Third, researcher can consult published tables, such as table 4.1. such tables require the researcher to specify a given set of criteria, namely the precision (i.e., sampling error), confidence level (i.e., entent to which the sampled attribute is reflective of the true population value) , anddegree of variability (i.e., distribution of attributes in the population). With respect to variability, a proportion of .5 indicates the maximum variability in a population; therefore, it is often used in determining the most conservative sample size. For example, if a researcher desired a precision level of ± 5 percent with a confidence level of 95 percent and a degree of variability of .5 in a population size of 300, the appropriate minimum sample size would be 172 participants. When using tables to project sample sizes, it is important to remember that the recommended values in the table reflect the number of obtained responses-not necessarily the number of subjects who should be contacted to participate. Researchers typically contact a higher number of potential participants in order to compensate for the predicted estimate of nonresponses.