Many thanks to all those to date who have participated in beer tastings as part of my M.Sc. thesis research. This research focuses on using un-malted purple wheat to make a Belgian style Wit Bier. Taste tests have been done using the tetrad test. In this test, panelists are presented with flights of beer containing 4 cups. Two of the cups contain Wit Bier made from White wheat and the other cups contain Wit Bier made using Purple wheat. Panelists are asked to nose the samples and taste them. They are then asked to divide the 4 cups into 2 identical pairs. The goal is to see if the human senses can detect a difference between Purple wheat and White wheat beer. In some of the early testing, the beers were brewed with reduced amounts of orange, coriander and hops. Statistically, people were able to detect a difference (taste and aroma) between the beers. Some “high science” GC-MS testing on the beers proved that indeed the Purple wheat beers have elevated levels of longer chain alcohol molecules which explain the flavor difference between the beers. Why exactly the Purple wheat generates these elevated levels of long chain molecules remains the burning question.

The testing that was conducted recently was done on beers brewed with increased levels of orange, coriander and hops. So far, taste panelists are unable to definitively separate the 4 samples of beers into their proper pairings. In other words, the added orange, coriander and hops are covering up those long chain alcohol molecules.

As to how I analyse the data from a testing event…..I use the Binomial function which is:

In this messy looking equation, p = probability that a person identified the correct pairings. That probability is fixed at a value of 33% because believe it or not in a tetrad test there is a 33% chance of you guessing the proper pairings. In the equation q=probability of not identifying the proper pairings. N is the number of test panelists and n=the number of correct responses.

After crunching the numbers, if the resulting value P is less than 0.05, it can be stated with certainty that there is a difference between the beers in the flights.

The most recent data from the evening of Nov 18th, shows P values of between 0.19 and 0.225. In other words, statistically we cannot make the claim that the beers are different.

Obviously the tetrad test contains some imperfections. Maybe with deeply experienced beer judges, the results would be different. Maybe with more panelists participating the results would be different.

But for here and now, I am prepared to say that the average beer drinking patron at a Regina brew pub is unable to detect the difference between a White wheat Wit Bier and a Purple wheat Wit Bier.

This will have profound implications for a craft brewer. If people cannot detect the flavor difference in a Purple wheat beer, does that brewer use less costly White wheat to save $$$ on operating expenses ? Or, does that brewer spend extra money on the more expensive Purple wheat and go on to create a unique, powerful marketing message about using an ancient grain that originates in Ethiopia and now grows in Saskatchewan etc…