Abstract
Existing theories for the growth of order in unstable systems have successfully exploited the use of a Gaussian auxiliary field. The limitations imposed on such theories by assuming this field to be Gaussian have recently become clearer. In this paper it is shown how this Gaussian restriction can be removed in order to obtain improved approximations for the scaling properties of such systems. In particular it is shown how the improved theory can explain the recent numerical results of Blundell, Bray, and Sattler [Phys. Rev. E 48, 2476 (1993)] which are in qualitative disagreement with Gaussian theories.