## What is Randomness?

- Details
- Written by Omkar

Randomization of sampling is essential if you want your sample to be representative of the true population. The laws of nature underlying statistical theory ensure that any random sample from a population results in the sample being representative of the population. Without this randomness the sample will not be representative of the population. For sampling to be randomized,** each individual in the population that is going to be sampled must have an equal chance of being selected**. In a particular sampling design, the less likely this is to happen, the lower the degree of randomization of the sampling process and the less representative a sample is to the population being sampled. This basically means that if the sampling process isn't actually random, the data you collect is not a true representation of the characteristic of the population. For example if you want to know what the average weight of all the frogs in an area is, you have to sample in such a way that each frog you select has an equal likelihood of being selected for collecting data on it's weight. Only then will the average weight of the frogs I have sampled be close to the true average weight of all the frogs in the area. If the sampling is not random the result you get i.e. The average weight of the frogs you have selected will not be close to or similar to the true value of the average weight of the frogs in the area.

How do you randomize a particular sampling methodology?

About the Author of this Article

References

McDonald, J.H. 2009. Handbook of Biological Statistics, 2nd ed. Sparky House Publishing, Baltimore, Maryland.