Yuki and I have known each other for over five years and next to our shared love of coffee, we also love to throw around ideas. This has led to many joint endeavors such as founding a student-run NGO in Munich and presenting our research at conferences. In late 2017, on a visit to Oxford where Yuki is doing his PhD, we were again enjoying some coffee and discussing the environmental impact of food consumption and how people’s attitudes are changing – the focus of my PhD in Munich.
What did make us order blueberry waffles instead of a typical English breakfast with bacon and sausage just now? And more generally, was it just within our circle of friends or were plant-based diets generally becoming more and more popular? When the caffeine kicked in, we knew where we could get answers to at least some of these questions: online food communities. Not only do they shape what we eat at home, they also provide an impressive log of human behavior.
Yuki and I come from very different disciplines: Yuki has a background in physics, data science, and machine learning, while I focus on environmental studies, life cycle assessment, and the food system. This topic intrigued us both: answering questions of food sustainability with technology-mediated behavioral data.
Interdisciplinary sweets
We started wondering what sort of insights could be drawn from online recipe platforms, where users go to search for inspiration and authors post their creations. We started scrolling through the most used German recipe site on our smartphones. Was there a trend toward more sustainable food hidden in the data? Or maybe even in the opposite direction? Could we trace the evolution of user’s preferences in the platform usage data?
The pressure of the food system on the environment, through land-use change, water use and greenhouse gas emissions is increasing due to population growth and the growing demand for animal-protein. In our study, we were therefore especially interested in comparing meat and plant-based recipes, such as vegetarian and vegan ones.
Searching for recipes online
Looking into the data, we found very encouraging trends: Not only were authors posting more vegetarian and vegan recipes than meat-based ones, users of the platform were also using/viewing them more often over the past 14 years, in relative and absolute terms. Even on the microscopic per-user-level, we found the same trend: more users were converting from a diet including meat to a plant-based one, than the other way around and they were also sticking to their new diets. We wanted to triangulate our results using qualitative methods, to add another dimension of insight. We contacted users that had undergone significant changes in their recipe rating behavior and invited them to talk to us. In the interviews, our findings on the role of meat-replacement products as an important step from meat-focused to vegetarian and ultimately vegan diets was confirmed and we learned more about why people change their diets.
Not only did we thoroughly enjoy the project, we also saw how large the potential of online behavioral data for sustainability studies can be. I am looking forward to the next brainstorming session at a café somewhere and the research project ideas that may come out of that.
Find our paper here: https://www.nature.com/articles/s41893-019-0316-0
Gesa & Yuki
Please sign in or register for FREE
If you are a registered user on Research Communities by Springer Nature, please sign in