So what do we mean when we say ‘research translation’?
In its most simple form, research translation is a systematic approach to convert research findings into practical applications to improve health and wellbeing. Beyond this, the definitions vary widely in interpretation. In the medical field, it is sometimes called “bench to bedside”, meaning the translation of lab findings to real-life recommendations and practices.
Research translation takes applied research and adds action steps to transform findings into practice or policy. Let’s say a research study finds that housing instability is associated with HIV vulnerability. Now that we know this, what practices, programs, and policies should be put in place? To “translate” research findings, researchers might partner with community organizations, advocates, and practitioners to understand what action steps and strategies are most important and realistic. Research translation adds a level of accountability and transparency not always seen in typical social science discourse.
In 2011, SexLab partnered with AIDS Partnership Michigan, the Ruth Ellis Center, HIV/AIDS Resource Center, and Detroit Latin@s on the United for HIV Integration & Policy (UHIP) Project. Using an approach based in community-based participatory research, UHIP aims to understand and begin to address the social and structural factors in the Detroit Metro Area that are shaping the HIV/AIDS epidemic among Black and Latino young men who have sex with men (YMSM) and trans youth. Structural factors include social, economic, and political conditions and issues like transportation, unemployment, homophobia, racism, and housing. We conducted 50 in-depth interviews with people who work in this field and surveyed 429 YMSM and trans youth. The translation stage of this project included creating one-pagers, digestible “data snapshots” that summarize research findings and provide policy recommendations.
For this post, I decided to show an example of research translation by breaking down the process of how we made a data snapshot about transactional sex. This topic was particularly difficult to distill into one page because there are a lot of divergent ideas on transactional sex and the ways it differs and doesn’t differ from commercial sex work. Unlike our other data snapshots that simply sort data into large themes, this one provides a conceptual blue print for how or why situations of transactional sex might occur and how it relates to HIV risk.
Who is your audience and what’s your goal?
When you create any kind of communication, these are some of the first questions you ask. In our case, the audience is those who work in LGBTQ health, community organizations, health agencies, and policy makers. Our goal of the data snapshots is to distill our research findings into common language and terms and share practical applications.
How do you actually do it?
First, we look at our data, and we have A LOT of data. We use a statistical program called SPSS to analyze our quantitative data:
Really exciting stuff! In all seriousness, it’s like detective work. We run different statistical tests to look for descriptions and patterns. Even though we are dealing strictly with numbers, it’s actually a creative process. You need to “ask” the data questions in different ways and then interpret those results.
Our qualitative data from the interviews was transcribed and housed in a program called NVivo. This program allows you to search for specific terms and do thematic coding. Thematic coding entails reading through all the interviews (in our case, over 80 hours of interviews) and marking or “coding” when a specific topic was brought up. So let’s say we’re interviewing an outreach coordinator of a local organization. We make a special note or tag whenever they mention “transportation”, “stigma”, “family”, “discrimination”, “jobs”, “church”, and so on. We take all these tags and organize them into groups of topics called nodes. This helps us go back later and find exactly what we’re looking for. Except it doesn’t always work that smoothly (SPOILER: it didn’t work out smoothly)! Organizationally, Nvivo provides a huge, extremely detailed map of your qualitative data. It looks like this:
For this data snapshot, we went in looking for quotes already tagged with “transactional sex”. Some topics like “access” yield a lot of results, while others are more difficult to search for because the people we interviewed don’t use research terminology like “transactional sex”. Sometimes we have to do word searches in Nvivo. For this one, we also looked for “slept with”, “couch”, “crash”, and related common words. Later on, these quotes get added to the data snapshot to give perspectives from people in the field.
Gather what is already out there on the topic.
Whatever work you are doing, chances are there are already people invested in this work. In the research world, we search for articles, reports, and comb through SexLab’s treasure chest of (free) publications and presentations. We sorted through a presentation on transactional sex that was given earlier this year and decided how much of it was appropriate and feasible to include.
Building a story.
Data needs to be interpreted and understood within specific contexts and its implications for practice. This brings us to a question that we would literally ask out loud: What is the story we want to tell? Who or what is often left out of this story? At this point, we step away from our computers, talk, and sketch:
Iterations, iterations, iterations,
Research translation is a cyclical process. We do a lot of group work at SexLab, especially collective editing, and while it’s time consuming, it strengthens our work. Having multiple hands and eyes on a project allows us to tell a more nuanced story.
People learn in different ways and the way you present data is just as important as what you present. To navigate this, we use multiple types of information like quotes, numbers, and images. Some people are really visual, others are not, some have a high need for cognition (they like to dig apart complex ideas), while others don’t. It’s important to take off your research hat and deeply consider elements of health communication like literacy, numeracy, and layout.
Based on our data, input from community partners, and current research, we create a list of policy recommendations and resources. Later, these get added to the back of the data snapshots. We took our hand drawn sketch and turned it into a first graphic draft:
And after that, more editing…
After more input, we make it to a final version:
Our final steps include more proof reading, cleaning up the document, and adding hyperlinks. You can find out more about UHIP and check our data snapshots at our website.