Abstract: A multi-attribute group decision-making model(MAGDMM) based on a rapid self-adjusting method is presented to solve problems that take place in the procedure of conventional Delphi decision making model,such as a complicated and cumbersome processing procedure with a notable outcome lag,biased decision results caused by group psychological factors,etc.This model introduces the comprehensive satisfaction evaluation technology,developed according to biased information sampling model,into the Delphi method to improve its rapid processing and self-adjusting abilities.Moreover,as an illustrative example,the curricula-variable problem in colleges is solved by using this new model.Results show that there are no complex trade-offs when judging decision values and weights,and it significantly improves the performance of convergency.