SAMPLING METHODS

SAMPLING METHODS

Probability and Non-Probability Sampling

The final decision you have to make is about the method to use to select the sample. The choice will depend on the:

• ... aim of the survey
• ... type of population involved, and
• ... time and funds at your disposal.

An important distinction is made between probability and non-probability sampling. In probability sampling, every item in the population has a known chance of being selected as a sample member. In non-probability sampling, the probability that any item in the population will be selected for the sample cannot be determined. The methods from which the choice of sampling is usually made are listed below:

Probability sampling:

• ... simple random sampling
• ... systematic sampling
• ... stratified sampling
• ... multistage sampling
• ... cluster sampling.

Non-probability sampling:

• ... quota sampling
• ... judgemental sampling
• ... snowball sampling
• ... convenience sampling.

In the next section we will define, explain and discuss the major advantages and disadvantages of these methods.

Simple Random Sampling

The word random has a definite and specific meaning in the statistical theory of sampling. The dictionary definition of random is "haphazard" or "without aim or purpose", but the statistical definition is: a process by which every available item has an equal chance of being chosen. So simple random sampling is probability sampling in which every member of the population has an equal probability of being selected. For example, looking at the bank survey again and given that the sampling frame is everybody over 18 shown on any electoral roll throughout the UK, everyone on the roll is given a unique number from 1 to n, (n being the total number of people in the sampling frame). Each number is now written on a slip of paper and put in a box. If you want a sample of a thousand people you mix up these slips thoroughly and draw out a thousand slips. The numbers on these slips then represent the people to be interviewed. In theory each slip would stand an equal chance of being drawn out and so would have been chosen in a random manner. It is fundamental to simple random sampling that every element of the sampling frame stands an equal chance of being included in the sample. This method sounds almost foolproof but there are some practical difficulties. For instance, if there are 52 million people in the sampling frame, another method of drawing a sample in a random fashion is needed – using a computer, for example. The most convenient method for drawing a sample for a survey is to use a table of random numbers. Such a table is included in your copy of Mathematical Tables for Students. These tables are compiled with the use of a computer, and are produced in such a way that each of the digits from 0 to 9 stands an equal chance of appearing in any position in the table. If a sample of a thousand is required, for example, then the first thousand numbers falling within the range 1 to n that are found in the table form the sample (where n is the total number in the sampling frame). Many pocket calculators have a built-in program for selecting random numbers.

Advantages - the advantage of this method of selection is that it always produces an unbiased sample.

Disadvantages – its disadvantage is that the sampling units may be difficult or expensive to contact, e.g. in the bank survey sampling units could be drawn in any area from any part of the country.