Whatever survey you are looking to create and distribute, you’ll want to know what sample size you need to reach in order to achieve a desired level of accuracy in your results.
What do we mean by sample size?
In its purest form sample size refers to the count of individual samples or observations in a statistical setting such as a scientific experiment or a survey. More specifically in the context of an online survey, the sample size refers to the actual number of completed responses a specific survey generates.
The influence of different population groups in defining your sample size
Samples, also known as sample populations are essentially clusters of people that represent a larger group of interest - commonly referred to as the target population.
In most cases surveying the entire target population of interest is just not feasible, due to the large volume of people that exist within it. So, by surveying samples of that population instead, researchers can uncover insights and make some presumptions about the behaviours and opinions of that larger target population.
For example, let’s say you were a researcher who wanted to know how dog owners felt about different brands of dog litter. Surveying the entire population of global dog owners would be impossible both in terms of time and money. To gain the views you needed, you would be much better off selecting and surveying a sample population of dog owners that best represented the larger target population.
However, it is worth mentioning at this point that regardless of the survey sample size you ended up choosing, that sample group would need to accurately represent the larger target population in order to yield valid survey responses.
The fine lines that determine a statistically significant sample size
The margins between obtaining or falling short of your desired sample size can be extremely narrow.
If your sample size was made up of too few responses, your resulting data may not be representative of your target population. This would make your results both inaccurate and therefore not sufficiently helpful to inform your decisions.
In contrast, if your sample size yielded too many responses, the analysis of this data could put a huge strain on the time and budgetary resources available to you. Researchers or survey creators using consumer panel services or offering incentives to survey takers could without careful planning be particularly vulnerable to this.
Ultimately before you create and distribute your survey you need to be asking yourself, ‘how many survey responses do I need to reach in order to maximise my available time and budget, as well as ensuring I get the most accurate results possible, which I will be able to take meaningful actions on.”
Using a Sample Size Calculator to help achieve a statistically significant sample size
Fortunately, there are some steps you can take to make this process simpler, with the sample size calculator offering one of the best tools for achieving this.
The SmartSurvey sample size calculator can be used to determine how many survey responses you’ll need to achieve a desired level of accuracy in your results, which is also representative of your target population.
To start working out your ideal sample size, why not try our sample size calculator now.
Further points to consider when determining a statistically significant sample size
The sample size calculator is a great tool to get you started, but it’s not the only determining factor when trying to work the right sample size. Depending on your area of research or the survey you are looking to carry out, there could be further questions you’ll need to ask yourself, which can make a difference to your final sample size. These can include:
What type of statistical analysis will I be looking to conduct with this data? Will I want to compare subgroups? If so, I might need a larger sample size?
What is the probability of the event I’m investigating occurring in this sample population? If no previous data exists, it might be best to use a 50 percent confidence level to gain a conservative estimate.
How confident do I need to be that the true population value falls within my confidence interval?
(Note: with regards to the last two points, it will mean that when the data is a close call, you're going to need to collect more data to be sure you've got it right.)
What available budget do I have for this study? Can I really afford the sample I want or need?
What is the target population size? Is it large? small? finite?
How precise do I need my survey data to be? In other words, how much error am I able to tolerate? This affects your confidence interval, also known as your margin of error
For more information about the ‘margin of error’ including how to calculate it and then what to do next, why not take a look at our ‘margin of error calculator’ page.
Ultimately, there is no single magic formula that will enable you to determine your exact sample size with total confidence. However, by asking yourself the right questions and utilising tools such as the sample size calculator, you will be much better placed to achieve more accurate results and inform your decision-making process going forward.