TYPES OF SAMPLING:
Random Sampling: Random sampling is used to reduce bias so that a set of data is not one sided. Random sampling can be obtained by using a calculator. When using random sampling you must ensure that all samples are equally likely to be chosen.
Stratified sampling: Strata means ‘layer’. A stratified sample is made up of different layers of the population. The sample size is proportional to the size of its layer. To do this the formula is the number in the year group *50 /number in the whole school. Random samples are then taken from each section of the population.
Systematic sampling: If a sample of size t is to be taken from the population of size u then every t/u member of the population is to be tested. Every member of the sample is chosen at regular intervals from the list. A sample chosen in this way can be biased if low or high values come in a certain pattern. The starting point is chosen randomly.
In order to test, my hypotheses I must take a sample from across a range of years. This is because the sample of year 7’s must be in proportion o the size of the whole school.
My stratified sample (males) My stratified sample (females)
Stratified sample of 50 males Stratified sample of 50 females
Year 7 males: 150/601*50 = 12 Year 7 females: 131/589*50 = 11
Year 8 males: 144/601*50 = 12 Year 8 females: 125/589*50 = 11
Year 9 males: 117/604*50 = 10 Year 9 females: 153/589*50 = 13
Year 10 males: 106/601*50 = 9 Year 10 females: 94/589*50 = 8
Year 11 males: 84/60*50 = 10 Year 11 females: 86/589*50 = 7
I have rounded my figures up I have rounded these figures up they add
they add up to 53. up to 55.
As with every stratified sample rounding errors could have occurred, so they couls have added to 54 or 56.
I will now select my samples from the whole population using a random method.
Cluster sampling: First you split the whole population into groups or clusters. Then you choose certain clusters using random sampling. This method is called cluster sampling. The sample is then every member of the cluster chosen. Cluster sampling is very cheap but it can be biased.
Quota sampling: This is a sample where they try to involve as much of the population as possible. It is a method often used by research companies. The interviewer will normally be given instructions but they then will be left to choose the interviewees.
Attribute sampling: In this type of sampling you would pick your sample based in an attribute which is totally unrelated to the variable being investigated.
SAMPING I WILL CHOOSE
I will choose to use random sampling and stratified sampling because these will make the results less biased as possible and the stratified sampling will give us the exact of people we need to use.