[CH] Re: Scovies

George Nelson (70431.3065@compuserve.com)
Sun, 24 Jan 1999 12:31:50 -0500

Dave Anderson wrote: "...the Scoville Organoleptic Test...This test used a
panel of five human heat samplers who tasted and analyzed a solution made
from exact weights of chile pepper dissolved in alcohol and diluted with
sugar water. The pungency was recorded in mutiples of one hundred "Scoville
Units" A majority of three of the tasters had to agree before a value was
assigned to a given chile or food;"

One important concept omitted from the above is that Scoville diluted his
samples way out before presentation to his panel.  They were asked simply
to say whether the sample had any detectable heat at all.  The Heat Units
were the dilution at which the heat was barely detectable by 3 of the 5
panelists.  It is interesting the resolution of the method was to the
nearest hundred SHU.  This implies three, or maybe even four significant
digits, though I would be surprised to have more than two significant
digits.  However, if range of the samples is from 0 to less than 10,000
SHU, it could be possible to get to the nearest 100 units (2 significant
digits) in a maximum of three tastings, each one narrowing the range of
SHU.

George Nelson
70431.3065@compuserve.com