An environmental group wants to use a survey to determine the extent to which its members would be willing to protest a recent decision of the governor of a particular state. Which of the following procedures would be most appropriate for obtaining a statistically unbiased sample?

A. Survey a selection of people whose names are randomly chosen from the telephone directory.

B. Have signups of which workers gather signatures at local malls throughout the state.

C. Survey members attending the annual convention of the environmental group.

D. Survey a selection of people whose names are randomly chosen from a list of members of the environmental group.