Random sampling of broad populations is possible for survey research in which slight demands are made on the subjects. For example, most public-opinion polls are able to obtain random samples since they typically ask only a few questions and take only a few minutes of the respondent's time. Demands on the subject are much greater in most educational research; consequently, even if the researcher selects a random sample he can rarely get cooperation from all the subjects selected. (You will recall that when some subjects refuse to participate in a study, the remaining subjects no longer constitute a random sample because persons who agree to participate are likely to be different from those who do not.) For educational studies that employ other methods than survey, such as correlational or experimental research, the demands on the subject are usually much greater and consequently it is virtually impossible to obtain the cooperation of all subjects selected by random sampling. |
0 komentar:
Posting Komentar