+ 2 × Ks{1(yt ≤ g(xt ))- F(g(xt)| Z)}{F(g(xs)| zs)- θ}
+ Kts {F (g ( xt ) | zt ) - θ} {F (g ( xs ) | Zs ) - θ}
= H1T(s,t,g)+2H2T(s,t,g)+H3T(s,t,g)
(A.17)
1 TT
Then let J.[g] =-----------∑ ∑ H.τSs,t,g) , i = 1,2,3 . We will treat
j T(T- 1)hm jT
Jj[Qθ]-Jj[Qθ-CT] for j=1,2,3 separately.
[1] Thm/2 [ Ji( Qθ ) - Ji( Qθ - Ct )] = Op (1):
By simple manipulation, we have
J1(Qθ)-J1(Qθ-CT)
=-----1-----∑ ∑ [H1 t (s, t, Qθ ) - H1T (s, t, Qθ - Ct )]
m 1Tθ1TθT
T(T- 1)h t=1 s≠t
1 TT
= τ,τ nm∑∑ Ks{ [11(Qθ)-F(Qθ)][1 s(Q)-F(Qθ)]
T(T- 1)h t=1 s≠t
-[1t(Qθ-CT)-Ft(Qθ-CT)][1s(Qθ-CT)-Fs(Qθ-CT)] } (A.18)
To avoid tedious works to get bounds on the second moment of J1(Qθ)-J1(Qθ-CT )
with dependent data, we note that the R.H.S. of (A.18) is a degenerate U-statistic of order 2.
Thus we can apply Lemma 2 and have
Thm/2 [J1(Qθ) - J1(Qθ -Ct)] → N(0,c12) in distribution, (A.19)
where the definition of the asymptotic variance σ12 is based on the i.i.d. sequence having the
same marginal distributions as weakly dependent variables in (A.18). That is,
σ12 =E%[H1T(s,t,Qθ)-H1T(s,t,Qθ-CT)],
where the notation E% is expectation evaluated at an i.i.d. sequence having the same
marginal distribution as the mixing sequences in (A.18) (Fan and Li (1999), p. 248). Now, to
show that Thm/2 [ J1 (Qθ ) - J1(Qθ - Ct )] = Op (1) , we only need to show that the asymptotic
14
More intriguing information
1. The name is absent2. A Review of Kuhnian and Lakatosian “Explanations” in Economics
3. Special and Differential Treatment in the WTO Agricultural Negotiations
4. The name is absent
5. The name is absent
6. Industrial Cores and Peripheries in Brazil
7. A Unified Model For Developmental Robotics
8. The name is absent
9. Can a Robot Hear Music? Can a Robot Dance? Can a Robot Tell What it Knows or Intends to Do? Can it Feel Pride or Shame in Company?
10. The name is absent