0000000000390348
AUTHOR
Nerida. Abbott
Critical evaluation of two commonly used techniques for the treatment of data from extract dilution sniffing analysis
The data from the extract dilution sniffing analysis of beer samples have been treated by two methods of analysis to give either «charm» or «FD» values. The results obtained from these two methods were compared and demonstrated that the rank order of intensity of the odor-active regions was different for most panelists when the data were presented as charm rather than FD values. Points of uncertainty observed while using this method such as between- and within-individual reproducibility and gaps in the coincident response five also been discussed
Evaluation of the representativeness of the odor of beer extracts prior to analysis by GC eluate sniffing
The importance of obtaining and proving that the odor of an extract is representative of the odor of the original product from which it was obtained, before analysis by GC-FID, GC-MS, or GC sniff, is discussed. The sensory methods, such as triangle tests, matching tests, and quantitative analysis, used to determine the representativeness of the odor of an extract are described. Beer extracts obtained by three methods were used to illustrate the interest of the sensory tests. A method using a mixture of XAD resins was proved by sensory analysis to give some extracts with sensory characteristics representative of the particular beers from which they were obtained
GC sniffing analysis : olfactive intensity measurement by two methods
The intensity perceived by sniffing after GC elution of six volatile compounds was measured by ten judges using two pieces of apparatus: a PC mouse which is moved on a 60-cm length scale and a rheostat apparatus which measures the finger span. The choice of the components mixed was influenced by purity, elution time, presence in food and known Steven's slopes. The histograms obtained by summing the responses (determined by measuring the areas under the peaks) of ten people show no significant differences between the two pieces of apparatus. Histograms realised in parallel by Charm analysis with three other judges showed great differences between subjects. However, the mean of the three Char…