ABSTRACT
Ε { [ Υ - ( α + βχ)}2} Fisz’s type II regression criterion 157 E(Y\x) expectation o f variate Y given variable 82
Ε[{Ζω - Ε(Ζω )}2] variance o f j th ordered Z variate 264 E(Zj) expected value j th unranked variate 262 E(Z(j)) expected value j th ranked variate 262 Ε(α,β) expectation as a function o f the para
Ε} φ - Θ )2 mean square error o f estimator Θ (o f parameter Θ)
Ε(μ) - μ μ is an unbiased estimator o f μ 317 Ε2 (α,β) expected squared residual function o f
e x p ( z ) preferred notation for the exponential function
F*(xj') = j/n j f is unranked index o f the j th ordered variate
[ f {F -\p ) )W p{ l -p )/n asymptotic scale parameter o f pth quantile’s m odel
[φ(\χ - μ}/σ)}/ / “ ψ(\χ - μ]/σ)άχ left truncated and translated standard normal
density <t>([y - μ/σ)/σ 99
E(X )]2})3/2 233
ρ σ ^ σ ^ covariance 115
τ' truncation parameter of an indicator function like I^ T ,00\y)
Var([X + Y]/2) variance of the composite variate [X + Y} /2
X common symbol for a variate 27 X**2 common computer notation for X
Y common symbol for a response variate 130 Υ = μ + σΦ~1(Ζ) standard Normal variate generation
Y\x € Ν(ν;σ\χ,β) Y given x is homoscedastic normal with a location parameter that is asso ciated with x as parametrized by slope parameter β