Every Project Needs Its Own Research

A key part of our process in a behavior change design project is to do a literature review.  We comb the published peer-reviewed literature to find research that will help us understand the current project. For example, on a recent project where we wanted to design a wellness app for people on Medicaid and Medicare health plans, we looked at research on how social determinants of health (SDoH) affect access to wellness services and care and outcomes associated with community-based health and wellness models. The information we learn from the literature review helps us shape our own research by understanding what, if anything, we might be able to apply from the previous work, the types of questions we might want to ask, and the types of solutions that have worked on similar problems.

What’s key is that we do still do our own research that’s designed specifically for the project. We have to. Why?

There are three key tenets to keep in mind when designing for behavior change.

  1. People are different
  2. Context matters
  3. Things change

Every project I work on is designing something for a unique combination of people, problem, and product within a particular historical context. That means that there is never a pre-existing research study that provides the precise set of insights that will lead our team to the best solution. We can learn a lot from previous work, but I can’t think of a single example where conducting our own focused research didn’t help us come up with something better, more tailored, more right.

Even if we did magically find an existing study looking at a very similar problem space to a given project, chances are the people who participated in that research don’t look like the people we’re designing for. In particular, the existing body of research, especially in psychology, suffers from being WEIRD. That is, the people who participated in the research are largely from Western, Educated, Industrialized, Rich, and Democratic societies, which means any findings are filtered through that kind of worldview. On top of the WEIRDness, many research studies focus on undergraduates as the participants, which is due to convenience and affordability and is deserving of its own critical scrutiny. And the average undergraduate is going to be young, quite possibly a full-time student, literate and numerate, and otherwise not necessarily like the average person. To learn about the people we’re designing for, we have to talk to them directly, now.

To go back to my recent project on wellness apps for Medicare and Medicaid plan members, we could look at population health trends (group/macro level) next to the efficacy of digital health interventions (individual/micro level) but the latter was usually not tested with people resembling our design targets. When we actually gathered together plan members for research, our rooms included immigrants and first generation US citizens, people spanning from 18 to 70+ years of age but skewing older, people who had left school without a high school diploma as well as people with graduate degrees, some people who struggled with English, and all manner of professional backgrounds and skills. We would not have learned from the published research, for example, that one reason why eating healthily is an issue for some people is that when they get off the overnight shift and need dinner, the only available takeout options are fast food breakfast sandwiches. That insight opens design opportunities around food service that could not have been gleaned from other types of research.

When we design products and services, we’re designing for a diverse audience who live in a very particular moment of time. That’s why existing research, while important and valuable for framing up a project, is only a start. So get out there and talk with your design targets.