population: all 480 Year 13 students at the college.

• Sample: the 50 selected students.

• Participants: each of the 50 individual students.

• Sampling: the process used to choose those students.

• Generalisation: applying the results from the 50 participants to the 480 students.

Another worked example

A psychologist wants to investigate sleep among nurses working night shifts in hospitals within one city.

The researcher interviews 25 nurses from one hospital.

• Target population: all nurses working night shifts in hospitals within the city.

• Sample: the 25 nurses interviewed.

• Possible limitation: nurses from one hospital may differ from those working in other hospitals.

The researcher should not automatically claim that the sample represents:

• All nurses.

• All shift workers.

• All healthcare workers.

• Nurses throughout the UK.

Population and sample in an experiment

Suppose a psychologist investigates whether background speech affects memory among A-Level students.

The researcher recruits 40 students from one college.

• Target population: the group of A-Level students to whom the researcher intends to apply the findings.

• Sample: the 40 recruited students.

• Independent variable: background speech or silence.

• Dependent variable: number of words recalled.

The sample concerns who participates.

The experimental variables concern what happens during the study.

These are separate parts of the research design.

Population and sample in a correlation

A psychologist investigates whether revision time is related to assessment performance.

The researcher obtains paired scores from 100 students.

• Target population: the wider student group of interest.

• Sample: the 100 participating students.

• Co-variable 1: revision time.

• Co-variable 2: assessment score.

The use of a correlation does not remove the need to consider sampling.

A relationship found in one narrow sample may not appear in other populations.

Population and sample in an observation

A researcher observes 12 classes in one school to investigate pupil participation.

Possible populations include:

• All classes in that school.

• All Year 12 classes in that school.

• All Psychology classes in the school.

The correct population depends on the aim.

The sample might consist of:

• The 12 selected classes.

• The pupils observed within those classes.

Researchers need to explain exactly what or who was sampled.

Population and sample in a case study

A case study may investigate only one person, group or institution.

The case is the sample being studied.

For example:

• Target population: schools introducing a new timetable.

• Case studied: one school introducing a new timetable.

The detailed findings may explain that school well, but the sample of one institution creates major limits on generalisation.

This problem is considered in detailed investigations and generalisability.

Why researchers use samples

Investigating the whole population may be impossible

Some target populations are extremely large.

For example:

• All UK university students.

• All adults aged over 65.

• All secondary-school teachers.

• All users of a particular type of service.

The researcher may be unable to:

• Contact every member.

• Obtain consent from everyone.

• Arrange participation.

• Collect and process all the data.

• Complete the study within a reasonable period.

A sample makes the investigation possible.

Time

Research procedures take time.

Participants may need to:

• Complete tasks.

• Answer questions.

• Attend interviews.

• Be observed.

• Provide repeated measurements.

The researcher must also:

• Arrange sessions.

• Give instructions.

• Record data.

• Score responses.

• Analyse results.

• Store information securely.

Studying a manageable sample allows the researcher to complete these stages properly.

Cost

Psychological research may involve costs such as:

• Researcher time.

• Travel.

• Equipment.

• Testing rooms.

• Participant expenses.

• Recording and transcription.

• Data storage.

• Specialist materials.

Investigating every member of a large population would usually be too expensive.

A sample reduces the resources required.

Access

Researchers may not be able to contact everyone in the target population.

For example:

• Personal contact details may be unavailable.

• Organisations may refuse access.

• People may live across a wide geographical area.

• Some members may be difficult to identify.

• Some people may not be available during the study.

The sample is normally selected from people the researcher can identify and reach.

This practical reality can also produce bias.

Participant willingness

Even when people can be contacted, they may not agree to participate.

Possible reasons include:

• Lack of time.

• Lack of interest.

• Privacy concerns.

• Discomfort with the topic.

• Concern about recording.

• Difficulty travelling to the research setting.

Researchers therefore obtain data from those who both meet the criteria and participate.

Those who participate may differ from those who refuse.

Detailed methods require manageable samples

Some methods collect large amounts of information from each participant.

Examples include:

• Long unstructured interviews.

• Repeated observations.

• Detailed case studies.

• Complex experiments.

• Studies lasting several months.

Using a smaller sample may allow the researcher to collect higher-quality information from every participant.

A very large sample may be unsuitable if the researcher cannot conduct the procedure consistently.

Data management

Researchers must manage:

• Consent information.

• Participant codes.

• Test scores.

• Questionnaire responses.

• Interview recordings.

• Transcripts.

• Observational records.

A manageable sample helps researchers:

• Record information accurately.

• Check for mistakes.

• Protect confidentiality.

• Complete appropriate analysis.

Ethical burden

Studying more people than necessary may expose additional participants to:

• Inconvenience.

• Tiredness.

• Sensitive questions.

• Deception.

• Possible distress.

Researchers should collect enough evidence to address the aim without involving people unnecessarily.

The ethical design of research is examined later in Ethics in psychological research.

Samples make research efficient

A well-selected sample can provide useful information about a much larger population.

For example, researchers do not need to question every student in a college if a smaller group adequately represents:

• Different year groups.

• Different subjects.

• Relevant demographic characteristics.

• Other characteristics connected with the aim.

The aim is not simply to obtain the largest possible sample.

The aim is to obtain a sample capable of supporting a justified conclusion.

Sampling

What is sampling?

Sampling is the process through which participants are selected from a target population.

A sampling method determines:

• Who has a chance of being selected.

• How participants are approached.

• Whether particular groups are likely to be included.

• How much researcher choice affects selection.

• How representative the resulting sample may be.

The AQA specification requires knowledge of:

• Random sampling.

• Systematic sampling.

• Stratified sampling.

• Opportunity sampling.

• Volunteer sampling.

These methods are covered in the next three lessons.

Overview of sampling methods

Sampling method	Basic selection principle
Random sampling	Selection is determined by chance
Systematic sampling	Members are selected using a fixed interval
Stratified sampling	Important population groups are represented in appropriate proportions
Opportunity sampling	Available and convenient people are selected
Volunteer sampling	People choose to respond to a request for participants

The method chosen affects:

• Practicality.

• Time and cost.

• Risk of bias.

• Representativeness.

• Generalisation.

Random and systematic sampling

Random and systematic methods use planned selection procedures.

They may reduce the researcher’s personal influence over who is chosen.

However, they require an appropriate way of identifying members of the population.

Their procedures and limitations are examined in Random and systematic sampling.

Stratified sampling

Stratified sampling aims to reproduce relevant proportions found in the target population.

For example, if a population contains different year groups, a stratified sample may include participants from each year in the appropriate proportions.

This can improve representativeness for the characteristics used to form the strata.

The procedure and calculations are covered in Stratified sampling.

Opportunity and volunteer sampling

Opportunity sampling uses people who are available.

Volunteer sampling recruits people who respond to a request.

These methods may be:

• Convenient.

• Quick.

• Inexpensive.

However, they may produce biased samples.

The methods are examined in Opportunity and volunteer sampling.

Representativeness

What is a representative sample?

A representative sample reflects relevant characteristics of the target population.

This means that the sample resembles the population in ways that could affect the research findings.

Relevant characteristics might include:

• Age.

• Gender.

• Educational background.

• Occupation.

• Experience.

• Location.

• Cultural background.

• Level of attainment.

The characteristics that matter depend on the research aim.

Representativeness depends on the investigation

A characteristic may be important in one investigation but less important in another.

For example, suppose a psychologist investigates attitudes towards online learning.

Relevant characteristics might include:

• Age.

• Access to technology.

• Previous experience of online learning.

• Educational setting.

Eye colour would be unlikely to affect the topic.

A sample does not need to match the population on every imaginable characteristic.

It should represent characteristics that may influence the measured behaviour or response.

Representative does not mean identical

A sample cannot include every feature of every population member.

Representativeness means that important patterns are reproduced sufficiently well to support generalisation.

For example:

• Population: 60% Year 12 and 40% Year 13 students.

• Sample: approximately 60% Year 12 and 40% Year 13 students.

The individual participants differ from one another, but the group proportions resemble those

of the population.

Representativeness and variation

Populations contain individual differences.

A useful sample should capture enough of this variation.

Suppose a college contains students with:

• High, medium and low previous attainment.

• Different subject combinations.

• Different revision habits.

• Different backgrounds.

A sample containing only high-attaining Psychology students may not represent the wider college population.

A sample can be representative in one way but not another

Suppose a sample matches a population’s age distribution but contains only volunteers from one department.

The sample may be representative in age but unrepresentative in:

• Occupation.

• Experience.

• Motivation to participate.

• Attitudes towards the research topic.

Representativeness is not a simple all-or-nothing judgement.

Researchers should identify the specific characteristics that are well or poorly represented.

Sampling bias

What is sampling bias?

Sampling bias occurs when the sample differs systematically from the target population in a way that may influence the findings.

A biased sample overrepresents or underrepresents particular types of people.

This means that some population members are more likely to be included than others.

Example of sampling bias

A researcher wants to investigate stress among all students at a college.

They recruit participants from the library at 5.00 pm.

The sample may overrepresent students who:

• Remain at college late.

• Use the library.

• Are completing independent study.

• Have transport allowing them to stay.

• Are willing to stop and participate.

It may underrepresent students who:

• Leave immediately after lessons.

• Study at home.

• Have work or caring responsibilities.

• Avoid the library.

• Are absent that day.

The sample may therefore produce a misleading estimate of stress across the whole college.

Bias is systematic

Sampling bias is not simply any difference between a sample and population.

It is a consistent selection problem.

For example, if a researcher selects only people available in a sports centre, physically active people are likely to be overrepresented.

The difference arises from the way the sample was selected.

Researcher selection

Bias may occur if the researcher chooses participants based on personal judgement.

They might select people who:

• Appear cooperative.

• Are easy to approach.

• Seem likely to understand the task.

• Are similar to the researcher.

• Are available at a convenient time.

This may exclude less accessible or less confident population members.

Location bias

Selecting participants from one location may produce a sample with shared characteristics.

For example:

• One school.

• One hospital.

• One workplace.

• One neighbourhood.

• One online group.

The participants may be influenced by:

• The same policies.

• The same environment.

• Similar local experiences.

• Shared resources.

• The same institutional culture.

Findings may not apply to people in other locations.

Time-related bias

The time at which recruitment occurs may affect who is available.

For example:

• Recruiting in the morning excludes people present only later.

• Recruiting during examination season may produce unusual stress levels.

• Recruiting on one weekday may exclude people with different schedules.

Researchers should consider whether the recruitment period represents the target

population’s usual circumstances.

Volunteer bias

People who volunteer may differ from people who do not.

Volunteers may be more:

• Interested in psychology.

• Confident.

• Available.

• Motivated to help.

• Comfortable discussing the topic.

• Interested in any reward offered.

People who find the topic embarrassing or distressing may be less likely to volunteer.

The resulting findings may not represent the target population.

Restricted samples

Research samples may be restricted to people who are easy to recruit, such as:

• Students.

• People from one institution.

• People living near the researcher.

• People with internet access.

• People willing to attend a laboratory.

A restricted sample may allow a study to be completed efficiently but limits wider conclusions.

Sampling bias and data quality

A study can collect highly reliable data from a biased sample.

For example:

• The task may be standardised.

• Scores may be measured precisely.

• The procedure may be replicated.

However, the findings may still apply only to the type of people included.

Good control and reliable measurement do not remove sampling bias.

Generalisation

What is generalisation?

Generalisation means applying research findings beyond the people who directly participated.

Researchers may wish to generalise from:

• The sample to the target population.

• One setting to similar settings.

• One period to another period.

• One group to related groups.

In this lesson, the main focus is generalisation from the sample to the target population.

From sample to population

Suppose 60 students complete a questionnaire.

The researcher finds that 42 report using practice questions each week.

The direct finding is:

Forty-two of the 60 sampled students reported weekly use of practice questions.

A broader conclusion would be:

Most students in the target population use practice questions each week.

The broader conclusion is justified only if the sample provides a reasonable basis for representing the population.

Generalisation is an inference

The researcher has direct evidence only from the sample.

Applying the result to people who did not participate requires an inference.

The strength of that inference depends on:

• How the population was defined.

• How the sample was selected.

• Whether important groups are represented.

• Sample size.

• Sampling bias.

• Whether the research setting reflects the circumstances of the population.

• Whether relevant people refused or were unable to participate.

Appropriate generalisation

Suppose a researcher randomly selects students from all year groups at one college.

The findings may reasonably be generalised to:

• Students at that college, with appropriate caution.

They should not automatically be generalised to:

• Every sixth-form student in the UK.

• University students.

• Adults who are not in education.

• Students in a different educational system.

The conclusion should match the defined target population.

Overgeneralisation

Overgeneralisation occurs when conclusions are applied more widely than the evidence supports.

For example:

Thirty Psychology students at one college preferred online practice questions. Therefore, all teenagers prefer online learning.

This conclusion is unjustified because:

• The sample studies one subject.

• Participants attend one college.

• The sample may be self-selecting.

• Preference for practice questions is not identical to preference for all online learning.

• Teenagers outside education are not represented.

Generalisation to similar populations

A researcher may suggest that findings could apply to people with similar characteristics.

For example:

The findings may be relevant to students of a similar age attending comparable colleges.

This is more cautious than claiming universal application.

Further research with:

• Other institutions.

• Other locations.

• Different participant groups.

• Different time periods.

would be needed to support broader generalisation.

Generalisation and research aim

The research aim should identify the intended population.

Aim:

To investigate examination anxiety among students at Parkside College.

A sample from Parkside College may address this aim.

Aim:

To investigate examination anxiety among all UK A-Level students.

A sample from one class is unlikely to represent this much broader population.

The same sample can therefore be more or less appropriate depending on the stated aim.

How sample size affects generalisation

Larger samples may capture more variation

A larger sample may include a wider range of:

• Ages.

• Experiences.

• Abilities.

• Attitudes.

• Backgrounds.

• Behaviour.

This can reduce the influence of unusual individuals on the overall result.

For example, the results from 200 students may be less affected by one unusually high anxiety score than the results from five students.

Larger does not automatically mean representative

A large biased sample can still produce weak generalisation.

Suppose a researcher recruits 2,000 volunteers through a revision website.

The sample may still overrepresent students who:

• Use online revision.

• Are motivated to engage with educational content.

• Have internet access.

• Notice and respond to research advertisements.

The large number does not remove volunteer bias.

A smaller carefully selected sample may be stronger

A smaller sample that includes relevant population groups appropriately may support better generalisation than a much larger convenience sample.

For example:

• Sample A: 100 students selected across year groups and courses.

• Sample B: 1,000 volunteers recruited from one Psychology revision forum.

Sample A may represent the college population more effectively, despite being smaller.

Very small samples

A very small sample may:

• Fail to include important population differences.

• Be strongly affected by unusual participants.

• Produce unstable patterns.

• Make it difficult to justify generalisation.

However, small samples may still be valuable in:

• Case studies.

• Detailed interviews.

• Research involving rare cases.

• Early exploratory investigations.

The researcher should match their conclusion to the evidence.

Sample size and practical quality

Increasing sample size can create practical problems.

If resources are stretched, researchers may:

• Give less careful instructions.

• Record data less accurately.

• Reduce the depth of interviews.

• Rush observations.

• Fail to check data properly.

A suitable sample size balances:

• Representativeness.

• Practical feasibility.

• Quality of data collection.

• Ethical responsibility.

Sample composition

Composition can matter more than total size

Sample composition refers to the characteristics of the participants included.

Two samples of the same size may differ greatly.

For example:

Sample A

• Includes students from several subjects.

• Includes different year groups.

• Includes a range of attainment levels.

Sample B

• Contains only high-attaining Year 13 Psychology students.

If the target population is the whole college, Sample A is likely to be more representative.

Overrepresentation

A group is overrepresented when it forms a larger proportion of the sample than of the population.

Example:

• Target population: 20% Psychology students.

• Sample: 70% Psychology students.

Psychology students are overrepresented.

If subject choice affects the measured response, the findings may be biased.

Underrepresentation

A group is underrepresented when it forms a smaller proportion of the sample than of the population.

Example:

• Target population: 40% Year 13 students.

• Sample: 10% Year 13 students.

Year 13 students are underrepresented.

The sample may not reflect the experiences of the wider population accurately.

Complete exclusion

A group may be absent from the sample.

For example, an online questionnaire excludes people without suitable internet access.

If those excluded differ from those who participate, the findings may not generalise to the entire population.

Relevant characteristics

Researchers should explain why a characteristic might affect the findings.

Weak evaluation:

The sample contains more Year 12 students, so it is biased.

Developed evaluation:

Year 12 students are overrepresented. They may have less experience of A-Level examinations than Year 13 students and may report different levels of examination anxiety. The sample may therefore provide an inaccurate estimate for the whole sixth form.

The developed answer links sample composition to the measured variable.

Sampling and different research methods

Experimental research

An experiment may have high control but use an unrepresentative sample.

For example:

• Laboratory procedure is standardised.

• Participants are all volunteer university students.

The experiment may provide a valid comparison within the sample, but its findings may not generalise to:

• Other age groups.

• Non-students.

• People who would not volunteer.

Control over variables and representativeness are separate issues.

Correlational research

A correlation found in one sample may not appear in another.

For example, the relationship between revision and assessment performance might differ according to:

• Age.

• Subject.

• Educational setting.

• Prior attainment.

• Access to resources.

Researchers should avoid assuming that one sample produces a universal relationship.

Questionnaire research

Questionnaires can reach large samples, especially online.

However, respondents may differ from non-respondents.

People who answer may be:

• More interested in the topic.

• More confident expressing their views.

• More available.

• More likely to hold strong opinions.

Large response totals do not automatically remove sampling bias.

Interview research

Interviews are time-consuming and often use smaller samples.

This may limit generalisation.

However, the purpose may be to obtain:

• Detailed explanations.

• Personal experiences.

• Rich qualitative evidence.

Researchers should not criticise a small interview sample without considering the aim.

Depth may be prioritised over broad population estimates.

Observational research

Researchers may observe people who happen to be present in one location.

For example:

• Customers in one shop.

• Pupils in one playground.

• Passengers at one station.

The observed sample may differ according to:

• Time.

• Day.

• Location.

• Weather.

• Activity.

• Who chooses to use the setting.

Generalisation beyond the observed context should be cautious.

Content analysis

Content analysis samples communication rather than always sampling participants directly.

Researchers might sample:

• Newspaper articles.

• Television episodes.

• Interview transcripts.

• Social-media posts.

The same principle applies.

The selected material should represent the wider body of content about which conclusions are drawn.

Case studies

A case study usually contains one case or a very small number.

Its strength is detailed understanding rather than population representation.

Findings may:

• Demonstrate that a pattern is possible.

• Generate a hypothesis.

• Challenge a universal claim.

They do not usually show how common the pattern is in the population.

Random selection and random allocation

These are different processes

Random selection concerns how people are chosen for the sample.

Random allocation concerns how sampled participants are assigned to experimental conditions.

For example:

1. Forty students are randomly selected from a college population.

2. The 40 selected students form the sample.

3. The 40 students are then randomly allocated to a music or silence condition.

The two procedures address different problems.

Purpose of random selection

Random selection aims to:

• Reduce researcher choice.

• Give population members an equal chance of selection where possible.

• Reduce systematic sampling bias.

It affects the representativeness of the sample.

Purpose of random allocation

Random allocation aims to distribute participant variables across experimental conditions.

It affects the fairness of the experimental comparison.

Random allocation does not make the sample representative of the population.

A sample of 40 volunteers remains a volunteer sample even if those volunteers are randomly allocated to conditions.

Random allocation is examined further in Control procedures.

Assessing a sample in an exam scenario

Step 1: identify the target population

Ask:

Who does the researcher want the findings to apply to?

Use the research aim rather than assuming the broadest possible population.

Step 2: identify the sample

Ask:

Who actually provided the data?

State the number and relevant characteristics where given.

Step 3: identify the sampling method

Ask:

How were participants selected?

This may be:

• Random.

• Systematic.

• Stratified.

• Opportunity.

• Volunteer.

Step 4: compare the sample with the population

Consider:

• Which groups are represented?

• Which are underrepresented?

• Which are absent?

• Does the sample come from one location?

• Did participants choose to take part?

• Were only convenient people included?

Step 5: link differences to the research variable

Explain why the sample characteristic might affect the outcome.

For example:

Recruiting only students attending an optional revision session may produce unusually motivated participants. Motivation could affect both revision behaviour and examination confidence.

Step 6: judge generalisation

State exactly where the findings may reasonably apply.

For example:

The findings may apply to motivated students attending similar revision sessions, but cannot confidently be generalised to all students at the college.

Step 7: suggest an improvement

A suitable improvement might involve:

• Selecting participants from several locations.

• Including relevant population groups.

• Using a method with less researcher choice.

• Increasing sample size while maintaining representativeness.

• Defining the population more narrowly.

• Replicating the research with another sample.

Worked scenario: examination anxiety

Research description

A psychologist wants to investigate examination anxiety among all Year 12 and Year 13 students at a college.

She asks 40 students attending an optional after-school revision session to complete a questionnaire.

Target population

All Year 12 and Year 13 students at the college.

Sample

The 40 students attending the optional revision session.

Sampling concern

The students were available at one particular activity.

They may be more:

• Academically motivated.

• Concerned about examinations.

• Willing to stay after school.

• Interested in revision.

Effect on findings

Their anxiety scores may be higher or lower than those of students who do not attend revision sessions.

For example:

• Greater concern may produce higher anxiety.

• Greater preparation may produce lower anxiety.

The direction is uncertain, but the sample remains potentially unrepresentative.

Generalisation

The findings should not be generalised confidently to all students at the college.

They may apply more closely to students who attend optional revision sessions.

Possible improvement

The researcher could select students from:

• Both year groups.

• Different courses.

• Different timetable periods.

A stratified procedure might help ensure that relevant college groups are represented proportionately.

Worked scenario: online questionnaire

Research description

A psychologist advertises an online questionnaire about social-media use through a social-media account.

Five thousand people complete it.

Strength

The sample is large and allows extensive data to be collected efficiently.

Limitation

The participants are volunteers who saw and responded to an online post.

They may:

• Use social media more frequently.

• Be especially interested in the subject.

• Have strong views.

• Be more comfortable sharing information online.

Generalisation

The researcher should not assume that the 5,000 participants represent all people.

The large sample remains self-selected.

Main lesson

Sample size cannot correct a biased selection process by itself.

Worked scenario: memory experiment

Research description

A psychologist investigates whether background noise affects memory in adults.

The sample consists of 30 Psychology students from the researcher’s college class.

Target population problem

If the target population is all adults, the sample is narrow because it contains:

• Students.

• Psychology students.

• People from one college.

• People of a limited age range.

Effect on findings

Psychology students may:

• Be familiar with memory tests.

• Recognise the research aim.

• Have similar educational experiences.

The findings may not apply to adults with different ages or backgrounds.

Better conclusion

Background noise affected memory performance among the Psychology students tested.

This is more justified than:

Background noise affects memory in every adult.

Sampling and generalisation trade-offs

Practicality and representativeness

Convenient samples are easier to obtain.

Representative samples are often more difficult and expensive to construct.

Greater practicality	Greater representativeness
Participants are quick to access	More planning may be required
Research costs may be lower	Researchers may need several locations
Data can be collected rapidly	Population information may be required
Sample may be narrow or biased	Relevant groups are more likely to be included
Generalisation may be limited	Broader conclusions may be more justified

No method guarantees perfect representation.

Researchers must explain the trade-off.

Depth and breadth

A small sample may allow:

• Long interviews.

• Detailed observation.

• Repeated testing.

• Extensive case information.

A large sample may allow:

• Better population coverage.

• More numerical comparison.

• Reduced influence of unusual individuals.

The appropriate balance depends on the aim.

Improving generalisability

Define the population accurately

Researchers should avoid claiming a broader population than the study addresses.

A carefully defined target population allows a more precise conclusion.

For example:

Students attending one sixth-form college.

is more realistic than:

All young people.

Select from more than one setting

Including participants from several:

• Schools.

• Workplaces.

• Hospitals.

• Communities.

• Locations.

may reduce the influence of one institution or local environment.

Include relevant groups

Researchers should identify characteristics likely to affect the findings and ensure that these groups are not excluded.

This is the purpose of Stratified sampling.

Reduce researcher choice

Using a planned procedure can reduce the researcher’s personal influence over who participates.

Random and systematic selection are examined in Random and systematic sampling.

Replicate with new samples

Researchers can repeat the study using:

• Different age groups.

• Different institutions.

• Different locations.

• Different cultural groups.

• Different time periods.

Similar findings across samples provide stronger support for broader generalisation.

Use cautious conclusions

Researchers should report:

• Who took part.

• How they were selected.

• Which population was targeted.

• Which groups were absent.

• Where conclusions may reasonably apply.

Scientific caution increases the credibility of the report.

Writing an effective population-sample distinction

A strong answer might state:

The target population is the complete group about which the researcher wants to draw conclusions, whereas the sample is the smaller group selected from that population to take part and provide the research data.

This answer:

• Defines both terms.

• Gives their relationship.

• Identifies the function of the sample.

Applied distinction

Scenario:

A researcher selects 80 employees from a company employing 1,200 people.

Strong answer:

The target population is all 1,200 employees in the company. The sample is the 80 employees selected to participate.

Writing an effective sampling evaluation

Weak answer:

The sample was biased.

This does not explain:

• Which people were overrepresented.

• Why the sample was biased.

• How the bias could affect the results.

• Which conclusions are limited.

Developed answer:

The researcher recruited volunteers from an optional revision class. These students may be more motivated and concerned about academic performance than students who did not attend. Their revision and anxiety scores may therefore differ systematically from the wider college population, limiting generalisation to all students.

Writing an effective generalisability paragraph

The findings may have limited generalisability because all 40 participants were Psychology students from one college. They share a similar educational setting and may be more familiar with psychological research tasks than other students or adults. The findings can therefore be applied most confidently to students with similar characteristics rather than to the wider adult population.

This paragraph:

• Identifies the sample.

• Explains how it differs from the wider population.

• Links the difference to behaviour.

• Gives a cautious judgement.

Overall summary

A target population is:

• The complete group the researcher wants to understand.

• Defined by the research aim.

• The group to which findings may be generalised.

A sample is:

• The smaller group selected from the population.

• The people who actually provide the data.

• Chosen through a sampling method.

Researchers use samples because studying a whole population may be:

• Impossible.

• Too expensive.

• Too time-consuming.

• Difficult to access.

• Unnecessary.

• Inappropriate for a detailed method.

A sample supports generalisation when it:

• Represents relevant population characteristics.

• Avoids systematic selection bias.

• Includes appropriate population groups.

• Is large enough for the investigation.

• Is selected using a suitable procedure.

A sample limits generalisation when it:

• Comes from one narrow setting.

• Contains only available or willing people.

• Overrepresents particular groups.

• Excludes relevant population members.

• Is too small to reflect important variation.

• Is treated as representative without evidence.

The key principle is:

Findings are collected from a sample, but conclusions are often intended for a population. The quality of that generalisation depends on how well the sample represents the population.

Key Words 🔑

Key word	Student-friendly definition	How it may be used in an exam
Target population	The complete group about which the researcher wants to draw conclusions.	Identify the wider group named or implied by a research aim.
Population	The full group relevant to an investigation.	Distinguish the wider group from the participants studied.
Sample	The smaller group selected from the population to participate.	Identify who actually provided the data.
Participant	One individual who takes part in the research.	Distinguish an individual from the complete sample.
Sampling	The process used to select participants from a population.	Explain how the sample was obtained.
Sampling method	The procedure used to choose the sample.	Identify random, systematic, stratified, opportunity or volunteer selection.
Representative sample	A sample that reflects relevant characteristics of the target population.	Evaluate whether findings can be generalised.
Sampling bias	A systematic difference between the sample and target population caused by selection.	Explain why some groups are overrepresented or excluded.
Generalisability	The extent to which findings can be applied beyond the sample studied.	Evaluate conclusions drawn about a target population.
Overgeneralisation	Applying findings to a wider group than the evidence justifies.	Identify an unsupported research conclusion.
Sample size	The number of participants in a sample.	Evaluate whether population variation is likely to be represented.
Sample composition	The characteristics of the people included in a sample.	Explain why size alone does not ensure representativeness.
Overrepresentation	Including a group in a larger proportion than it forms in the population.	Identify a source of sampling bias.
Underrepresentation	Including a group in a smaller proportion than it forms in the population.	Explain why a sample may not reflect the population.
Random sampling	Selecting participants through a chance procedure.	Explain a method intended to reduce researcher selection bias.
Systematic sampling	Selecting population members using a fixed interval.	Identify a planned sampling method.
Stratified sampling	Selecting participants so relevant population groups are represented proportionately.	Explain how representation may be improved.
Opportunity sampling	Selecting people who are available and convenient.	Evaluate practicality and sampling bias.
Volunteer sampling	Recruiting people who choose to respond to a request.	Explain volunteer bias and limits on generalisation.
Random selection	Using chance to choose members of the population for the sample.	Distinguish participant selection from allocation to conditions.
Random allocation	Using chance to place sampled participants into experimental conditions.	Explain a control procedure rather than a sampling method.

Common Mistakes ⚠️

Mistake: Describing the sample as the whole group the researcher is interested in.

Why this is incorrect:The whole group is the target population. The sample is the smaller group that participates.

How to improve:Ask who the researcher wants to understand and who actually provided the data.

Mistake: Assuming that the population means everyone.

Why this is incorrect:A target population is a clearly defined group relevant to the research aim.

How to improve:Include the appropriate age, setting, experience or membership criteria.

Mistake: Identifying the population from the sample rather than the research aim.

Why this is incorrect:The sample may be narrower than the intended target population.

How to improve:Use the stated purpose of the investigation to define the population.

Mistake: Saying researchers use samples only because they are easier.

Why this is incomplete:Researchers may also face limits involving time, cost, access, data management and participant burden.

How to improve:Explain a specific practical reason and link it to the investigation.

Mistake: Saying a large sample is automatically representative.

Why this is incorrect:A large sample may still overrepresent volunteers, one location or one type of participant.

How to improve:Evaluate both sample size and the method of selection.

Mistake: Saying a small sample is always useless.

Why this is incorrect:Small samples may provide detailed or valuable evidence, particularly in interviews and case studies.

How to improve:Evaluate whether the sample suits the research aim while recognising limits on generalisation.

Mistake: Describing sampling bias without identifying the biased group.

Why this is incomplete:The examiner needs to know who is overrepresented, underrepresented or excluded.

How to improve:Name the group and explain how its characteristics may affect the findings.

Mistake: Saying a convenient sample represents the population because all participants belong to it.

Why this is incorrect:Belonging to the population does not mean the sample reflects the population’s variation.

How to improve:Compare the sample’s composition with the wider population.

Mistake: Assuming volunteers are representative because anyone could volunteer.

Why this is incorrect:People who choose to participate may differ systematically from people who do not.

How to improve:Explain a relevant volunteer characteristic, such as greater interest or willingness to discuss the topic.

Mistake: Generalising from one school to all students.

Why this is incorrect:Students at one school share an environment, policies and local circumstances.

How to improve:Restrict the conclusion or replicate the research across several settings.

Mistake: Saying sampling bias makes every finding within the sample invalid.

Why this is incorrect:The data may accurately describe the participants tested while failing to represent the wider population.

How to improve:Separate the quality of the sample data from the reach of the conclusion.

Mistake: Confusing random selection with random allocation.

Why this is incorrect:Random selection chooses the sample. Random allocation places participants into experimental conditions.

How to improve:Identify whether the procedure occurs before or after the participants enter the study.

Mistake: Saying random allocation produces a representative sample.

Why this is incorrect:Allocation balances experimental conditions but does not change how the sample was recruited.

How to improve:Evaluate the original sampling method separately.

Mistake: Stating that a sample is unrepresentative without linking this to the measured

behaviour.

Why this is incomplete:The difference matters only when it could influence the research results.

How to improve:Explain why the overrepresented characteristic could affect the DV, questionnaire answer or observed behaviour.

Mistake: Generalising beyond the target population stated in the aim.

Why this is incorrect:The research was not designed to represent the wider group.

How to improve:Keep conclusions within the defined age, location, institution or experience.

Mistake: Claiming a representative sample guarantees a universal finding.

Why this is incorrect:Results remain limited by the measures, procedure, setting and time of the investigation.

How to improve:Use cautious language and support broader claims with replication.

Exam-Style Questions ✍️

Question 1

Which one of the following best describes a target population?

A. Every person who agrees to participate

B. The complete group about which the researcher wants to draw conclusions

C. The participants allocated to one experimental condition

D. The people who obtain the most typical scores

[1 mark]

Question 2

Explain one difference between a target population and a sample.

[2 marks]

Question 3

A college has 850 students. A psychologist selects 70 students to complete a questionnaire about examination stress.

Identify:

a) The target population.

[1 mark]

b) The sample.

[1 mark]

Question 4

Explain two reasons why a psychologist might use a sample rather than investigate every member of a target population.

[4 marks]

Question 5

A researcher wants to investigate revision habits among all students at a sixth-form college. She recruits 40 students attending an optional after-school revision session.

Explain one way in which the sample may be biased.

[4 marks]

Question 6

A psychologist obtains 3,000 responses to a questionnaire advertised on a social-media platform.

Explain why the large sample size does not necessarily mean that the sample is representative.

[4 marks]

Question 7

A memory experiment uses 30 A-Level Psychology students from one college. The researcher concludes that the findings apply to all adults.

Explain why this conclusion may not be justified.

[4 marks]

Question 8

Explain why a smaller representative sample may support stronger generalisation than a much larger biased sample.

[4 marks]

Question 9

A researcher randomly selects 50 employees from a company and then randomly allocates them to two experimental conditions.

Explain the difference between random selection and random allocation in this investigation.

[4 marks]

Question 10

Discuss how sampling affects the generalisation of psychological research findings.

Refer to population, sample, representativeness and bias in your answer.

[8 marks]