total population of India is over , crore you can see that the five-yearly survey of the NSSO involves a sample that is only about . per cent or just over one twentieth of one per cent — of the Indian population! But because it is scientifically selected to be representative of the total population, the NSSO sample is able to estimate population characteristics despite being based on such a tiny proportion. researching attitudes towards trade unions it would be important to consider workers, managers and industrialists, and so on.
The second principle of sample selection is that the actual unit — i.e. person or village or household — should be based purely on chance. This is referred to as randomisation, which itself depends on the concept of probability. You may have come across the idea of probability in mathematics course.
Probability refers to the chance (or the odds) of an event happening. For example, when we toss a coin, it can fall with the ‘head’ side up or the ‘tail’ side up. With normal coins, the chance — or probability — of heads or tails appearing is exactly the same, that is per cent each. Which of the two events actually happens when you toss the coin — i.e.
whether it comes up heads or tails — depends purely on chance and nothing else. Events like this are called random events. We use the same idea in selecting a sample. We try to ensure that the actual person or household or village chosen to be part of the sample is chosen purely by chance and nothing else.
Thus, being chosen in the sample is a matter of luck, like winning a lottery. It is only if this is true that the sample will be a representative sample. If a survey team chooses only villages that are near the main highway in their sample, then the sample is not a random or chance sample but a biased one. Similarly, if we choose mostly middle class households, or households that we know, then the sample is again likely to