Statistical analysis as a guide to product improvement.

Sfat, M.R. and Strohm, J.A.

Abstract
Although most improvements to beers are made by a planned change in the process based on a stimulus external to the day to day operations, most breweries accumulate the basic quality control data from which predictive models can be developed. The statistical technique of regression analysis has been applied to various brewing situations in order to produce improvements in the beer. Regression analysis of quality control data from a malting plant produced a predictive model which eventually led to new specifications for steep moisture, germination temperature and watering rate. In another case the technique focussed attention on calcium, oxalate, anthocyanogen, air and carbon dioxide contents as correlating well with haze formation. The method has also been used in the improvement of beer flavour and gave one brewery control over beer diacetyl content by showing the necessary control of the raw materials.
Keywords: brewing quality statistics