Previous Chapter Chapter 84: Sample Sizes Versus Usable Observations Next Chapter
Chapter 84: Sample Sizes Versus Usable Observations
An issue of concern in estimating sample sizes that is rarely discussed but should be considered is which observations are actually usable. Researchers can estimate the number of sampling units that must be surveyed. But often not all the initial responses are usable because of inactive mailing addresses, telephone numbers no longer in service, incomplete responses, etc.
A question can be raised concerning the sample size value that is incorporated into many of the statistical formulas used to estimate the defined target population's parameters from a survey's sample statistics. For example, if in the sampling process researchers estimate that one sample size – say, 1,850 – is needed but fewer sampled units – say, 1,500 – are ...