Notes

Notes

Notes Quick Questions Answers to Questions

Back Home Next

Sampling

Census

-Sample everything

The disadvantages are; not practical and sometimes destructive testing has to be used.

 

Sample

-Usually, samples take data from around 10% of the population of interest

In exam answers you will get marks for using the words in red... "the sample must be representative of the whole population  to avoid being biased".

 

1) Simple random sampling

-The idea behind this method is that each member of the population has an equal chance of being questioned.

 

2) Stratified sampling

-The population is subdivided into categories and the sample has the same PROPORTION of each category as the population.

For example; in a school with 200 girls and 100 boys a 10% sample would include 20 girls and 10 boys.

Once you know how many of each category are included, use simple random sampling to choose which girls and boys are asked.

The advantage of this is that it is more representative of the population than a simple random sample

 

3) Systematic sampling

A systematic sample questions people/tests products etc. AT REGULAR INTERVALS.

For example; for a 10% systematic sample (i.e. questions every tenth person)...

Generate a random number between 1 and 10 and then go up in tens.

For a 20% sample  (i.e. question every fifth person)...

Generate a random number between 1 and 5 and then go up in fives.

The advantage of this is that it is quick

The disadvantage is that subdivisions can seriously affect results e.g. you could get the third-best pupil in every class

 

4) Cluster sampling

A group is chosen at random from the populations and a simple random sample is conducted within that group. For example if I want to find out about patients in the UK I might choose just one hospital and just sample patients from that hospital.

The advantage of this is that it is quick & convenient

The disadvantage is that one group in a particular location may not be representative of the whole population. For example if I chose my hospital in London, their diseases might not be the same as in Scotland.

 

5) Quota sampling (also called convenience sampling)

This is exactly the same as cluster sampling except for once the group is chosen, no precautions are carried out to ensure that people are chosen totally randomly from the within group. For example, people would be stopped on the hospital corridors instead of random numbers to select particular patients being generated by a computer. This obviously has problems, for example, very ill people who do not leave their beds will not be asked.

The disadvantage is that one group in a particular location may not be representative of the whole population (as with cluster) AND that there is is no RANDOM selection from the cluster. EACH MEMBER OF THE GROUP DOES NOT HAVE AN EQUAL CHANCE OF BEING CHOSEN.

The advantage of this is that it is very quick & convenient.

 

Back Home Next

 

2006-2007 mathsMAD