ABSTRACT
W e now t u r n to detai led issues bear ing first on model formulat ion and then, i n Chap te r 4, on fi t t ing and est imat ion. For quanti ta t ive variables a central role is played by means and covariances and less commonly by higher-order moments. T h e first two give the basis for a summary of associations when interrelationships between variables are essentially linear and a complete summary of these associations when the special form of joint d is t r ibut ion , the mul t ivar ia te normal d i s t r ibu t ion , holds.
