The evaluation of the returned forms had two primary objectives: first, to quantify a correlation between statistical quality assessments and chill filtration / non chill filtration. Second, to analyse the identification rate for chill filtration or non chill filtration respectively. Emphasis was put on obtaining statistical significance for all statements through a high number of samples.
Statements about the ability of connoisseurs to correctly guess the distillery in blind tastings are a by-product of this big blind tasting.
A total of 1,331 evaluation forms were sent back by 111 participants. 12 samples were taken from original bottlings, 12 from independent bottlings. Each sample was sent out in a chill filtered version and a non chill filtered version, resulting in a total of 48 different samples.
Of each whisky three 0.7 litre bottles (2.1 litres) were available for the experiment. 1.05 litres were bottled unaltered. 1.05 litres were chill filtered first and then bottled. The sample bottles were filled by hand and not filled to maximum capacity. This resulted in 26 to 28 samples per whisky for chill filtration and non chill filtration respectively (together 52-56 samples per whisky).
Non chill filtered single malt whiskies are usually bottled at 46% abv in order to prevent clouding under temperature changes. 22 whiskies had an alcohol content of 46% abv. The Aberlour 12 year-old Unchillfiltered was bottled by the manufacturer at 48% abv. An exception was the whisky Ardmore 12 year-old Dun Bheagan, which had cask strength. This whisky was diluted to 46% abv with deionised water before bottling.