We do a sample study to determine the true parameter value of the population from the sample characteristic value.
Sample size is the optimum number of observations from the population you intend to study.
As variability increases, the sample size required becomes more to cover the entire variability.
As you need more precision, the sample size required becomes more to get the required accuracy.
How close we want our estimate to be? What is our Margin of Error, ME?
How confident we want to be? What is our α?
What should be our sample size? What is our n?
n≥(z/ME)^2 p ̂(1-p ̂ )
What is proportion we consider now? What is our p ̂?
Refer previous report for similar study (literature survey).
For which value of p ̂, sample size will be maximum?
It is 0.5.
A sample problem will be discussed later in another post.