Sometimes the answer that you want aren't a matter of binary choice (e.g. what we discover best via Triangle Testing) Sometimes it's just about "which do you prefer?". That's where hedonistic testing comes in. Also, we usually ask testers to do a hedonistic test with those who've successfully identified the different beers from a triangle test for further data. (This content adapted from our book Experimental Homebrewing.)
All things pleasurable are good is the core of hedonism as a philosophy. In evaluation, hedonism is simply the measurement of your enjoyment of a sample. A taster is given a set of samples, with or without information about differences, and asked to rate the samples on a scale of 1 to 10.
The evaluators can rank the sample anywhere they want; there’s no need for unique scores. If an evaluator feels that all of their samples are worth a 10 rating, then that’s their call. These ratings can show you the general preference of tasters. Whereas a number of these other tests focus on discerning single changes, a hedonistic test shows how a person actually trying the beer feels about it in general.
Rating Difference Testing
A rating difference test is like the hedonistic test in that it allows evaluators to rate a beer and assign a score. Unlike the hedonistic test, it focuses on a particular characteristic. Instead of asking, “Do you like this beer?” you ask evaluators something like, “Rate this beer’s hoppiness.” Tasters then assign a score according to the characteristic you ask about on a scale of 1 to 10. Using a rating difference test allows you to hone a taster’s responses and detect patterns that might otherwise be obscured.
Sometimes you’re faced with a set of samples that need to be sorted into discrete positions.
For instance, maybe you have five samples of different levels of dry hopping. You then ask your evaluators to rank the five samples in terms of absolute dry hopping character. No two samples can have the same ordinal value.
You can test other characteristics here as well. For example, by asking evaluators to do things like put two beers in order of color from light to dark, you can determine any color difference contributed by a process like wort reduction or late extract additions. If the evaluators don’t show consistent rankings, you can safely ascribe no discernable difference to the processes involved.
Paired Comparison Testing
Sometimes you’ll want a focused response from your tasters. With a simple question like “Which of these two beers is hoppier?” you can use your evaluators to determine if there’s a pointed directional difference between two samples. If enough evaluators choose the same sample, you know that
difference is significant. However, if there isn’t a clear tendency toward one of the samples, then you know the change between the two versions isn’t overt and can be considered a wash.
The samples don’t always have to be different, either. Some panels use a comparison test to weed out evaluators. They do this by presenting the same sample twice to a taster. An accurate taster should note that there’s no difference.