Sampling is potentially the most powerful tool to manage the resources available for a research study. While careful design and execution of a sampling plan is key to obtaining reliable results, creative use of sampling can lower the cost of a study.

One of the common questions we hear about sampling is, "How many respondents should I interview to obtain a 95% confidence interval?" There is no immediate answer to this question because many more factors must be considered to even calculate an answer – and the mere calculation of an answer does not address other factors that should be considered to provide maximum power from minimum sample size. The question of sample size requires evaluation along two lines: a) the "scientific" perspective, by which we estimate the statistical margin of error, and b) the "practical" perspective, by which we evaluate whether that error margin is adequate for the purposes of the study.

Good sample design requires integration of the sample specifications with the development of study objectives, expectations, questionnaire, and analytical methods. As we work through these other issues with you, we will keep you informed of sampling implications as well as the potential for using advanced sampling techniques. These may reduce the number of respondents required, with no loss of information.

back to top

Copyright © 2006 nVision Research. All Rights Reserved.