Cereal Chem 39:16 - 29.  |  VIEW ARTICLE

Multi-Factor Responses of Cake Quality to Basic Ingredient Ratios.

L. T. Kissell and B. D. Marshall. Copyright 1962 by the American Association of Cereal Chemists, Inc. 

The main and interaction effect of ingredient variation on baking quality are difficult to determine by single- factor experimental procedure. An analysis of these effects on volume and contour of lean-formula white layer cakes was obtained, using a central composite design of the Box-Wilson multiple-response surface type. Amounts of five basic formula ingredients were transformed to a series of four linearly independent ratios, variation increments were established, and batters representing combinations of the ratios were baked to fulfill the requirements of the design. Statistical analysis of layer-volume and top-contour scores provided multiple regression equations describing the quality responses to simultaneous variation of the ratios. A series of response surface drawings illustrate the significant ingredient-ratio effects and the areas of formulation in which both volume and contour are acceptable. The effects of baking powder and sugar/water ratios were curvilinear and critical, with optimum volume and contour occurring at a narrow range of the variables passing through the center of the design. A strong interaction resulted in markedly reduced quality if leavening, sugar, and water were out of balance. The main effect of the shortening ratio was linear and nonsignificant, but significant interactions were found with flour and sugar/ water ratios. Quadratic and all interaction effects of the flour ratio were significant and resulted in larger than normal cake volume at higher levels of the variable.