YumInside the science of what delights us with OSU sensory labs.

Nothing matters more than taste.

We might pride ourselves on healthy eating, frugal shopping or ethical sourcing. But let’s face it: If we don’t like how a food tastes, we’re not gonna buy it.

That’s why knowing what a target audience finds delicious is valuable — and why OSU has spent more than six decades asking volunteers questions: Is this too sweet/sour/salty? Did you find this soft/firm/crunchy enough? Do you like the mouthfeel? And so on.

Even after 20-plus years of experience, Ann Colonna is regularly surprised by findings. She directs the sensory and consumer testing program at the Food Innovation Center in Portland. Located downtown near Union Station, the FIC is a public service program of OSU and the Oregon Department of Agriculture, created to help the state add value to the abundance of raw agricultural products grown and harvested here.

“It’s a really rewarding job,” Colonna said.

OSU Food Innovation Center in Portland. Photos by So-min Kang

One of the things she enjoys most is helping growers develop new varieties. She’s worked on multi-year projects with pears (did you know Oregon is No. 2 in the nation for pear production?) and hazelnuts (tester comment: “I had no idea hazelnuts could be so different!”).

Some products are less familiar. Testers compared syrup produced by Pacific Northwest bigleaf maples with syrup from sugar maples. Several bigleaf maple syrups were statistically preferred, with testers finding their flavor more complex and higher in quality. (Watch out, Vermont!)

Another test evaluated olive oils: two from the Willamette Valley, as well as oils from California, Italy and Greece. This was the good stuff — early harvest with the strongest flavor — and testers sampled it plain, with bread and over vanilla ice cream (!). Oregon’s bounty was again preferred, showing our state’s products can compete on a global scale.

How to host a taste test

The lab isn’t the only place you can hold a taste test. It’s also fun to have one with friends. But doing it right is more complex than you might think — potential bias abounds. Let’s say you want to compare three kinds of honey. Follow these basic guidelines:

1. Tell guests to come prepared: They should avoid perfume, smoking, brushing teeth, chewing gum, eating candies or drinking coffee for at least one hour before they arrive. (For your part, don’t serve refreshments until after the test.)

2. Ideally, separate the testers from one another. No chatting during the test.

3. Prepare samples so each kind is presented in the same way: same portion size, same temperature. If serving honey on bread, for instance, cut off the crusts.

4. Assign random numbers to the three samples — 122, 413, 709 — and label them accordingly. Mix up the order when presenting them to your testers, so they’re not all tasting the same thing at once.

5. Wait a minute between tests and provide a palate cleanser. The sensory labs typically use water but sometimes also give a single unsalted saltine cracker (for a fruit test, say) or apple slice or grape (for cheese tests).

6. If you want to be really serious about the data you collect, check out the DIY resources OSU offers to entrepreneurial food makers and farmers.

Beyond bringing more yum into the world, the results of these taste tests can sometimes have an unexpected, and powerful, impact. An initiative that got a lot of press recently compared fresh and frozen seafoods. Conventional wisdom is that fresh is best, but in multiple blind taste tests of salmon, albacore tuna, sablefish, scallops and more tasters declared that frozen seafood was as good as or better than fresh — a finding with huge potential benefits for the environment. A third of supermarket seafood is thrown away because it isn’t purchased fast enough, Colonna said.

When it opened in 1999, OSU’s Food Innovation Center was the nation’s only branch experiment station like this in an urban area. (Since then, a few other universities — like University of Hawaii at Manoa — have modeled facilities after it.) A key advantage of the urban location is the availability of taste testers. A food company may be interested in a very specific group of consumers, and to get good results, a test needs to be statistically sound. In Portland, a test with 100 tasters is typical. To find them, the center has a database of about 50,000 people.

OSU’s original lab in Corvallis, the Center for Sensory and Consumer Behavior Research, has a smaller database, around 2,500 people. The minimum needed for a test there is 67. 

Sara Maruyama, ’15, M.Sc. ’21, who manages this center in Wiegand Hall, noted that even the time of day can affect test results. She’s been doing a series of coffee creamer tests and tries to schedule them in the morning. But don’t ask testers to eat ice cream at 10 a.m., she said. “People aren’t ready for it that early.”

6 taste tester tips

It’s easy to sign up as a volunteer tester. First, fill out the online form. (Sign up for the Portland lab or for the Corvallis lab.) You’ll be contacted if you fit a test’s desired criteria. Typical payment for testers is $40 in Portland and $10 in Corvallis.

 

But not all who apply are chosen, and not all who are chosen are invited back. Ann Colonna explains what the labs are looking for:

1. Fit the demographic. The client is looking for “purchase intent”: in other words, how likely you are to buy their product. If they’re targeting people like you, they want to know what you like.

2. Be honest. On screening surveys, some people check every box, thinking this will increase their chances of being chosen. But sometimes those lists include distractors, such as a fake brand. Select one, and you’ll be disqualified.

3. Open up about your preferences. “We’re looking for people who are really engaged, who care,” Colonna said. Go ahead and geek out on open-ended questions to fully describe your opinions.

4. Follow instructions. Most tests begin with a visual assessment of the sample. The saying is true: We eat with our eyes, so appearance can be very important. Take time to do just that. (If you’re heard munching when you’re supposed to be looking or smelling, you won’t be invited back.)

5. Embrace your palate. Both wildly adventurous eaters and peanut-butter-and-jelly lovers are welcome here. “There’s no such thing as having a golden tongue,” said Colonna. “One person’s opinion is just one person’s opinion.” That’s why food companies survey large groups of people.

6. Take time to taste. You generally don’t need to eat everything provided but, bafflingly, occasional testers send back all but the tiniest sip or morsel of a sample. They won’t be invited back, either.

In the summer, Maruyama avoids scheduling in the late afternoon because the testing booths can get too hot. Grumpy testers aren’t inclined to like anything and therefore don’t typically provide helpful feedback to food companies.

The Corvallis sensory lab hires undergraduates like Emily Hull, pictured top left, to help run taste tests. It’s relevant research experience but sometimes means counting peanuts, literally. Making samples exactly the same is so important that when trail mix is being tested, every cup gets the same number of each ingredient, starting with 14 whole peanuts or 28 halves. Finished samples are passed through the booth windows to testers. Photos by Karl Maasdam, ’93

Still, in general, she said, testers come in and leave happy. They’re going to get paid for having a snack, and they’re excited to know that their opinions matter. In July, the Portland center asked kids ages 5 to 18 to taste crackers for a school snack program. When receiving his $40 payment, one 7-year-old jumped for joy.

In this case, testers too young to read questions on the computer had an adult helper. Sometimes kid surveys use emojis. But rest assured, even the youngest testers have strong opinions about taste. An assessment of baby foods, Colonna said, gets down to basics: Did the infant eat it or spit it out?