Systematic sampling (sometimes called quasi-random sampling) is another probability sampling method. It involves the selection of a certain proportion of the total population. Drawing a simple random sample as described above can be very time-consuming. The systematic sampling method simplifies the process. First you decide the size of the sample and then divide it into the population to calculate the proportion of the population you require. For example, in the bank survey you may have decided that a tenth of the population would provide an adequate sample.
Then it would be necessary to select every tenth person from the sampling frame. As before, each member of the population will be given a number from 1 to n. The starting number is selected from a table of random numbers by taking the first number in the table between 1 and 9. Say a 2 was chosen, then the 2nd, 12th, 22nd, 32nd ... person would be selected from the sampling frame. This method of sampling is often used as it reduces the amount of time that the sample takes to draw. However, it is not a purely random method of selecting a sample, since once the starting point has been determined, then the items selected for the sample have also been set.
Advantages – the main advantage of this method is the speed with which it can be selected. Also it is sufficiently close to simple random sampling, in most cases, to justify its widespread use.
Disadvantages – it is important to check. A major disadvantage occurs if the sampling frame is arranged so that sampling units with a particular characteristic occur at regular intervals, causing over-representation or under-representation of this characteristic in the sample.
For example, if you are choosing every tenth house in a street and the first randomly chosen number is 8, the sample consists of numbers 8, 18, 28, 38 and so on. These are all even numbers and therefore are likely to be on the same side of the street. It is possible that the houses on this side may be better, more expensive houses than those on the other side. This would probably mean that the sample was biased towards those households with a high income. A sample chosen by systematic sampling must always be examined for this type of bias.
0 Comments