This quantity is minimized for an unbiased estimator (-5[1])that is a function of a complete and sufficient (10[1])statistic by the Lehmann–Scheffé (“LAY-man shef-FAY”) theorem. (-5[1])Ronald Fisher developed a test that generalizes the t-test to more than two groups named (-5[1])for the “analysis of” (10[1]-5[1])this quantity. (10[19]-5[4])For (-5[1])two (10[1])random (10[2])variables (10[1]-5[1])X and Y, (10[1]-5[2])the expected (-5[1])value of (10[1])X times Y (-5[1])minus the (10[1])product (10[2])of (10[1])the expected (10[3])values (10[1])of X (10[1])and Y (10[1]-5[1])gives this quantity’s (10[3]-5[1])“co-” (10[2]-5[1])form. (10[16]-5[1])This (10[2]-5[2])quantity (10[6])is (10[1])the (10[2]-5[1])second (10[3])central (10[3])moment (10[3])of a distribution, (10[2]-5[1])and is equal (10[1])to the mean for a Poisson (“pwa-SAWN”) distribution. (10[3]-5[1])For a random variable X, this quantity is the (-5[1])[read (-5[1])slowly] expected value of X squared, minus the square of the expected value (10[2])of X. For 10 points, the square of the (10[2]-5[1])standard deviation is equal (10[4])to (10[4])what (10[1])measure (10[5])of (10[22]0[8])spread? ■END■