Methods of Sampling
By Tom B
Different methods of sampling
Everything possible item of data has
an equal chance of being used in the
Use a random number table
Use a random number generator or a
Put the numbers into a hat
Divide the population into categories
Male or Female, age, social class,
From within each strata take a
The size of each sample is in
proportion to the strata from which it
Chosen at regular intervals from an
E.g.Take every 5th student from a
Used when population is very large
May not always be truly
The population being sampled is split into
groups, or clusters.
The clusters to be sampled are randomly
Every cluster is looked at.
Clusters should be more or less an equal
A large number of clusters is best so that
it minimizes the chances of the sample
Popular with scientists
Used in market research
Information is given about the quota or
amount of each section with a certain
characteristic that is to be sampled.
No sampling frame is required
Major disadvantage is that the choice of
the sample is the surveyors decision which
can lead to questions of bias.
This is where the most convenient sample
For a sample of 50 students from a school
it will simply be that they choose the first
50 names on the register.
Very quick and is not time consuming
Major disadvantage is that it can lead to
bias and the sample can be
Sometimes you have to use more than
one sampling method.
Opinion polls in politics use combination of
quota and cluster sampling
A major disadvantage of the conclusions
drawn is that opinions can change
Each method of sampling has a drawback
you have to choose the right one for what
you need it for.