This week we will consider strategies for selecting samples. In an everyday situation, we often use some examples to illustrate our points. While these examples can be convincing, their representativeness of the overall population is questioned. In other words, they can be special cases so that others can use some opposite examples against your points. In scientific research, we should choose cases that can “represent” the population we intend to study. For example, if you plan to investigate Houston Police Department officers’ attitudes towards the death penalty, you should consider how many officers to be selected for your research and how will they be chosen. In order to have a representative sample, there are lots of situations to consider so that your samples can represent the overall HPD population, not a biased sample. When we plan a survey, one of the key considerations is that we should have at least 100 cases for the survey in order for meaningful statistical analysis. If the population is large (i.e. thousands), OR if there are many variables to be surveyed, we should consider greater sampled cases. At least, each variable should have 10-30 cases on averages. How many samples do you plan to get for your research (explain why)? What is your sampling process? Please describe the process in as detail as possible. While you decide your sampling methods, you should carefully consider all the possible difficulties you may encounter. Please JUSTIFY your sampling process?
