In Chapter 6 of Luker's (2008), Salsa Dancing Into The Social Sciences, a method of researching called sampling, where random subjects are observed and compared to each other, is discussed. Luker states that researchers use sampling as a form of data collection "because there is no way [they] can gather all of the possible bits of information that would illuminate research question[s]" (p. 101). Personally, I find that this method carries a great danger of resulting in narrow-viewed conclusions. If only several random subjects are observed, one cannot see the big picture in how they relate to whole population of subjects. Those selected for sampling may not represent what is wide-spread or accepted as the norm in the population. While I agree with Luker that sampling would save time and energy for researchers (p. 101), those factors should not compromise the quality of research that one must do to answer their research question accurately.
Luker goes on to say that sampling is not meant to represent the larger population, but rather the larger phenomenon (p. 103). In this case, researchers would have to pick and choose their subjects in order to ensure that they are relevant to the cause at hand. This, however, could certainly result in the manipulating of research conclusions because if researchers have control of those they study, they could promote their own agenda in getting the results they are looking for.
Luker, K. (2008). Salsa dancing into the social sciences: Research in an age of info-glut. Cambridge, MA: Harvard University Press.