In order to conduct a good survey, you would need to do three things. First, you would need a target group. A random sample is best and if the group is rather large, then for the study to be accurate, you would need a good amount of people to participate.
Second, you would need a method to conduct the survey. There are various ways in which to do this. You can send out a questionnaires, do something on-line, or even resort to phone calls. The only caveat is that the questions must not be biased to lead a person to answer a certain way. If you are conscious of this, it should be fine.
Third, since you are doing a survey on the popularity of ice cream flavors, how about these two other options. Ask the local ice cream parlor the most popular flavors. This can bypass a lot of work. Another option would be to talk with the manger who orders ice cream at the local supermarket. He or she can provide easy information.
If the community is small, you might want to interview the whole group. There is an inverse relations. With a small population, the room for error can be greater.
What method do you think would be best to use?
-the testing method
-the case-study method
-the naturalistic method
-the laboratory method
I'm torn between either the testing method or the cross-sectional method..