Being able to taste and evaluate your product is critical to getting it ready for market, not to mention meeting your own standards. But what does it mean to taste and evaluate?
The Institute of Food Technologists defines sensory science as “a scientific method used to evoke, measure, analyze, and interpret those responses to products as perceived through the sense of sight, smell, touch, taste, and hearing.” If that sounds straightforward, putting it into practice can be more of a challenge.
Currently, there are no universally agreed-upon standards for distillers who want to conduct sensory analysis. That means that each distillery—and, indeed, each distiller—must decide for themselves how best to implement the tenets of sensory science.
Here, we introduce two major pillars of sensory science: the major testing types and some categories of bias.
Types of Testing
Although there are myriad specific sensory tests that distillers can employ, almost all sensory tests can be classified as one of three main types: descriptive, difference, or hedonic. When distillers are first setting up a sensory assessment, it’s vital that they understand what kind of test to conduct to get the answers they need. If they choose the incorrect test, or if testing methods are blurred together, the conclusions may not be valid.
Descriptive-analysis testing is an important tool that distillers can use to learn more about their own products and how they compare to their competitors.
In some ways, descriptive-analysis testing is the most straightforward of all sensory testing. As the name suggests, the goal of descriptive analysis is to develop a comprehensive description of all the aromas, flavors, and textures of a specific sample. At its most basic, this can be accomplished by simply asking a sensory panel to list out all the flavors that they taste in a sample.
However, because all humans have inherently different tastes, it’s very common for panelists to describe certain flavors differently. For example, one panelist may describe an aroma as “fresh hay,” while another panelist describes the same aroma as “fresh-cut grass.” To mitigate this problem, sensory scientists commonly sit down with a panel before sampling and agree on a set vocabulary. Spirit-tasting wheels can be used as a good starting point for such discussions.
Difference testing is perhaps the most common kind of sensory testing that distillers conduct. The goal of difference testing is to determine whether a tasting panel can consistently identify a sample when compared to other similar-tasting samples.
Difference testing is incredibly important for quality control, and distillers often use it to determine whether process or ingredient changes have influenced product quality—whether for better or worse. The triangle test, in which you present two identical samples and a different one, is one of the most common and statistically powerful difference tests; however, there are many others. Mastering the proper use of difference tests is one of the most valuable tools distillers can have in their arsenal.
The third type of testing that sensory scientists commonly use is hedonic testing. Simply put, hedonic testing is used to measure how much a taster likes a sample.
While hedonic tests can be as simple as asking panelists the yes/no question—do you like this product?—most hedonic testing is designed to determine exactly which aspects of the product panelists like. Typically, this involves asking panelists questions about the product and allowing them to rate each aspect using a nine-point scale—i.e., one for liking it the least, nine for liking it the most.
While hedonic testing seems straightforward, it’s one of the trickiest types of sensory testing because people often have a hard time explaining why they like what they like; it’s easy for panelists to latch onto one specific aspect of a product, even though it’s not the reason they like it. Hedonic testing is also tricky because it can’t be mixed with descriptive testing or difference testing. If hedonic testing is done with one of the other two forms of testing, testers risk skewing the data for both tests.
Despite its difficulty, hedonic testing is a powerful tool, especially when developing new products.
In sensory science, the term bias is used to describe anything that could cause the results of a sensory assessment to be untrue.
Unlike machines, which are the same every day, people’s perception and senses are in a constant state of flux. A drink that tasted good one day may taste bad the next day. This means that if you conduct a sensory test with the same product and the same person on two separate days, you may get completely different results.
The goal of sensory science is to try to remove this uncertainty by reducing bias. Although there is no way to completely remove bias from sensory testing, several methods and techniques have evolved over the years to both classify and reduce the effects of bias.
Below are three of the most common forms of bias and ways to reduce them.
Situational bias is the effect that situations have on how we perceive things.
For example, most people will tell you that smoky whiskey tastes better when sipped next to a fire, and that the best way drink a piña colada is on the beach. It’s common sense that our environment plays a large part in how we perceive certain sensory stimuli.
When designing a sensory assessment, it’s important that the testers take this into account. Sensory tests should be conducted in a space that either eliminates or minimizes the effects that the environment will have on the testers. Many large distilleries have a room specifically designed for sensory testing. These rooms are painted neutral colors and kept at a constant temperature and noise level, all to minimize distractions and create situational consistency.
For smaller distilleries, a dedicated room isn’t always an option. Therefore, it’s important for the distiller to find a space that will be as consistent as possible to minimize distraction and maintain situational consistency.
Serving and labeling samples is also a place where bias can commonly creep into sensory tests.
Humans often have unconscious preference for certain letters or numbers. For example, someone born on the second of a month may have an unconscious preference for the number two. This means that when presented with three equal samples labelled one, two, and three, they would unconsciously prefer the one labelled two.
To combat this, a good practice is to code every sample with a randomly assigned three-digit number. For example: Samples one, two, and three would become samples 620, 231, and 937. This helps to minimize unconscious bias.
Finally, mutual-suggestion bias is another major form of bias that sensory scientists try to limit. Humans are social animals, and we crave acceptance, especially from those we perceive to be of higher status. This means that when testers share their opinions, junior or less-experienced testers will almost always conform to those with the most experience.
This conformity may be entirely unconscious, but it presents a problem because it can skew the data of a test. To curb this problem, it’s common to have testers not talk or interact when conducting sensory tests, but rather to write their opinions out. In instances where testers are required to interact, it’s important to have a proctor who helps mediate the conversation and ensure that all participants give their own, honest feedback.
This is by no means a comprehensive list of biases. There are hundreds, if not thousands, of different types of bias and many methods have been developed to try to minimize it.
Watch this space for more on method to reduce bias, as well as some specific tests and their associated protocols. The goal is to consider options for adopting a more rigorous sensory program in-house—so you can better create the product that you envisioned.