by peacecup on Thu Oct 18, 2007 1:13 pm
Mark Prince on coffeegeek is writing a series of essays about tamping - might be worth a look. He's presenting some complex ideas.
The whole coffee world seems to be searching for quasi-scientific ways of evaluating methods, equipment, and coffees. There are two very serious potential problems with this:
1. if one is to draw statistically valid conclusions, experimental designs need to be well-planned. This basically means lots of replication, i.e. testing the exact same thing over and over, BUT,
2. The ultimate response variable is usually taste, and in my opinion this is completely non-replicable. Not only does every individual's tastes differ (i.e. between-subjects), but I believe the taste experience of an individual for any given shot of espresso (replicate) is dependent upon a whole suite of variables that cannot easily be replicated. Time of day is one very obvious one, but the list is potentially infinite; how well one slept, what s/he had for breakfast, mood, etc. This is all within subject-variation, and as much of it as possible should be controlled for. Statisticians have ways of dealing with some of this, but if coffeegeeks plan to actually represent the results of their tests as "objective" they have a lot of work to do to meet the scientific rigor.
So, I believe those seeking scientific/statistical methods of rating/ranking equipment, methods, coffees, have a long and very hard road ahead.
Those of us who view espresso as a gateway experience, perhaps seeking nirvana, but realizing that there are many, non-linear, paths by which to reach it, shall always treasure this cup while anticipating the next.
PC
LMWDP #049
Hand-ground, hand-pulled: "hands down.."