Method, Data, and Ethics

“No one can say what a ‘result’ is in the ‘human sciences’.” (Roland Barthes)

If your theory section is about the reader’s expectations, your methods section is about your competence to address them. Indeed, you have to present yourself as someone who is qualified to challenge your reader’s expectations of your research object. This will be the main point of my talk this afternoon in the Craft of Research Series (or “Weekly Wednesdays” as we’re starting to call them internally). The talks are directed at students who are working on their year-end research projects and I’m rather encouraged by the turnout. It gives me a chance to talk about things that I normally pass over rather quickly in my “How to Write” and “How to Structure a Research Paper” lectures. Today, for example, I get a chance to go into some depth about the difference between “methods” and “methodology” as well as “experience” and “empirical data”. I’ll also get a chance to discuss the ethical implications and limitations of your methodological choices. As ever, you decide how to write by considering who you’re writing for: a serious, engaged and intelligent peer.

Think of your method as what you do to learn about the world, and think of methodology as a way of holding you accountable. “Method” comes from the Greek word hodos for “way or manner” and, more literally, “path”. The suffix “-logy” indicates some sort of “discourse” about (or “account” of) why you did things the way you did (a methodology is a “doctrine of method”). Where exactly did you go to look for answers to your research question? What did you look at? When and how long did you observe? How many people did you talk to? Did you survey them or interview them? What questions did you ask them? Or did you draw on census, polling, or financial data? How did you decide which data to include and which to exclude? Notice that all these are questions of “method” and are answered straightforwardly and factually. At the end of the day, you just have to be honest about what you did. The methodological question is: why did you do it that way? And here you will be invoking the methods literature, including handbooks and the papers you found in your review of the literature. The trick here is to demonstrate an awareness of what might go wrong–sources of error–more specifically, you must anticipate what your reader thinks might go wrong.

Your methods section, then, should be centered on what you actually did to collect and organize your data. But what is data? Here again, we can go back to the Greek roots of the word: the “given”, “in classical use originally ‘a fact given as the basis for calculation in mathematical problems.'” Today, we generalize beyond mathematics and take data to consist of facts given as the basis for any analysis. Notice the difference between this and, say, a fact derived from an analysis, i.e., a fact that is not the basis of something but is, rather, based on something. Both are facts, but what counts as “data” depends on what happens in your analysis. We can say that what is “given” to us can be “taken for granted” in the analysis. What the people you study think or believe will come out as a result of your analysis; but what they said (in an interview or response to a survey) will be a fact that is given to us as analysts. The purpose of your methods section is to give you the basis, if you will, to let the you take your data for granted in this way. That means describing how you collected it in a way that lets your reader trust it.

This sets up an essential rhetorical tension in your methods section. On the one hand, you want to present your approach to data collection in the most persuasive way possible. On the other, you must be honest about what you did. If you say you sent out a thousand surveys that may impress your reader. But if you only sent out a hundred, you’re lying. Likewise, if you say, truthfully, that you sent out a thousand surveys, and report, truthfully, that 80% of respondents were satisfied with their work load, you are obligated to report, just as truthfully, the, let’s say, 48% non-response rate. Otherwise your reader may be mislead into thinking that 800 people said they were satisfied. Of course, careful readers (with high methodological standards) will have noticed you left out the non-response rate. They will find your data less trustworthy, less credible. They won’t be able to take them as “given”. In an important sense, you haven’t “given” them what they need.

(There are lots of other ethical issues to consider. Does your method respect your respondents privacy? Did you secure informed consent where necessary? But I think I’ll need to take those up in a post of their own.)

The important thing is to make sure that the reader gets a good sense of both the potential and the limitations of your data. Whatever conclusions you’re going to reach, you need to make sure that the data you have gathered provides a plausible basis for them. And you have to show your reader that you understand your own limitations. (This is why I always recommend including the “limitations” in your methods section, not saving them for your discussion.) You demonstrate a working understanding of the difference between your personal experiences (planning your research, calling up subjects, travelling to their places of work, setting up the recorder) and the empirical data that this produces if you do it right, i.e., according to the standards that govern your field. If your theory section made a spectacle, let’s say, of your “disciplined imagination”, your goal in the methods section is to persuade your reader that you also have a disciplined sensibility. You look at the world in an orderly and reliable way. As your reader would. In fact, if the reader wanted to, they could see almost exactly what you have seen, simply by following your instructions, and doing as you did.

Leave a Reply

Your email address will not be published. Required fields are marked *