Aiming Your Research for Effect


Jeffrey A. Butts
September 2, 2015

All researchers want their studies to have an impact on policy and practice, but few do. It’s often the researcher’s own fault — at least in part. Here are some basic strategies for increasing the chances that your research will have an effect.

Research products need to be prepared with a specific audience in mind. If you’re writing for academics alone, you can use complex statistical models and the language of hypothesis testing. But, if you’re trying to reach practitioners and policymakers, you need to speak their language. This means translating your findings for people who may not love equations and 48-cell tables.

I’m often surprised by how many talented researchers fail to communicate properly — not always for lack of interest but because their professional culture discourages efforts to communicate with people involved in the application of research knowledge.

Researchers are taught to gather evidence, analyze it, and report it as if they have no stake in its application. Secretly, of course, most researchers want people to read their work and they hope that their studies have an impact on policy and practice.

A very bright colleague once asked me to review a draft of a report he was writing. I told him that it was indecipherable unless his readers were already deeply familiar with the research literature. When I suggested that he clarify and simplify the writing to help more readers understand it, he responded, “No, I’m not going to dumb it down; the work speaks for itself.” … Speaks to itself was probably more accurate.


By working with good writers and designers, any researcher can learn to translate technical knowledge into more accessible and usable products. Here are a few guiding principles for translating your research for larger audiences:


The act of writing about research findings is akin to teaching, or at least it should be. Even if your study findings are derived from a complex research design, and even if statistical training is required for a full understanding, your results can still be presented in a way that actually helps your audience to understand the material. That may seem obvious, but many researchers never seem to appreciate the audience’s perspective.


Technical information does not have to be presented badly, in tedious run-on sentences that seem to never end, with so many clauses, and excessive punctuation — like semi-colons, dashes, and parentheses — that by the time you’ve reached the end of the sentence, you are feeling bored, irritated, frustrated, and even angry, and you are beginning to lose all respect for the author. … See what I mean? By the time most researchers have finished their academic training, they have consumed so much bad writing that they are actually impressed by bad writing, mistaking complexity for importance. It may be too late for some of us to learn how to write clearly and plainly, but we can always work with partners or hire someone to help.


Very few people will ever see the peer-reviewed journal articles on which academic researchers spend so much time. Until more scholarly journals are brought out from behind the internet pay wall, restricting your publications to peer-reviewed journals is like sending an email to a few dozen friends and hoping that they spread the word.



Even if your study is first published in a journal, the basic findings should be disseminated in accessible policy briefs and data summaries that are available freely and promoted by a series of posts on professional blogs and social media sites. Research is not complete until you have published several short versions designed for a broad audience. Video is often a good idea as well.


Most people learn more efficiently from visual information. Data graphics should be designed carefully for both accuracy and ease of interpretation. Ideally, the complete message of a research report is discernible by perusing the data graphics alone. Data tables are also crucial, but they should communicate effectively to the audience. Tables are not there simply to impress the nontechnical reader with small font sizes and Greek letters.


Graphics should be reserved for the most important aspects of your study that are complex and multidimensional. They should not be wasted on simple, descriptive tasks. Graphics should expedite your readers’ comprehension of complex information. I recommend asking several colleagues who are not otherwise familiar with your study to look at a draft of each data graphic. Ask them what it means. If they pause for more than 3 or 4 seconds and are struggling to summarize the meaning, you need to redesign the graphic.


Web pages should be your primary dissemination platform. PDF reports should be offered to your audience as an extra bonus, not as the main product. More web traffic is shifting to mobile devices every day and PDF files are not easy to read on many mobile devices. If your research is not accessible via mobile, your audience is automatically half the size.



Some investigators need to work in the rarefied world of basic research, but most of us work in the practical world. When we do, we need to deploy the techniques and tools required to do the job well. The first step is admitting that we need help.


[ originally posted on LinkedIn ]