Charting Feelings About Food

After co-launching her own food company with her husband in her homeland of Brazil, Laura Coelho had questions. The company didn’t do as well as they’d hoped, and that led Coelho to a degree in food science.

“I ended up in sensory because I was curious about how you really understand what consumers are thinking about the food,” Coelho says.

She’s now completing her PhD in sensory science at the University of Guelph’s Food Science Department. While working on a project to test a food product developed for those with dysphagia (difficulty swallowing), Coelho had the idea of developing a sensory evaluation test specifically for this group. “The issue we had is that this population can’t come several days for longer sessions, like a focus group,” she explains. “Some of them are elderly, some of them are in treatment.”

She created EmoMap—Emotional Projective Mapping—an emotional mapping tool inspired by projective mapping, a common sensory method where participants group products based on their perceptions of similarities and differences. EmoMap was designed to be done quickly and also can be done online. The results of a pilot study using the tool were recently published in the Journal of Sensory Studies.

Other sensory methodologies for food testing provide a list of emotions and the participants are asked to check them off as they sample a food. They are quick and straightforward, says Coelho, and are also quick for data collection. But they’re also less precise than a list of emotions developed for a particular product, and also, the researchers note, can deliver inaccurate results. By contrast, the EmoMap doesn’t provide emotions; it also doesn’t require a focus group to generate a list. “We didn’t use a list of emotions because we were not sure what exactly we were expecting these people to tell us about their emotions regarding this food,” Coelho says.

The system is easier on the panelists, but it takes more work for the researcher. “After doing the session, the heavy job is going to be done by me,” she says. “I’m going to spend more time analyzing data, but the hard time is going to be mine and not the panelists.’”

Tracking Emotional Responses

The entire pilot was performed online. Sixty-three participants were shown pictures of six foods: tofu, rice pudding, vanilla pudding, meringue, hummus, and carrot crackers. The foods were chosen for their range of textures and sweet and savory flavors. The foods were similar in color because the researchers weren’t looking for visual descriptions or responses—they wanted to focus on expected flavors and textures.

The participants completed a questionnaire weekly for a month. EsSense25 and emotion circumplex, both of which are tools used in emotional measurement that use preexisting word lists, were used as control tests.

They were asked to group the samples according to the emotions they felt when looking at each food. For example, foods that made them feel happy would be near each other and foods that made them feel sad would be grouped close together.

The researchers ended up with 1,887 words and sentences, which they winnowed down to 27 categories, 12 of which were positive and 15 of which were negative.

“I ended up discovering that product-specific methodologies collect more precisely than methods that have a predetermined list for a general population,” says Coelho.

After the pilot, the researchers tested out EmoMap again, but this time with a different group of panelists who came to their lab, sampled real food, and wrote their responses on paper.

The researchers continue to test EmoMap with larger panels, and they are also planning to do a cross-cultural study. Directly translating emotions doesn’t always work because emotional responses to food are linked to cultural background, explains Coelho, so emotional mapping needs to be tailored.

“It’s not easy to translate one list of emotions to another language,” she says. “We hope that the methodology will help this out and [that] it’s going to be easier to collect data in different languages.”ft