What is Inferential Statistics?

Inferential statistics is also a set of methods used to draw conclusions or inferences about characteristics of populations based on data from a sample.

µ – The mean calculated for a population
σ – The standard deviation calculated for a population

Open Full-Size Diagram ↗

Inferential Statistics

Population : A complete set of data “N”
Sample: a subset of the population, of size n; its usefulness for inference depends on how it was selected.

We do Sampling all the time. Whenever we execute a project, it has to be managed under many constraints such as time, cost, resources, among others. Thus, it may not be always feasible for the project to study 100% of the population to derive its inferences. For example, if we are improving the quality of ammunition manufactured in an ammunition factory, we may not be able to do quality test of 100% of the products. This is mainly because the product will get destroyed after testing. Thus, sampling is used in these cases where only a sample of products is taken in for quality testing and inferences are made for the population basis the result of this sampling.

Another example is lot acceptance in car manufacturing. For illustration, a company might sample 10–15 cars from a lot of 400 and use the results to decide whether to accept or reject the lot. That is not a universal sample-size rule: a justified sampling plan specifies the selection method, sample size, acceptance criteria and the risks of an incorrect decision.

Sampling helps in managing the project by utilizing lesser resources and is still effective in getting results. Sampling by and large is done by all of the organizations and thus, it is an important topic for our discussion.