No matter where I have taught, one job that often challenges graduate students is the writing of a research proposal. When first encountered, it is a peculiar piece of academic writing that most undergraduate students may only have to attempt once or twice, if that. But they become increasingly important in graduate school, when writing thesis proposals and grant applications.
Below is a refined version of an explanation of how I approach writing questions for a research proposal, that I have shared in the past with Master’s students. Later installments will address other aspects of the proposal.
Questions are key to any type of research proposal. They make clear to the reader what your research is going to be about, and helps them to imagine the range of things you might do in your research. Formulating and refining your questions is also useful for helping you clarify your own thinking. You will draft many questions that are close but not quite right, and the process of writing better questions will help you figure out what you are actually doing.
Usually, students are asked “What is your research question?” and research questions are certainly important in a proposal. But in my view, the research question is just one of three types of questions that any effective research proposal must address.
The three types of questions are: conceptual questions, research questions, and field questions.
These question types are arranged in a logical hierarchy, with conceptual questions being the most abstract and field questions the most concrete.
These are the “big” questions, posed at the most abstract level. Conceptual questions can include the biggest questions, like “What does it mean to be human?” or “What do people value in their lives and relationships?” but in research proposals, they usually connect to the broadest issues of academic concern in anthropology. They might be “What is capitalism?” or “How do new technologies affect human life?” or “How do people remember the past?” or “How does race shape peoples lives?”
Note how at this stage, these questions do not yet ask where or with whom your research will be done. Neither do they ask how the research will be conducted. Those matters are addressed by the other question types. What’s important in a conceptual question is to be able to express the potential significance of your contribution as a scholar who is doing this research, and not primarily of this specific research. Conceptual questions should suggest what you want to say about issues of general concern, which will certainly emerge out of specific research projects, but which rely more on what it is you want to be able to say about the world as a scholar, and what you want to see change, or argue is changeable.
Therefore, conceptual questions will often be about a big category of things in the world, like “capitalism” or “kinship,” that you want to critically understand. If you think capitalism is messed up and destroying the world, then your conceptual questions should probably be about capitalism and the environment. If you think what really matters to most people is religion, then your conceptual question should be about religion.
Note also that even though you may be doing anthropology, these conceptual questions may be shared with researchers in other disciplines. A sociologist, biologist, political scientist, or psychologist might also ask “What is race?” An anthropologist will answer this question differently, but it allows partial connections and dialog to happen interdisciplinarily.
For an anthropological reader, and for you as an anthropological proposal writer, conceptual questions will also indicate more or less what kind of anthropologist you are. For instance, if your conceptual question is “What does it mean to be “healthy?” a reader might place you as a medical anthropologist. But perhaps you’re actually interested in what people consider to be “good” or “virtuous,” above what they desire as “healthy,” in which case you might be an anthropologist of ethics. Or, maybe you are a medical anthropologist of ethics. In any case, by providing the reader with a sense of what your conceptual question is will allow those who might not know anything about your research sites a way to begin to understand how to position your work within their view of the discipline.
For most research proposals, you’ll only have one or two conceptual questions. You may not even explicitly pose them in your proposal, though a reader should be able to reasonably guess at what yours are. But knowing your conceptual questions will help you develop your other questions, and orient you within the discipline.
The abstract things that you ask about in your conceptual questions must have some concrete manifestations in the world. Anthropologists may talk about “kinship” but what really exists is people doing things with each other in more-or-less repetitive ways that anthropologists have decided can be usefully discussed using a concept called “kinship.” Research questions show the reader where specifically you think these concepts can be fruitfully investigated in the world. Thus, research questions are more concrete questions than conceptual questions: they get down to the specific who and what of your research. They are the key questions you will ask in your proposal and answer in your thesis.
So, if your conceptual question is “How do new technologies affect human life?” then your research question might be: “How is being an avatar on Platform X changing the way users navigate interpersonal relationships?” If your conceptual question is “What is capitalism?” then your research question might be “How is traditional textile production in Taiwan being affected by global trade?”
In anthropology, research questions are usually “How?” questions, even if they are not written in that way. This is because, in the majority of cases, your research will be about ethnographically establishing that some unstudied or understudied relationship exists between one thing and another thing. This is the case in most research. In the natural sciences, a study often tries to understand a causal relationship between one measurable thing—the independent variable—and another measurable thing—the dependent variable. (e.g. How is the motion of an electron affected by the presence of a strong magnetic field?) In anthropology, the relationships being explored are not usually causal, but analogical or otherwise co-occuring, but the point that the key aim of a study is to find the relationship between a thing and another thing that contributes to what that thing is holds, and your research question should reflect what you understand to be an important relationship between them.
Where the research questions are not explicitly posed as “How” questions, the relationship between one thing and another thing is usually present in more subtle ways. To take a classic example, Evans-Pritchard may have asked “What are magic and witchcraft among the Azande?” In this case, the dependent variables are “magic” and “witchcraft” and the independent variable is “among the Azande.” Unwieldy as it is, you could restate the question as “How are magic and witchcraft changed by being among the Azande?”
By formulating a research question in this way, you articulate a starting point for empirical study. For instance, the avatar question above shows clearly that you should start by trying to get information about interpersonal relationships from users of Platform X, and also that you should pay special attention to how people talk about or experience changes in those relationships associated with their use of the platform.
When you are writing a research proposal, you may only have a vague sense of what the relevant relationship is, but you hopefully have some idea. This can come from pilot research, personal experience, or a reading of the literature. In other words, you will have a hypothesis about what you will find in your research. Therefore, another function of a research question is to give you the opportunity to state this hypothesis. A hypothesis is a statement that answers the research question, and which can be tested for validity. For example, you may believe that users of Platform X become more isolated in their offline relationships as they invest more of themselves into their avatar. In your actual research, you can then go and test this hypothesis. Are users really becoming more isolated in real life? If so, is it because of their use of the platform?
Your research question must be posed in a way that a clear hypothesis can be given in response. One thing to be careful of here is that the research question should not pre-suppose too much of the hypothesis. You must be able to clearly separate the research question you are asking from the answer you think you will establish. A poor research question would be something like “How do people deal with the isolation in their offline relationships caused by using Platform X?” This question already implies that the major effect of using Platform X is isolation, and moreover that using Platform X causes that isolation. The question is posed in a way that already takes things you should probably find out for yourself as already known. Taken as a starting point for empirical study, it means that you might end up ignoring all the ways that people might become less isolated offline, or if they do appear to be becoming more isolated, all of the other things that could be contributing to that isolation that have little to do with using Platform X.
In practice, writing a clear research question is tricky, but it becomes easier with experience, and also by knowing how other people construct and answer their own research questions. More on this later, when I discuss literature reviews.
Field questions are the most concrete questions. If your research question is “How is being an avatar changing the way people navigate interpersonal relationships?” then good field questions might be: Who are the “people” you are interested in? What is an avatar, and how do people choose or create them? Who do they create or maintain relationships with, and how? Field questions are what you will end up asking directly to people in an interview, or trying to answer through participant observation, or asking on a survey, so that you can gather information that will help you answer the research questions.
Good field questions will help bring clarity, detail, and nuance to key terms in your research question. A key term in the earlier example “How is traditional textile production in Taiwan being affected by global trade?” is “production.” Field questions about “production” will help you be able to explain what it is in the context of traditional textiles in Taiwan: Who is doing the producing? What activities do they perform to produce? What do people say about their productive activities? What are the types of products they produce? Where and when does production happen? What other types of production are going on? And so on…
A Dynamic Hierarchy of Questions
As I’ve been discussing, conceptual questions are the most abstract, field questions are the most concrete, and research questions come in between. By “abstract,” I mean that they are about things that they are about things that are present generally, if not universally, in human life. This means that conceptual questions are relatively devoid of specific details. They are about the concept “the sacred” rather than about Christianity or Islam or Buddhism. By “concrete,” I mean that they are about specific details: this person, instead of “people”; this ceremony happening at this time and place, rather than “ritual.”
Therefore, you must be able to arrange conceptual questions, research questions, and field questions in a hierarchy from abstract to concrete. In this hierarchy of questions, you should be able to see that at each level, questions are about different things, i.e. “ritual” rather than “a baptism.” It should be possible to see the research questions as a specific manifestation of a general conceptual question, and each of the field questions as manifestations of the research questions.
You should also be able to see how the key terms in those questions are related to each other in a hierarchy of abstract to concrete. So if a conceptual question is about “the sacred” then there should probably be research questions about “Christianity” or “Buddhism,” and then field questions about “baptism” or “enlightenment” or whatever else. (Exactly what terms go at what level in the hierarchy will depend completely on how you design your project.)
There is also a hierarchy in the number of possible questions that can be posed at each level. For every conceptual question, there are hundreds of research questions that could be asked. Each research question should lead to hundreds of field questions. In any single research proposal, you will probably only have one conceptual question, maybe one or two research questions, and many, many field questions. Conversely, it should be possible to see how answering enough of the field questions can lead to answering the research questions, thereby helping to answer, at least partially, the conceptual questions.
In the actual practice of research, the questions at each level while cause you to rethink and revise questions at other levels. To give a trivial example, you may find that users of Platform X do not in fact use avatars at all, but are taking care of virtual pets. In response, you may decide to change your research question so that it is no longer about avatars but about virtual pets, or maybe you will change your question to be about a different platform. In either case, answers to your field questions will have significantly changed your research questions. You will then adjust your field questions to suit the new research questions. This is normal, and is a feature of ongoing ethnographic research. You discover things are not quite what you thought, and you adjust.
Notice, however, that the conceptual question (“How do new technologies affect human life?”) does not change. In some cases, the conceptual question may change, but this means that things were really not what you thought. Or, it might mean that you want to do a different kind of project as a different kind of anthropologist. This happens too, but probably less frequently.
The point is that, when you’re actually doing research, the hierarchy of questions will turn out not to be as rigidly top-down as it might appear in the research proposal. Nevertheless, the research proposal should usually be presented in a way that a reader can follow how certain conceptual questions imply the research questions which imply field questions; and be able to see how answering the field questions will answer the research questions, and then the conceptual questions. This is often what a reader of a proposal means when they say “logical organization” or something similar about a research proposal
The rest of the proposal
In terms of research design, having clarity about your questions and the relationships between them is crucial for building up the justification for your research project. Depending on the type of proposal that you are writing, you may spend the first one to three pages introducing and explaining your questions. If you’re successful in this, then the literature review and methodology sections, which usually follow, should be on firm logical footing. (I plan to address these other sections in other posts.)