Determining the Appropriate Sample Size
Sample size refers to the number of respondents who complete a survey or project. The number of individuals initially asked to participate is based on the expected response rate.
To determine the appropriate sample size, locate on the chart below the size of the population in the first column and cross-reference it with the level of confidence and margin of error desired. For example, if one were interested in studying an IEEE Society with 10,000 members at a confidence level of 95% and a margin of error of +/-5%, the resulting sample size of respondents would need to be approximately 370. The sample sizes assume a randomly drawn sample.
Notice that as the population exceeds 10,000 the sample size hardly increases. Although this is contrary to common sense, this has been statistically proven and is common research procedure.
For small populations, less than 200, the general rule-of-thumb is that one needs to have at least 50% of the population respond.
|95% Confidence||99% Confidence|
|Population Size (N)||Plus or Minus 3%||Plus or Minus 5%||Plus or Minus 10%||Plus or Minus 3%||Plus or Minus 5%||Plus or |
Source: Rea, Louis., and Richard A. Parker. Designing and Conducting Survey Research: A Comprehensive Guide. 2nd ed. San Francisco, CA: Jossey-Bass, 1997, page 121
For most projects, we aim to achieve results with a confidence level of 95% and a margin of error of +/-5%. These parameters ensure results that can be used to generalize to the population and to make data-driven decisions.
Typically we expect a 25% return rate when surveying members. Longer, more complicated surveys on difficult topics may have response rates below 10%, while short surveys that are on extremely interesting topics can achieve response rates as high as 40-50%. Therefore, we generally ask 1,000 - 2,000 individuals to participate in our surveys, although depending on the goals of the study, the sample sizes may vary.