How Response Rates Affect Survey Results

Surveys are used to infer something about a group of people by surveying a representative sample from its population. The alternative, gathering information from all members of the population, usually isn’t economically feasible. In order for a survey to be valid, the sample must represent the population. If it does not, the results can be misleading.

The first step of sampling is to ensure that names chosen are representative of your entire circulation. Random or systematic (nth name) selection, rather than judgment or convenience, must determine who will be chosen for the sample.

Furthermore, to legitimately project survey results to the total population, the people who respond to the survey must still represent that population. Since surveys rarely achieve 100% response rates, there is always the danger that people who didn’t respond might have answered differently from those who did.

If they would have answered differently, the survey’s estimates are flawed by non-response bias.

As an example of this phenomenon, you conduct three surveys from the same population of students, hoping to learn the percentage that received A’s in a recent class:

The example above illustrates that as the response rate of a survey increases, the results will more closely represent the total population—thus, the higher the response rate, the more accurate the survey.

By employing techniques that have been refined over the past 60 years, Readex Research maximizes response rates to make sure the data a survey yields truly represents the population of interest.