Database results:
examBoard: AQA
examType: GCSE
lessonTitle: Stratified Sampling
Stratified sampling is a probability sampling method where the researcher divides the entire population into different subgroups (called strata), then randomly selects participants from each of these subgroups. This ensures that each important segment of the population is properly represented in the sample.
Key Definitions:
Psychologists use stratified sampling when they want to ensure that specific subgroups of a population are adequately represented. For example, if you're studying attitudes toward mental health services, you might want to ensure your sample includes people from different age groups, genders and socioeconomic backgrounds in the same proportions as they exist in the population.
Stratified sampling works best when:
• The population naturally divides into distinct groups
• These groups might respond differently to what you're studying
• You need to make comparisons between these groups
• You want to increase the precision of your overall population estimates
Conducting a stratified sample involves several clear steps:
A psychologist wants to study stress levels among 1,000 secondary school students. The school has the following year groups:
If the researcher wants a sample of 100 students using stratified sampling, they would select:
Each group would be randomly selected from their respective year groups.
Ensures all important subgroups are properly represented in the sample, even relatively small ones that might be missed in simple random sampling.
Generally provides greater statistical precision than simple random sampling if the strata are homogeneous internally but different from each other.
Allows researchers to make comparisons between different subgroups in the population.
Requires researchers to know the distribution of the population across the strata before sampling begins.
More complex and time-consuming than simple random sampling, requiring more effort to organise and implement.
Can become very complicated if stratifying by multiple characteristics simultaneously.
When studying the effectiveness of a new therapy for depression, researchers might stratify their sample by severity of depression (mild, moderate, severe) to ensure each group is properly represented and to see if the therapy works differently depending on severity.
When investigating learning styles, researchers might stratify by year group, academic achievement levels, or specific learning needs to ensure their findings represent all types of students.
The UK Office for National Statistics conducted a mental health survey using stratified sampling. They divided the population by age, gender, region and socioeconomic status to ensure their findings accurately represented the UK population. This allowed them to identify that young women (16-24) had the highest rates of common mental health problems, while men aged 65+ had the lowest rates. Without stratified sampling, these important differences might have been missed if certain groups were under-represented.
Let's work through an example of how to calculate a stratified sample:
A psychology researcher wants to survey student satisfaction at a university with 10,000 students. The student population is distributed across faculties as follows:
If the researcher wants a sample of 500 students, the stratified sample would be:
The formula used is:
Number in stratum = (Size of stratum ÷ Total population) × Total sample size
Simple random sampling gives every individual an equal chance of selection, but might miss smaller subgroups.
Stratified sampling ensures all subgroups are represented proportionally, but requires more knowledge about the population.
Opportunity sampling uses whoever is available and willing to participate, making it quick but potentially biased.
Stratified sampling is more systematic and representative, but takes more time and effort to implement.
When evaluating stratified sampling in your exams, remember these key points:
In your exams, always explain both how stratified sampling works and why it might be appropriate for the specific research scenario described. Use examples to show your understanding and be prepared to calculate how many participants would be needed from each stratum if given population information.
Log in to track your progress and mark lessons as complete!
Login NowDon't have an account? Sign up here.