Survey is a very effective tool through which we can make assumptions about a population using responses from a much smaller group of people within that population. In order for those responses to be truly representative, a correct sample size must be used. There are a few factors that determine the required sample size for each audience; we’ve simplified the number crunching with a sample calculator built specifically for healthcare market research. If you’d like to know more about the variables that are used to determine sample size, scroll down and read on.
In order to know how many participants you need to survey, you first have to know the total population you are trying to represent. For example, if you are trying to understand the perception that US cardiologists have about a certain drug, you would need to know how many total cardiologists there are in the US in order to know how many you should survey. We’ve made it easy to get the population without any research: select your healthcare audience from the available options in the dropdown list and we’ll automatically prefill the correct population size (US only). The data is sourced from the NPI registry and other healthcare databases. If you know the population size, you are welcome to enter it yourself.
Other than the population, margin of error and confidence level (see explanations below) determine the size of the sample you should be surveying. We decided to designate these variables as advanced options and pre-selected settings that are commonly used for market research. If you are not sure what the margin of error or the confidence level should be, we encourage you to keep the default values.
Most people are already familiar with margin of error from political polling. For example, if you see a poll result in which a candidate is running at 40%, you will often see a margin of error reported that looks something like +/- 4%. This means that if you were to ask the question of the entire population, the response would be within 4% (in either direction) of the result produced by a sample of that population. Thus the actual result is likely to be between 36% and 44%. The lower you want the margin of error to be (result more precisely reflects the population), the more participants you would need to survey. As a rule of thumb, 5% is a reasonable margin of error that is commonly used in determining the required sample size for market research. For a more detailed and technical explanation, we encourage you to look here.
A confidence level represent the degree of certainty with which you can predict that the population you are trying to study would choose a response within the margin of error of what the population sample has selected in a survey. For example, if a political candidate was polling at 40% and you selected a margin of error that is 4% and a confidence level of 95%, you could say that you are 95% certain that if you were to survey the entire population of interest, that the response would be between 36% and 44%. The higher the confidence level you select, the more participants you would need to survey. In most cases, market researchers use a confidence level of 95%.
