Humane League Labs has completed a report titled Surveying US College and University Dining Services for Potential Collaboration on Diet Change Research 2017-2018 presenting the results of surveys designed to identify campuses with detailed data on food purchasing habits. I'm glad to answer any questions and would welcome feedback on building these collaborations.
Executive summary: Collaboration with colleges and universities may be a promising opportunity for collecting data on what foods people buy. This data could support research in animal advocacy and other fields seeking to change diet, including reducing the purchase and consumption of animal products. To help build these collaborations, information about the dining services at 66 campuses in the United States was collected. Using this information, we tried to identify campuses which were likely to have detailed information on the food students buy and be willing to collaborate with researchers. We identified three such campuses and 25 campuses that likely do not meet those criteria. Researchers should consider collaborating with the identified campuses, searching for additional campuses as well as alternative research methods that do not require such detailed information.
Thanks for sharing this solid (and concise) paper! I really like research on ways to collect better data.
Compared to grocery stores, "restaurant-style" areas like a college cafeteria seem likely to have artificially inflated prices on meat and meat-free dishes, such that the proportional price difference is smaller (example with fake numbers: chicken is $5/pound and veggies are $2/pound, while a chicken sandwich is $5 and a veggie sandwich $4). I wonder if that difference exists, and if so, whether it inflates meat consumption relative to what students would buy in a grocery store. (Sorry if this was addressed in the paper; if so, I didn't see it.)
Thank you, Aaron! I think your observation that animal product consumption differs systematically between restaurants, grocery stores and other venues is likely accurate. This study mitigated the problem by selecting for campuses where most of the food purchased can be tracked via the dining services, thus providing a more complete picture of individual diets. Of course, these diets may not be representative of the general population but at least a more complete picture of individual diet reduces selection biases between food venues. That said, we didn't find many campuses that met those selection criteria, so future field research will likely need to consider the limitation of sampling only a possibly biased portion of diet.