Cereal Chem 44:253 - 268.  |  VIEW ARTICLE
Optimization of White Layer Cake Formulations by a Multiple-Factor Experimental Design.

L. T. Kissell. Copyright 1967 by the American Association of Cereal Chemists, Inc. 

The effects of simultaneous variation of seven full-formula white layer cake ingredients were studied in a central-composite design of the Box-Wilson type. Six linearly independent ingredient ratios were defined, with five levels of each ratio establishing the experimental space. Seventy-seven treatment combinations were baked to sample the responses: cake volume, top-contour shape, and internal score. Multiple- regression analysis provided second-order equations for computing response surfaces. In the analysis of variance, significant effects on cake volume were found for water, sugar, leavening, and flour ratios and for five interaction terms. Top contour was responsive to the water x sugar interaction; internal appearance was affected by leavening, egg albumen, and flour ratios, and by several interactions. Some response surfaces are presented to show the range of the variables in which superior cake performance is indicated. Test baking confirmed the performance predicted by the equations within the experimental range. Stationary points for the system were located by mathematical methods and tested.