## Presentation on theme: "IB Business and Management"— Presentation transcript:

Sampling Methods (HL)

Sampling When conducting market research it would be ideal to use results from the whole market (all customers and potential customers). The whole market is known as the POPULATION A survey of the whole population is called a CENSUS. However this is not always a practical and sensible idea.

Why? Too Big Too Time Consuming Too Expensive

If a business decides not to use a census it must decide who it is, they are going to ask. This is called a SAMPLE.

Sample Size The SAMPLE SIZE is the number of people within the sample
The SAMPLE SIZE will be a trade-off between cost and accuracy

Factors affecting Sample Size
(1) No estimate taken from a sample is expected to be exact: Any assumptions about the overall population based on the results of a sample will have an attached margin of error. (2) To lower the margin of error usually requires a larger sample size. (3) The confidence level is the likelihood that the results obtained from the sample lie within a required precision. Statisticians often use a 95 per cent confidence level to provide strong conclusions. (4) Population size does not normally affect sample size.

Choosing the Sample Once the business has decided on how many people it will survey, they then need to choose who these people will be….. This is done using a sampling method

Random Sample Everybody within the population has the same chance of being picked for research. This is like putting the name of everyone in a “hat” and picking out a percentage of the total to research. Advantages/limitations + Risk of bias minimised The people selected may not be part of the target group Sample sizes need to be large to get meaningful results

Systematic Sampling After randomly selecting a starting point from the population, between 1 and "n", every nth unit is selected, where n equals the population size divided by the sample size Easier to extract the sample than via simple random; ensures sample is spread across the population Can be costly and time-consuming if the sample is not conveniently located

Example….. A firm has a workforce of 1000. Of these: 600 are male
400 are female What percentages are these? If the business wants to do a survey of 100 employees how many males and females should they ask?

Example……. Of the male teachers: 60 are part time 540 are full time
Of the female teachers: 80 are part time 320 are full time What percentages are these? So how many full timers and part timers should be surveyed

Survey Sample Category Sample size Males 60 Full Time 54 Part Time 6
Females 40 32 8

Stratified Random Sample
This is a way of choosing people making sure a fair spread of people are used. This is usually done by splitting up the population into certain characteristics. Once this has been done a random sample will be done on each group of people. This is an advantage from a pure random sample as it gives a fairer spread of the market.

Risk of interviewer bias
Quota Sample Aim is to obtain a sample that is "representative" of the overall population; the population is divided ("stratified") by the most important variables a required quota sample is drawn from each stratum chosen by the interviewer This is a very good way of sampling if exact figures of your market are known; it is like sampling a mini version of the market. Not random, so still some risk of bias; need to understand the population to be able to identify the basis of stratification Risk of interviewer bias

Cluster Sampling Units in the population can often be found in certain geographic groups or "clusters" A random sample of clusters is taken, then either all or a sample of people within the cluster are examined Quick & easy; does not require complete population information; good for face-to-face surveys Expensive if the clusters are large; greater risk of sampling error

Judgment Sampling This is a way of using people who the interviewer feels will give a good response. It is a quick way of sampling and least expensive however the results will probably not give a good indication of the views of the whole market.

Convenience Sampling This is a way of just using people who are convenient to the interviewer to ask. It is the quickest way of sampling and least expensive however the results will probably not give a good indication of the views of the whole market.

Snowballing Surveys/Interviews are carried out with individuals who then suggest other friends/family/colleagues Useful if the business is unable to get hold of appropriate respondents Large risk of bias

Criteria for choice of sampling method
Available budget Required accuracy Time constraints Availability of population

Types of Errors Sampling error Non response error
Data collection error Data analysis error

Sampling Error When there is bias due to the way the sample was selected