Sampling Flashcards
What is a population?
the group we are interested in knowing more about
What is a sample?
The group of individuals chosen from the population to represent it in a research study
What is a target population/sampling frame?
Refers to the population from which we will sample, but is more specific than vague population
What is a representative sample>
sample contains sub-groups of people in direct proportion to their prevalence in general population - sample accurately reflects characteristics of the population.
Expected to represent behaviours and attitudes of population we want to study
What is external validity?
The extent to which the results of one study can be generalised across settings, time and populations
What is generalisation?
The process of extrapolating the results
What is sampling bias?
selection bias - systematic tendency towards over or under-representation of some categories in a sample
What is a biased sample?
A sample in which members of a sub-group of larger population are over/under-represented
What is sampling error?
Any difference in the observations between the sample and population
- all research studies have some sampling error because the sample will never give the same observations as the entire population.
- the best we can do is minimise the sampling error as much as possible
What is standard error (SE)?
Measures how well the mean of a particular sample represents the mean of the population
What does a small standard error indicate?
Sample means are similar to population mean and the sample is likely to be an accurate reflection of the population
What is the equation for standard error?
SE = standard deviation / square root of the number of samples
What are 2 ways to reduce sampling errors?
- larger sample (needs to be representative)
- Choose a sampling technique that gives you the most representative sample for the population you want to study.
What are the 2 types of sampling?
- probability sampling
- non-probability sampling
What is probability sampling?
There is a specific/known probability of each member of the population being sampled
- requires a clearly defined population we can have access to
- most likely representative of whole population - should use wherever feasible
- can reduce sampling error
What is non-probability sampling?
There is NOT a specific and known probability of each member of the population being sampled
- often used in psych research
(often interested in identifying trends and relationships between variables)
- individuals not chosen randomly
What are 4 examples of probability sampling?
- simple random sampling
- systematic random sampling
- stratified sampling
- cluster sampling
What are 4 examples of non-probability sampling?
- convenience sampling
- purposive sampling
- quota sampling
- snowball sampling
What is simple random sampling?
Probability sample
A sample is chosen randomly from population so everyone has an equal chance of being selected
- reduces sampling error by choosing from all members of population to represent it
- sample units are selected randomly
What is one way to achieve simple random sampling?
The lottery method
- computer selection/random number tables/manual selection (each unit in population assigned a number, written on a piece of paper and put in a hat)
What is systematic random sampling?
Probability sample
Requires selecting samples based on a system of intervals in a numbered population
- random starting point but with a fixed, periodic interval
- Sample selected by taking every nth case from a list of target population (eg. every 5th)
- Sample units selected systematically via a single technique
What is stratified sampling?
Probability sample
Determine groups (strata) within population whose proportions we want the sample to reflect - enhances representativeness
- divide population into groups
- proportion of group in sample should be equal to proportion of group in population
- reduces bias
- similar to simple random
What is cluster sampling?
Probability sample
Clusters of individuals are identified then a subset of clusters is randomly chosen to sample from
- sample chosen randomly from clusters identified in population
- makes it easier to choose members randomly from smaller clusters to better represent population
- can ignore segments of population not in clusters chosen for sample
What is convenience sampling?
Non-probability sample
A sample is chosen from people who are available to participate in the research
AKA availability sampling or volunteer sample
- available to researcher, respond to an ad or volunteer
- used in many psych experiments
What is purposive sampling?
Non-probability sample
Researchers rely on their judgement when choosing members of population to participate in study
- sample consists of people who have a particular set of characteristics that are relevant to the study
AKA judgement, selective or subjective sampling
What is quota sampling?
non-probability sampling
Similar to stratified except uses non-probabilistic sampling technique
- identify relevant characteristics of units and their percentage in the population (quotas) and then apply convenience sampling to fill quotas
- representative
What is snowball sampling?
non-probability sample
Can be used by researchers looking for a specialised population
- used for research with hidden populations, excluded communities or invisible groups
- past ppts recommend future ppts from same population
- introduces bias into sample (not representative)
What are WEIRD populations?
W - Western
E - Educated
I - Industrialised
R - Rich
D - Democratic
Highly represented in psychological research - persistent sampling bias
What is non-response bias?
Can lead to non-representative sample of individuals who refuse to participate in study
eg. unhappy couples refuse to participate in study about marriage
How can sampling bias relate to culture?
Samples may not be comparable and variations in samples from one cultural context to the other can confound observer scores
