ABSTRACT
W e now t u r n to a more famil iar use of least squares i n esti ma t ion . T h i s underlies the s ta t is t ical methods of l inear regression analysis and to some extent of analysis of variance. Because we assume the reader is famil iar w i t h these we give here only an out l ine of the key results, together w i t h some discussion of points of in terpreta t ion.
