When we calculate sampling sizes, we are trying to gain accurate information. In short, a sample size refers to the number of responses we received for a survey. For example, if you’re trying to estimate the amount of people within a certain demographic that smoke, it’s important to find out how many people took your survey because it will then give you more confidence about your results in quantative data analysis.

Typically, your survey needs to have a confidence level (how confident you are that your results are spot on) of 95%, and a margin of error of just 5%.

For example, let’s say for arguments sake that you run a service that has 5,000 subscribers and you want to find out how happy they are with your product. You come up with a survey and feel that a sample size of about 1000 people creates a confidence level of the desired 95% and a margin error of just 5%. Because of this confidence level, you’re sure that 95% of the time your results will fall within your margin of error.

## Does It Matter How Large My Sample Size Is?

The best sample size is one that is larger. The larger the sample size, the more confidence you have. You’re happy, basically. The smaller the sample size, the less confident you are, and the greater your margin of error. You’re worried basically.

This is because the margin of error of a CI (confidence interval) decreases as the sampling size increases.

Therefore, when you’re creating a survey, you have to take into consideration how to achieve the best sample size. The more data you have – ergo, the bigger the sampling size – the more accurate your results are going to be.

If you are confident about the margin of error you’re willing to allow for, you can then conduct a sample size calculation. Essentially, you set the margin of error ahead of time.

If you want to estimate the size of a, say, population for example, you can use a sample size formula like the one below:

Z² p ( 1 — p )

n = ——————

MOE

When carrying out a sample size formula, you need your desired margin of error, desired confidence level, as well as your total number of people.

Now, it’s important to remember that you should avoid using samples that are too large – because they’ll just waste time. As well as cash and resources, you will lose time. A sampling size that is too small will be woefully inaccurate because your margin of error will be typically too large. Thus you may end up with results that are too far wide of the mark.

To get it just right, to get the best sample size needed to forecast the size of a population, you first need to:

Collect the sample data

Calculate the sample mean

The sample mean should be different from the population mean – it most often is. But when carrying out a sample size calculation, we can look at this difference as something of an error. The margin of error is the difference between the population mean and the sample mean.

To avoid having to carrying out a sample size calculation yourself, there are available online survey programs that do all the hard work for you.