Let’s suppose that I have to select 10 students from a class of 100 students for an unpopular task.
Naturally, nobody volunteers. I am forced to randomly select the 10 students I need. How do I select the 10 students?
One way is to follow the Lottery method:simply make 100 chits, each with the name of a student written on it. The chits are placed in a container of some kind, given a good shake, and 10 chits are picked. Those whose names are on the chits are selected for the task. This is probably the simplest way to perform simple random sampling (without replacement).
Another way is to use a Table of Random Numbers. In short, one drops a pen or pencil on the table of random numbers. From that point, one follows a pre-defined path along the table, using pre-defined conventions till the required sample is obtained.
Let’s assume that each student has been assigned a number from 1 to 100. A table of random numbers (or its copy) is obtained. A pencil is dropped onto the table, and the precise point noted (row number and column number). That gives us the starting point. From there, we move in a pre-determined manner, say left to right and then top to bottom. We observe the following conventions (Rules):
Blocks of three numbers will be taken.
Numbers greater than 100 will be rejected.
Duplicate numbers will be discarded.
The first 10 (non-duplicate, consecutive) numbers between 001 and 100 will be selected.
Based on the numbers obtained, we can choose the students with the corresponding numbers.
A third way is to use a computing device (calculator or computer) that can generate “random” numbers.
[If you remember what we learnt about randomness, it would occur to you that such numbers should be unguessable or unpredictable. However, a computing device uses complex algorithms to generate the numbers. Theoretically, someone could possibly hack into the system and figure out the algorithm(s). Thus, such numbers can at best be called “pseudo- random numbers”]
As we did in the previous case, we generate random numbers using any of these devices. Pre-determined processes (“rules”) are observed to obtain the required sample.
An important thing to note is that scientific method requires reproducibility. It is, therefore, a good idea to make a note of how exactly you went about obtaining the sample- where the pencil fell; what random numbers you obtained; what rules were followed, etc., etc.
Because it is a basic sampling technique, it is often performed within other probabilistic techniques, too (simple random sampling within the strata/ cluster, for instance).
It is simple, and rapid.
All members have an equal probability of selection.
It cannot be employed when one wishes to obtain a representative sample from a stratified population.
It cannot be employed to sample large populations (thousands or more).
Simple Random Sampling is commonly performed using one of the following methods:
Lottery method: Prepare chits; place in a hat or other receptacle; draw chits till required sample is obtained.
Using a Table of Random Numbers
Using a computing device like calculator/ computer to generate pseudo random numbers.
Simple Random sampling techniques may be employed along with/ within other probabilistic techniques.
Pingback: Stratified Random Sampling | communitymedicine4asses