Sampling Continued 14/10/2013
The method of choosing a smaller,
representative group of
Census respondents with which to
conduct research instead of the
whole population.
Population A survey of everyone in the
country.
Sample The total number of people who
could be included in a particular
survey
1 of 44
Keywords - sampling
systematic stratified
cluster random
convenience quota
Self selecting
2 of 44
Task One: Sampling V Census
1. Describe the advantages of using a census rather
than a sample.
2. Describe the disadvantages of using a census rather
than a sample
3. A large school has 1800 pupils. The head teacher wants
to find out how far the pupils have to travel to school.
a. What is the population he could use?
b. Would you use a census or a sample?
c. Explain why you would give this advice.
3 of 44
Sampling
There are three main factors when we think
about what sample we are going to take.
1. What size of sample can you expect to take,
given limited time, money and resources?
2. How is the sample to be collected to avoid
bias?
3. How is the data going to be collected to
avoid bias?
4 of 44
How
big? Sample Size
The sample size needs to be considered as part of the
data collection process.
A large sample will cost more time, money and effort but
may be more representative and may produce better
results.
For a small population, 5-10% is a large enough sample size.
For a larger population of size n, we take a sample of size
√n. E.g. Population of size 400, sample size 20.
5 of 44
Yum! Sample Size
Discretion can,
however, be used and
smaller samples taken
when necessary.
Well you can’t
sample ALL of
them, can you?
6 of 44
Sample Size – For your notes . . .
Key ideas:-
A larger sample may be more representative and may
produce better and less variable results. It should
give greater reliability.
For a small population, 5%-10% is a large enough
sample size.
For a larger population of size n, we take a sample of
size √n. E.g. Population of size 400, sample size 20.
7 of 44
Random Sampling
What do we mean by a random sample?
How could we create a random sample?
8 of 44
Using a random number table
In a class there are 30 pupils. The teacher decides
to take a random sample of 5 pupils to estimate
the mean height of the pupils in the class. Select a
random sample of 5 pupils from the list.
9 of 44
Using a random number table
We can now use a random number table like:
3 4 5 4 7 3 0 7 7 1 4 0 2 6 0 5 8 8 8 5
6 6 4 5 5 7 6 0 2 7 8 2 7 6 1 5
10 of 44
Task Two
On your desk is a worksheet on random sampling
techniques.
There is also a CHALLENGE question available for
you to complete.
11 of 44
Systematic Sampling
What do we mean by systematic sampling?
How could we create a systematic sample?
12 of 44
Systematic Sampling
We will look at the same class as before. What
would we do if we wanted a systematic sample of 6
people?
13 of 44
Task Three
Now complete the worksheet on systematic
sampling.
As an extension – can you create a systematic
sample of the people in the class?
14 of 44
Convenience Sampling
is quick to organise
there is no need to order or number the
sample frame
you work with what is available to you at
the time
. . . can lead to an enormous problem of bias
15 of 44
Quota Sampling
This is probably the easiest method to use in
the real world.
Companies use screening questions at the start
of any survey issued to make sure they can
quickly identify their target audience.
E.g. Do you use sun care products?
No? End of survey… Yes? Answer the following …
16 of 44
Cluster Sampling
Cluster is another word for group and we use
this method when the population can be put into
or naturally falls into groups.
Within each cluster there must be a good mix of
the different elements of the population with
the numbers in each cluster spread out as
equally as possible..
This means that any one cluster can be selected
at random and the population of that cluster
used as the sample set.
17 of 44
Cluster vs Stratified Sampling
What’s the
difference?
We group
We need a together
variety of members of
characteristics the population
in every with similar
cluster. characteristics
(age, gender,..).
18 of 44
Sampling
Please complete the mixed graded
exercise to test your understanding of
‘Sampling’. . .
. . .Next lesson we will look in greater
detail at ‘Stratified Sampling’.
19 of 44
