Like in audio, my preliminary proposed butter/margarine
A/B test would be a first step to establish if a difference can be discerned (and not just my imagination) before later potentially moving on to preference tests.
A test administrator opens and toasts a plain English muffin. They cut the halves into a grand total of 8 equal sections and they are randomized (or else the taster could potentially 'jig saw puzzle together" the edges to reconstruct which four came from the same half). Brand
A margarine/butter (an exact measured amount) goes on 4 quadrants and Brand
B on the other 4 quadrants and placed on a big plate/tray in an order that is only known to them but was also randomly ordered by coin toss, and labelled "Mystery margarine X samples, trials 1 through 8". They leave the room and then I enter. [By never being in the room at the same time, so I can't potentially be influenced by their inadvertent body language and facial expressions, this escalates the test from being just single blind to
double blind.]
My task is to identify, by taste alone, which are actually
A and which are
B. A palate cleanser is allowed, say a glass of water. I write my answers on a score sheet.
If it turns out at the end I get 4 right and 4 wrong (so no better than random guessing) it suggests I
can't truly taste a difference. Whereas if I get all 8 correct it suggests with a p-value of 0.0039 (0.39%) certainty—I just looked it up—I was either
incredibly lucky that day
or more likely a 99.61% certainty I can indeed discern a difference between the taste of
A and
B under these conditions.
If openly
labelled samples of both
A and
B are additionally provided to me on a different plate simultaneously, at my disposal to optionally taste at my discretion to get a "frame of reference for the two target tastes" during my test, this improves the test because it makes it more like an ABX test.
Ever take an ABX test, my friends?
You are about to, albeit a very short single question ABX test instead of a more meaningful 8 or more, by picking either
A or
B to answer the following question:
View attachment 79865
Psst: The correct answer to this single trial test is: