In any field of research involving the participation of humans, many complicating factors present problems that need to be addressed before research can begin. These questions are not unique to sociologists and sociological research but are present in studies involving economics, education, and other similar fields. Since most forms of research include elements of human behavior, being transparent about these problems is critical to research being credible and applicable.
One potential problem is the subject the researcher chooses. By identifying a problem and selecting a specific subject of study, the researcher immediately signals a bias. There may well be a long-established foundation of research on the subject—however, the researcher’s selection of a topic indicates the researcher believes more needs to be done or believes their line of inquiry will add nuance to the already published body of data. Addressing why a researcher chooses to pursue a particular study is essential in interpreting their conclusions.
Selecting target populations is risky as well. Research is not very good at generalizations, which means every element of a research study has to be specific to the target population and related to the hypothesis of the study. The selection of the population under investigation is crucial but can be limiting if a researcher desires to address a significant issue and apply the findings on a large scale. For example, a researcher might be conducting studies on why people become homeless. Within the homeless population are different sub-groups, whose reasons for being homeless vary (though there will also be overlap in these sub-groups). This shows that any given population cannot be generalized: resolving the homeless issues related to a veteran, for example, is not necessarily the same as addressing the problem for a single mother raising children in a rural area. Researchers have to avoid overgeneralization, or the application of conclusions to large swaths of a population may lead to incorrect assumptions.
Research is not cheap. It can be costly, and finding funding is often problematic, especially for studies addressing human behavior. Human behavior research requires a significant pool of people (all of whom must be paid) to conduct interviews, surveys, or other canvasses of the population to obtain the results. Think of the US Census as an example. You may have thought the only purpose of the Census is to count people living in the United States. When you look at the data collected, though, you can see the Census is far from just a count. There is a significant amount of sociological data compiled, which is then interpreted and applied to several different situations. The scope of the work requires a substantial workforce to obtain the information needed to address the problem.