4.4. Refining Your Research Question

Now that you’ve posed a working research question in a deductive or inductive fashion, what happens next? First, you need to read as much as possible about your topic. The processing you do of past research is called a literature review, a summary and analysis of the relevant publications on a topic. In Chapter 5: Research Design, we describe strategies for you to get the most out of the existing studies on your topic. As we describe, the literature review is essentially an argument for your research problem and the research question (and proposed research study) that addresses it.

Your working research question will evolve as you engage with the literature on your topic. You will learn new and important concepts that may shift your focus or clarify your original ideas. In fact, you should see the creation of a good research question not as a single moment of brainstorming, but as an iterative process: you create a working question, then do some reading of the existing literature, then refine your question, then do some more reading—rinse and repeat until you’re good. Trust that a strong research question will emerge from this process at the end.

Once you figure out where the gaps in the existing literature lie, and what sorts of contributions your own research can make, you should revisit your research question and see what needs to be tweaked. Make sure to spend adequate time rethinking and refocusing your question. After all, the overall contribution of a research study often depends on the specific question that it asked. Even if you execute the study perfectly—with thoughtful interviewing, precise polling, sophisticated data analysis—your research may turn out to be entirely forgettable if your question was poorly chosen.

To conclude this chapter, let’s go over some pointers for crafting good research questions that are empirical, correctly structured, and feasible to answer:

1. A good research question is actually a question. You can’t say that your research question is “the opioid epidemic” or “oppression.” You need to frame your topic as a question, not just a subject area. As we noted before, this doesn’t mean you literally need to write your research question with a question mark! (Using rhetorical questions generally doesn’t come across well in academic writing.) For instance, rather than stating, “My research question is: to what extent do people find sponge creatures to be morally reprehensible?” you could write, “In this paper, I will examine the extent to which people find sponge creatures to be morally reprehensible.”

2. 

   A good research question is well-focused and clear. You won’t be able to answer everything everywhere all at once, and you should be kind to yourself, as a novice researcher, and not tackle more than you can handle. In addition to striking out any irrelevant information in your question, try your best to limit the number of concepts you mention. Break down broad concepts into their constituent parts and see if you are truly interested in just one of those parts. Of course, you may ultimately decide to keep a wide net in your data collection so that you can understand diverse, complex, and intertwined factors, which is a particular strength of qualitative studies. Even then, however, you should be able to prioritize to some extent—noting particular causal mechanisms or subjective viewpoints you intend to explore in your data collection, for instance, or at the very least having specific examples and categories in mind of what you might investigate. The definitions of any concepts mentioned in your research question should be clear, too. Those definitions do not necessarily have to be part of the research question itself, but they should be somewhere in your write-up if they aren’t obvious by context.

   Let’s say you’re conducting a study of the relationships between a married person’s gender and what household tasks they do after work. While you might have something specific in mind when you say “household tasks,” your respondents might have their own impression of what the phrase means. For instance, they might think that managing finances and performing home maintenance are household tasks, but you might be primarily interested in other kinds of tasks, like childcare and cleaning. Therefore, you’d probably want to say upfront what you mean by the term (perhaps in a sentence following the research question) and also discuss why you defined it in that particular way, based on your own reasoning or arguments made in the literature. Another pertinent consideration with this example study would be how broad or narrow to go. For instance, you might want to focus your research question by picking a specific household task, or category of tasks, that you find personally interesting or that the literature indicates is important—for example, childcare. Alternatively, you could justify having a broader view of household tasks—one that encompasses childcare, food preparation, financial management, home repair, and care for relatives. In general, you should opt to keep your research question tightly focused—most students pose questions that are way too broad—but that decision ultimately rests on the specific phenomenon you’re interested in and the specific research problem you want to address. For instance, your study might be more manageable if you just look at differences by gender in the efforts that married couples put into childcare, but if you ultimately want to understand how the unequal division of childcare responsibilities fits into broader inequalities in household labor, then you wouldn’t want to narrow your question to look at just one category of tasks.

   We appreciate that it can be challenging to be very specific when you are just starting out your reading on the topic. (And again, that’s generally the best way to develop your research question—read as much as you can.) Very broad research questions like “What are the causes of chronic homelessness, and what can be done to prevent it?” are common at the beginning stages of a project. However, as you examine more of the literature on the topic, you’ll gain the background you need to make your question more focused and clear, and you should push yourself continually in that direction. After all, a study with a poorly posed research question is likely to get derailed at some point, given how much harder it is to creatively build upon existing research and effectively employ research methods in the pursuit of an unwieldy and vague question.

   To help you focus and clarify your research question, Table 4.3 lists a few words to watch for when evaluating it. You should avoid these and other vague terms that do not say much; they are clues that your question is not as tightly written as it could be.

   Table 4.3. Words to Watch for in Research Questions

   Words to watch for	How to get more specific
   factors, causes, effects, outcomes	What causes or effects are you specifically interested in? What causes and effects are more important, according to the literature on your topic? What causes and effects have been understudied? Try to specify one or a handful to focus on, or home in on a specific type or category of causes or effects.
   effective, useful, efficient	Effective at doing what? Effectiveness, utility, and efficiency are vague terms. What specific outcome are you expecting?
   etc., and so forth	Be more specific. You need to know enough about your topic to clearly address the concepts within it. Don’t assume that your reader understands what you mean by “and so forth.”

3. A good research question considers the relationships between several variables. As we noted earlier, explanatory research questions involve at least two variables—an independent variable and a dependent variable—given that they are focused on explaining causal relationships. For any explanatory study, however, you will also want to consider the role of other variables that may influence the key relationship you’re interested in. Let’s return to our example study on how gender affects the performance of household tasks. Even if that’s the causal relationship you care about, you can’t ignore the fact that a person’s preferences in this area are shaped by their political and cultural perspectives. For instance, more traditional and socially conservative individuals in heterosexual relationships may be more likely to see household tasks as “women’s work” and may be less likely to expect their male partners to contribute to those tasks. Because gender doesn’t exist in a vacuum, your analysis will need to consider characteristics like political orientation that work together with gender to shape people’s behaviors and preferences. You might therefore choose to refocus your research question on how gender and political orientation affect the performance of household tasks. Note that if you are primarily interested in the independent variable of gender, you won’t necessarily want to give political orientation—a variable of secondary interest—equal weight within your research question in the way we’ve just done. As noted in Chapter 3: The Role of Theory in Research, you might instead use the language of control variables—variables of secondary importance that still need to be accounted for in the analysis—and phrase your research question as something like, “How does gender affect the performance of household tasks, controlling for political orientation?” Alternatively, you might mention moderating variables or mediating variables: “How does political orientation moderate (or mediate) the relationship between gender and the performance of household tasks?” Of course, you don’t want to throw too many variables into your question—remember, a good research question is well-focused!—and it’s ultimately for you to decide whether and how to connect your study’s variables within your research question. That said, if you choose not to mention these secondary factors in the question itself, you’ll probably want to discuss them somewhere early on in your eventual write-up of your study to make sure your readers know that you’ve considered them.
4. A good research question identifies a target population. To reiterate, your study’s target population (or population of interest) is the larger group of people your study is interested in learning more about. It is different from the sample, the portion of that target population that you can observe and that stands in for the population you’re really interested in (we cover these terms in depth in Chapter 6: Sampling). Your research question should be clear about the population it is targeting. After all, a study of the effect of gender on the performance of household tasks would look very different if we conducted it on a sample of older adults or a sample of newly arrived immigrants. The act of conducting research always implies that you have a target population in mind, and you should think carefully about this population as you craft your research question. Usually, you will want to decide your target population based on the existing literature: what groups have been adequately studied, and what groups have been relatively ignored by researchers? One easy way to create a research question is to apply a question asked by previous scholars to a new population, to see if the same findings hold up within that different context. You may even find a single type of community compelling to study because you have reason to believe—based on the literature or personal experience—that the phenomenon you’re studying looks different within that group than it does elsewhere. For instance, if you noticed issues with household tasks in your interactions with first-generation immigrants, you might want to frame your research question around that particular target population.
5. A good research question has scientific importance. Obviously, your research question should be sufficiently interesting to you that you are willing to put in the effort to complete the project. But it is crucial to remember that science isn’t just an exercise in individual curiosity. Your question needs to be important and relevant to the scientific literature in your topic area, too. Scientific relevance can be hard to assess, and we don’t want you to be intimidated by the challenge of doing something original that also builds on past work. (In the next chapter, we’ll take you through the process of identifying a worthwhile research problem in the social sciences that your proposed study can feasibly address—as we think you’ll see, it’s tough, but not rocket science.) At the end of the day, however, the research your question generates should have something new to say that we don’t already know. It can’t just regurgitate what’s already been discovered.

While this last point is more debatable, we also want to propose that a good research question is important to the broader society. This is not necessarily something that academic researchers toiling in the ivory tower think much about, given how focused we usually are on our next conference and next publication. Of course, scientific inquiry has inherent benefits for society, which in theory can use the knowledge obtained to make people’s lives better. In that very big-picture sense, any scientifically important work you do is also socially important—adding another tiny stone to the heap of knowledge that already exists. But we think you should also think about the more tangible impacts your work might have on whatever communities might read your research or be affected by the knowledge gained from it. (We discuss the different audiences for research in Appendix A: Presenting, Writing, and Publishing.) Research projects, obviously, do not need to address all aspects of a problem or fix all of society. But you should never forget that you as a sociologist are studying topics that by their very nature have social significance. They matter to everyday people—even if the public doesn’t necessarily understand or welcome the language you use to interpret that reality, or have access to all the theories and data you as a scientist can draw upon. It is within your power as a sociologist to ask worthwhile questions and find meaningful answers.