Correct and Incorrect research Design: How to Supply the Best Results
- Case study conducted by Edward J. Bloustein School of Planning in New Jersey
- Summary of the results
In a case study of the mental and physical effects of second-hand smoke on pregnant women, it would be understandably ineffective to use a single test group comprised entirely of women in their second trimester. This is because, when you are performing a case study, you need to be sure that you are properly representing the whole of your test group.
In the event of the aforementioned case study, it would be necessary to have a subject base of as many races and ethnicities of women as possible, who were exposed to second-hand smoke for recorded amounts of time through the entire process of their pregnancy. While the women in their second trimester could be of some help to research, they by no means account for a large percentage of pregnant women who are subject to prolonged exposed to cigarette smoke
[...] This group could be made up of, for example women per ?research group? whose partners/families/co-workers/housemates smoke multiple times a day, once a day, 3-4 times per week, once a week, etc They could have daily journals and weekly meetings with physicians and psychologists from the time they discover their pregnancy until after the birth, in order to determine their physical and mental health at every step along the way. When performing a case study that involves a certain of people, it is imperative that as many representations of that is present in the study. [...]
[...] In the study's abstract, we are presented with a summary of the results: Non-Hispanic blacks, non-Hispanic white, and English-speaking Hispanic Americans were significantly more concerned about environmental pollution problems than were Asian Americans and Spanish-language Hispanic Americans (Greenberg 369). These results can, considering all of the methods of random sampling and personal considerations, be called fairly accurate. Perhaps, with a portion that involved the face-to-face sampling of and homeless New Jersey residents, they could have widened their results to truly represent the population as a whole. [...]