Bias of the structural quasi-score estimator of a measurement error model under misspecification of the regressor distribution

In a structural measurement error model the structural quasi-score (SQS) estimator is based on the distribution of the latent regressor variable. If this distribution is misspecified, the SQS estimator is (asymptotically) biased. Two types of misspecification are considered. Both assume that the statistician erroneously adopts a normal distribution as his model for the regressor distribution. In the first type of misspecification, the true model consists of a mixture of normal distributions which cluster around a single normal distribution, in the second type, the true distribution is a normal distribution admixed with a second normal distribution of low weight. In both cases of misspecification, the bias, of course, tends to zero when the size of misspecification tends to zero. However, in the first case the bias goes to zero in a flat way so that small deviations from the true model lead to a negligible bias, whereas in the second case the bias is noticeable even for small deviations from the true model.     

Completeness and compactness for varieties over a local field

This paper shows that for a local field K, a subfield k ⊂ K and a variety X over k, X is complete if and only if for every finite field extension Kʹ | K, the set X(Kʹ) is compact in its strong topology. The author likes to thank Florian Pop, Jakob Stix, Stefan Wewers, Gunther Cornelissen and his own parents for their support. Received: 13 April 2006     

The Discounted Limit Theorems

The large deviation theorems, exponential inequalities and a non-uniform estimate of the Berry–Esséen theorem in a discounted version are proved.Dedicated to Professor Vytautas Statulevičius on the occasion of his 75th birthday.     

A study of the effect of ZrO2 on the magnetic properties of FePt/ZrO2 multilayer

Fe_xPt_100−x(30 nm) and [Fe_xPt_100−x(3 nm)/ZrO₂]₁₀ (x = 37, 48, 57, 63, 69) films with different ZrO₂ content were prepared by RF magnetron sputtering technique, then were annealed at 550 °C for 30 min. This work investigates the effect of ZrO₂ doping on the microstructural evolution, magnetic properties, grain size, as well as the ordering kinetics of FePt alloy films. The as-deposited films behaved a disordered state, and the ordered L1₀ structure was obtained by post-annealing. The magnetic properties of the films are changed from soft magnetism to hard magnetism after annealing. The variation of the largest coercivities of [Fe_xPt_100−x/ZrO₂]₁₀ films with the Fe atomic percentage, x and differing amounts of ZrO₂ content reveals that as we increase the ZrO₂ content we must correspondingly increase the amount of Fe. This phenomenon suggests that the Zr or O atoms of ZrO₂ preferentially react with the Fe atoms of FePt alloy to form compounds. In addition, introducing the nonmagnetic ZrO₂ can reduce the intergrain exchange interactions of the FePt/ZrO₂ films, and the interactions are decreased as the ZrO₂ content increases, the dipole interactions are observed in FePt/ZrO₂ films as the ZrO₂ content is more than 15%.     

On the genericity of chaos

Let M be a compact manifold with dimM?2. We prove that some iteration of the generic homeomorphism on M is semiconjugated to the shift map and has infinite topological entropy (Theorem 1.1).     

Prospective and retrospective analyses under logistic regression models

In logistic case-control studies, Prentice and Pyke (Biometrika 66 (1979) 403-411) showed that valid point estimators of the odds-ratio parameters and their standard errors may be obtained by fitting the prospective logistic regression model to case-control data. Wang and Carroll (Biometrika 80 (1993) 237-241; J. Statist. Plann. Inference 43 (1995) 331-340) generalized Prentice and Pyke's (Biometrika 66 (1979) 403-411) results to robust logistic case-control studies. In this paper, we extend the results of Prentice and Pyke (Biometrika 66 (1979) 403-411) and Wang and Carroll (Biometrika 80 (1993) 237-241; J. Statist. Plann. Inference 43 (1995) 331-340) to a class of statistics and a class of unbiased estimating equations. We present some results on simulation and on the analysis of two real datasets.     

On the moduli spaces of multipolygonal linkages in the plane

One can form a polygonal linkage by identifying initial and terminal points of two free linkages. Likewise, one can form a multipolygonal linkage by identifying initial and terminal points of three free linkages. The geometric and topological properties of moduli spaces of multipolygonal linkages in the plane are studied. These spaces are compact algebraic varieties. Some conditions under which these spaces are smooth manifolds, cross products or disjoint unions of moduli spaces of polygonal linkages, or connected, are determined. Dimensions in smooth manifold cases and some Euler characteristics are computed. A classification of generic multiquadrilateral linkages is also made.     

A Note on Log-Convexity of q-Catalan Numbers

The q-Catalan numbers studied by Carlitz and Riordan are polynomials in q with nonnegative coefficients. They evaluate, at q = 1, to the Catalan numbers: 1, 1, 2, 5, 14,…, a log-convex sequence. We use a combinatorial interpretation of these polynomials to prove a q-log-convexity result. The sequence of q-Catalan numbers is not q-log-convex in the narrow sense used by other authors, so our work suggests a more flexible definition of q-log convex be adopted. Received January 2, 2007     

Coalgebraic representations of distributive lattices with operators

We present a framework for extending Stone's representation theorem for distributive lattices to representation theorems for distributive lattices with operators. We proceed by introducing the definition of algebraic theory of operators over distributive lattices. Each such theory induces a functor on the category of distributive lattices such that its algebras are exactly the distributive lattices with operators in the original theory. We characterize the topological counterpart of these algebras in terms of suitable coalgebras on spectral spaces. We work out some of these coalgebraic representations, including a new representation theorem for distributive lattices with monotone operators.     

Nonminimal sets,their projections and integral representations of stable processes

New criteria are provided for determining whether an integral representation of a stable process is minimal. These criteria are based on various nonminimal sets and their projections, and have several advantages over and shed light on already available criteria. In particular, they naturally lead from a nonminimal representation to the one which is minimal. Several known examples are considered to illustrate the main results. The general approach is also adapted to show that the so-called mixed moving averages have a minimal integral representation of the mixed moving average type.