How we change what others think, feel, believe and do
Use when it is difficult to identify items using a simple random sampling method (with random numbers).
Use when it is easier to select every nth item.
Identify your sample size, n. Divide the total number of items in the population, N, by n. Round the decimal down. This gives you your interval, k.
Thus for a population of 2000 and a sample of 100, k = 2000/100 = 20.
Put the population into a sequential order, ensuring the attribute being studied is randomly distributed.
Select a random number, x, between 1 and k.
The first sampled item is the x-th. Then select every k-th item.
Thus if k is 20 and x is 12, select the 12th item, then the 32nd, then the 52nd and so on.
In brief: select every nth item, starting with a random one.
A study of people going to night-clubs first determines that there are about 250-300 people in the club (due to fire regulations). A sample size of 30 is selected, giving an interval of 300/30 = 10. A random number between 1 and 10 is generated and comes up with 7. Starting with the 7th person to enter the club, every 10th person is given a brief interview.
Other precautions are taken to neutralize any impact on the study of what time of night people people enter the club.
This method only works if you can sort the items being studied into a sequence in which you can ensure the studied attribute is random.
It gives a handy method when a random number would be difficult to apply or when counting every nth item is simply easier. In the example above, if sequential random numbers were used and the first random number was 250, then you would have to wait for the 250th person for your first item.
Systematic sampling is also called systematic random sampling.
And the big