In market research, the purpose of selecting a sample, rather than the whole of your target audience or target population, is to reduce cost and simplify the process.
When selecting a sample size it is important it represents the whole of your target population. Careful consideration should be given to the selection criteria, so that you create a representative sample. The size of the sample also needs to be sufficiently large to avoid any bias in the results.
So how do you know you have selected an adequate sample size? And how do you calculate the margin of error in your data so that it truly represents the whole? One way is to use a sample size calculator. In this blog we take a look at five popular online calculators, offering a review on each. We also provide an explanation of the data terms, so you understand what data you need to enter to get a meaningful result.
You can use sample size calculators to:
- Determine the number of participants you need to survey
- The number of respondents required to achieve an accurate representation of your target market.
- The confidence Interval or margin of error you need to account for when calculating your sample size
- The margin of error to include when presenting your survey findings
When using sample size calculators you will need to know some of the following points to get a valuable result:
- Population size – This is the total number in your target population. So the total number that your sample is representing. Sometimes this number is ‘large’ (greater than 100,000) or too large to be known. Probability mathematics shows that in this case the number is irrelevant, as long as the sample size does not exceed a few per cent of the total. Population size does need to be included when the number is relatively small, such as when you are surveying a membership organisations or a customer base of contractors.
- Confidence Interval or Margin of error – Margins of error must be allowed for when calculating a sample size, otherwise the sample will no longer be representative of that target population. This is calculated in percentage and is usually between 3% and 6%. This percentage needs to be taken in consideration when presenting your findings. For example if your margin of error is 4% and 47% of your sample picks an answer, you can be ‘sure’ that if you asked the same question of the total target population that between 43% (47-4) and 51% (47+4) would have picked the same answer.
- Confidence levels – This is commonly 90%, 95% or 99%, the most popular being 95%. It is the chance that the true value will be inside the confidence interval/margin of error calculated. In other words how confident you are about your margin of error calculation?
- Response distribution or percentage of accuracy – This is the number that gave a survey response or if yet to complete the survey the number you anticipate will give a response. This does involve some guess work. You can use past experience and/ or your knowledge of the target population. The nature of the survey can also affect response rates, is it a long survey, is it online or via telephone…all these elements will have an influence. Working this out allows you to understand how many people you need to survey in order to get your desired number of responses. If you are unsure then using a 50% response rate is one way to achieve a sufficiently large sample size, so to ensure accuracy of data.
Of course, if you are finding all of this confusing, you could just ask a research agency like Competitive Advantage to do it for you!
Sample size calculators reviewed:
Ease of use
Provision of supporting Information
Calculator options /
How practical are the results?
|Survey System||Confusing to use as the confidence interval needs to be worked out to calculate the sample size and vice versa.
||Clearly set out but needs to be written in a more straight forward manner.||Two calculator options, one to determine sample size and one to determine confidence interval/margin of error.
||Gives a sensible result, providing a manageable sample size.
|MaCorr research||Confusing to use as the confidence interval needs to be worked out to calculate the sample size and vice versa.
||Clearly set out, in plain English, giving details of each stage of the calculation.||Two calculator options, one to determine sample size and one to determine confidence interval/margin of error.
||Gives a sensible result, providing a manageable sample size.
|Raosoft||A simple format, presented on screen with no need to scroll. Suggestions are given on how best to complete the fields.
||No detailed explanation provided, just a one liner explanation. The more information section is confusing.
||One calculator is provided to determine sample size.||Provides alternative scenarios, on your results, demonstrating what they would be if you were to increase or decrease your confidence level / sample size.|
|Australian national statistics service||Presented in a clear and easy to use format. With guidance on which fields to complete, to get the data you need.||Introduction is confusing but overall the calculator is clearly presented, with step by step bullet points to follow. Also an explanation of the categories and an explanation of the results is provided.||It has one calculator that can be completed in different ways to determine your results. It clearly indicates which data to complete to find sample size and/or confidence interval.
||Gives a sensible result, providing a manageable sample size.|
|Custom insight||Straight forward to use and presented in a down to earth format that is jargon free.||Instructions written in a straight forward manner. Logically presented.||The first calculator takes you step by step through how to determine number of respondents needed, the next sample size and the final calculator how to determine the accuracy of your survey results.
||Gives a sensible result, showing sample sizes for several different confidence levels. You can also calculate what the confidence interval or margin of error would be for different confidence levels.
As a result of our review, Competitive Advantage Consultancy decided to develop its own independent calculator to provide a practical and meaningful result. Following standard sample size formulae, we work through a check-list, to accurately represent the target population in a meaningful sample size. We then balance this against the practical aspects to propose a sample size which is representative and affordable. It is this calculator that we refer to when proposing sample sizes for client projects.
If you would like help with your market research then please contact us for a no-obligation proposal.
Competitive Advantage specialises in providing marketing and sales services for the construction sector. We aim to provide high quality information at an affordable cost.