I‘ve run into three situations lately where I’ve advised people trying to start up businesses to ‘do their research’. What surprised me is that they didn’t know how to do research. They are brilliant at doing searches, and even figuring out what ‘boolean strings’ to use to get quickly to the best search results, but the rest of the research process is a mystery to them. So this is a bit of a primer on the research process.

The research process is the core of the three-component process of exposition (“a systematic interpretation or explanation, usually written, of a specific topic”), which is what virtually all non-fiction writing (and, to some extent, fiction writing as well) is all about. Exposition consists of (a) assembling pertinent facts and information, (b) using analysis and inference to extract meaning from these facts and information, and (c) composing a coherent and compelling statement of the meaning you have discovered. The process is iterative: In doing steps (b) and (c) you often have do go back and get more facts and information. And it is not at all uncommon (and very human) to start with (c) and then go back and get the facts and satisfy yourself that the analysis and inferences that led to your pre-determined conclusion were valid. Even if that means conveniently ignoring the facts and information and analysis and inferences that do not support your statement.

Since research is the lion’s share of steps (a) and (b), let’s take a look at these three steps in a little more detail.

You begin with a thesis (a statement of belief you will set out to prove) or a question (which you will set out to answer). An example of a thesis would be: There is a rigorous 12-step process that can be used to solve any problem creatively and effectively. An example of a question would be: Is there a rigorous process that can be used to solve any problem creatively and effectively? The choice between the two depends on how sure you are that you already have the answer. At this stage no matter what you choose is likely to change as you ‘do your research’, so it would be more accurate to describe your thesis as a hypothesis (a statement you will set out to prove or disprove).

Now you go back to step (a) and start assembling the relevant facts and information. There are two sources of facts and information: Primary (first-hand) sources, that come from your own observation or conversations with knowledgeable sources, and Secondary (second-hand) sources, such as the Internet, books, newspapers and reference papers.

Primary sources are more credible, but they are also harder, more time-consuming and more expensive to obtain. If you have the luxury of being able to tap primary sources, you need to learn how to document your observations (cameras and recorders are helpful), and how to interview people effectively. You also need to learn how to get the interview in the first place. Keep careful records of who you spoke to, and when and where your primary research occurred. And make sure you have your subjects’ permission to ‘publish’ what they’ve told you, and information about them.

Here’s a great summary from MIT on how to conduct an interview.

Secondary sources need to be cited. If you’re doing your research entirely online, this is relatively simple: Bookmark (add to your Favorites folder) each web page you find pertinent. Then, when you say something in your writing that substantially comes from that web page, put a hotlink to the page under the key word or phrase that draws on the source, so that readers who question what they’re reading (either because they doubt the facts or your interpretation, or want clarification) can quickly jump to the source you’ve cited. When I save pages to my Favorites folder I usually put a little ‘note to myself’ at the end of the page name to remind myself how I found the page — it’s common courtesy to thank someone in your article for bringing a particularly useful link to your attention. It’s also useful if you need to backtrack on your research later. If you’re citing a book or a report that’s not published online, it’s normal practice (if you’re publishing online) to put in a link to the author’s home page, a page where the book or report is reviewed, or even an online bookseller’s page about the book — anywhere readers can get more information, or a copy, if they want it. This is the online replacement of the more formal citation process (listing the book or article name, author, date and publisher in footnotes or endnotes, and page number if applicable).

Here’s an excellent link to an excerpt from Online!, a book about online research, explaining how to cite secondary sources more formally.

To determine what facts and information you need, you can draw on the journalist’s (and the detective’s) investigation process. This generally entails asking yourself, and others if you’re also doing primary research, the questions Who, What, When, Where, Why and How. And it entails following leads, clues and trails that come to light as you answer these questions. Whether it’s a trail of online links or a series of interviews of people, each suggested to you by the last, this is simple, patient, logical, bloodhound work. Eventually you will either end up with the facts and information you needed, or you’ll reach a dead end and have to backtrack. It’s time consuming, but investing in it will show in the quality and credibility of your final composition. And don’t forget to document everything along the trail, so you remember exactly ‘how you got there’.

Although you will inevitably have to cycle back to get more facts and information, you’re now ready to go to step (b) — extracting meaning from the facts and information through analysis and inference. These two processes — analysis and inference — are quite different intellectual processes: Analysis is a deductive process, drawing on the left, logical side of your brain, while inference is an inductive process, drawing on the right, creative side. Suppose you have two pieces of specific information: A study that says stories are more easily remembered than bullet points, and another study that suggests that most business presentations use Powerpoint slides with bullet points on them, and that most business presentations are a waste of time because no one remembers their message. From this information you could deduce that business presenters should tell stories instead of showing slides full of bullet points. But suppose you also know that very few business presenters actually do tell stories. What could you infer from that? Possibly that business presenters don’t know that their presentations are ineffective, or that business presenters have never learned the art of story-telling. If the deductive statement (that business presenters should tell stories) is valid, then it makes more sense to believe the second inference (that business presenters don’t know how to tell stories) because if it was untrue, deductively business presenters, by experience, would quickly learn that they were more effective than bullet-point presentations. You’ve used a combination of inductive and deductive thinking to extract meaning (business presenters need to be taught how to tell stories) from the facts and information you’ve gathered. If that was your original thesis (or hypothesis) you now have what you need to write a compelling article in support of it. If it wasn’t, then you need to get more facts and information (return to step (a)), or do some more thinking (step (b)), or perhaps revise your thesis or hypothesis.

I confess that half the time when I’m writing my daily posts I change my hypothesis (and with it, the article’s title) after I’ve thought through the meaning of my research. Since I save the article under the original name when I first start drafting it, my hard drive is full of articles with names that don’t come close to matching the names of the articles they turned out to be about.

mintochartLast year I wrote an article about one excellent way to do this step (b) analysis and inference stuff. It’s Barbara Minto’s Structured Thinking or Pyramid Principle process. There are other ways of doing step (b) but I like Structured Thinking because it’s rigorous and self-documenting. It requires you to create a pyramid, like the one at right, that lays out in detail both the deductive and inductive arguments that support your thesis or hypothesis (or allow you to answer your question), which appears as the top box in the pyramid. The facts and information are at the bottom of the pyramid, and you structure your argument in support of your thesis by working both bottom-up and top-down, until the argument, using a combination of fully articulated inductive (inferences) and deductive (analysis) logic, is air-tight. So if someone challenges your thinking on any particular point, you can immediately point to the underlying facts and information that supports it, and the process you used to deduce or infer your conclusion.

The other neat thing about Structured Thinking is that it makes writing your article much easier: The core part of your article is simply a top-to-bottom reiteration of your thinking process, including your citations. Then you need to ‘sandwich’ this between a compelling introduction and a memorable conclusion. Minto suggests that the introduction consist of:

  • a factual summary of the current situation,
  • a complicating factor, problem or uncertainty that the audience should care about, and
  • the explicit or implied question that this factor, problem or uncertainty raises in the audience’s mind, and which your thesis answers.

This introduction is most compelling if you tell it as a story. In fact, if you’ve done some primary research, then telling your interviewees’ stories in their own words (i.e. with quotation marks), or, if you were an observer of the story yourself, telling it in the first person, is a very powerful way to establish the credibility of your argument and provide a context for your audience to understand it better.

The conclusion, suggests Minto, should consist of:

  • a restatement of the thesis and the key (second pyramid row) supporting arguments,
  • a reminder of why it’s important and what’s at stake, and
  • a ‘who needs to do what by when’ action plan of next steps.

The ‘action plan’ at the end reinforces the value of what you’re saying by confirming that it’s actionable. It also forces your reader to get off the fence — if they’re kind of going along with you indifferently and then all of a sudden you logically set out what they should do if they ‘buy’ your argument, they’re likely to re-examine their own thinking and either agree and commit to these actions, or challenge you, in which case you’re ready with your structured thinking ‘map’.

All of this is, of course, easier said than done. It takes a lot of practice. But if you work at it, you’ll find you become an excellent researcher, and a more disciplined and critical thinker as well. And those are skills with value far beyond the world of writing.

This entry was posted in Working Smarter. Bookmark the permalink.


  1. Thanks for all the insight! It really covers the full spectrum of writting an article from begining to end. I found a great program to help with doing just with the online material. A program called OnFolio is very sweet for keep a paper trail of websites and certain aspects of what you are writting an article on.

Comments are closed.