Our sample size calculator is the perfect tool to let you find the perfect sample size for your surveys – without needing to getting bogged down by maths!
Here are the best practices around sample sizes and a free sample size calculator for you to use.
It will do the hard work for you!
Free sample size calculator
By using our sample size calculator, your survey is more likely to deliver accurate data that you can rely on.
Your research needs to reflect the wider population of your target audience – and choosing the right sample size plays a big role in this.
Our calculator offers the perfect solution. Just tap in the numbers and you’re ready to go.
Created by Snap Surveys – delivering excellence in survey software and experience measurement solutions since 1981.
Why should I calculate the perfect sample size?
Calculating the perfect sample size is a way to save time and money by interviewing fewer people, without compromising the accuracy of your results.
For example, if a large supermarket chain is looking for feedback on the shopping experience they offer, they can invite customers to answer a survey.
But do they want to invest in surveying customers in every store around the country? That requires a lot of time, effort and money because of the scale of the operation.
Instead, they can just select a small group of stores to run the survey. The amount of responses they receive will be the sample size, and can be used to represent the views of the entire customer base.
But how many responses do you need for a true representation of all the customers?
Below are the four important considerations for determining the perfect sample size.
What makes a perfect sample size?
To find out the perfect sample size for your survey, you need to know a few other things first.
- Population size (overall)
How many people will your survey represent? In our earlier example, the supermarket chain would consider their estimate of total customers around the country.
- Margin for error
How likely are your results to reflect the views of the wider population? By calculating your margin for error, you can use it to establish a percentage range for how accurate your results may be – such as between 68% and 72%.
- Sampling confidence level
This is a percentage that tells you how confident you are that your sample results can be achieved consistently. If you ran the survey 100 times, how many times would the answers fall within the margin of error? If this would occur 90 times out of 100, then your sampling confidence level is 90%.
- Response rate
How many participants will fill in your survey to completion? Sometimes requests get ignored or surveys only get partially completed. A good response rate is a key component to quality data that allows you to see the bigger picture.
Weighting your sample size
Survey data weighting – when referring to respondents – is a way of correcting issues in the demographics of those who respond to a survey.
For example, if a theme park is running a visitor survey, a representation sample of visitors is interviewed. The theme park will be able to identify the total number of visitors that day through ticket sales.
To weight their sample size, they just need to factor up to the total number of daily visitors. If 1,000 visitors were interviewed, and the total daily visitors was 10,000, then each visitor counts for 10 people.
It should be noted that there will likely be different profiles of visitors on different days of the week, so this will also need to be taken into account.
Increasing the value of certain responses helps to ensure your sample size data reflects the overall population.
Find the middle ground
Larger sample sizes = time and money
You don’t always need a large sample size.
While a bigger sample size lets you increase your confidence level or reduce the margin of error, it also means you need to survey more people. This can require more resources.
If time, money or manpower is of the essence, consider whether a smaller sample size is viable – perhaps in a more targeted way.
Small samples can be misleading
As we’ve said, sample sizes that are too small (or ones without the wider context taken into account) can provide misleading answers.
It’s important you get it right.
So what’s the middle ground?
Our calculator will help you find it! Scroll back up to use it.