90,261 research outputs found
Improving the reliability of bootstrap tests with the fast double bootstrap
No full textTwo procedures are proposed for estimating the rejection probabilities of bootstrap tests in Monte Carlo experiments without actually computing a bootstrap test for each replication. These procedures are only about twice as expensive (per replication) as estimating rejection probabilities for asymptotic tests. Then a new procedure is proposed for computing bootstrap P values that will often be more accurate than ordinary ones. This “fast double bootstrap” is closely related to the double bootstrap, but it is far less computationally demanding. Simulation results for three different cases suggest that the fast double bootstrap can be very useful in practice.Bootstrap
Improving the Reliability of Bootstrap Tests with the Fast Double Bootstrap
No full textWe first propose two procedures for estimating the rejection probabilities of bootstrap tests in Monte Carlo experiments without actually computing a bootstrap test for each replication. These procedures are only about twice as expensive (per replication) as estimating rejection probabilities for asymptotic tests. We then propose a new procedure for computing bootstrap P values that will often be more accurate than ordinary ones. This "fast double bootstrap" is closely related to the double bootstrap, but it is far less computationally demanding. Simulation results for three different cases suggest that this procedure can be very useful in practice.bootstrap test, double bootstrap, Monte Carlo experiment, rejection frequency, fast double bootstrap, FDB
Testing for Consistency using Artificial Regressions
No full textWe consider several issues related to what Hausman (1978) called "specification tests", namely tests designed to verify the consistency of parameter estimates. We first review a number of results about these tests in linear regression models, and present some new material on their distribution when the model being tested is false, and on a simple way to improve their power in certain cases. We then show how in a general nonlinear setting they may be computed as "score" tests by means of slightly modified versions of any artificial linear regression that can be used to calculate Lagrange multiplier tests, and explore some implications of this result. We show how to create a variant of the information matrix test to test for parameter consistency. We examine conventional tests and our new version in the context of binary choice models, and provide a simple way to compute both tests based on artificial regressions. Some Monte Carlo evidence suggests the most common form of information matrix test can be extremely badly behaved in samples of even quite large size.Durbin-Hausman tests; information matrix tests; binary choice models; Lagrange multiplier tests
Reply to MacKinnon and Sorfa
No full textKenneth MacKinnon
\'Curiously, Fetishism Can Be Fun\'
_Film-Philosophy_, vol. 5 no. 4, February 2001
David Sorfa
\'Hieroglyphs and Carapaces: The Enigmatic Real in Laura Mulvey\'s _Fetishism
and Curiosity_\'
_Film-Philosophy_, vol. 5 no. 5, February 200
Testing the Specification of Econometric Models in Regression and Non-Regression Directions
No full textThe asymptotic power of a statistical test depends on the model being tested, the (implicit) alternative against which the test is constructed, and the process which actually generated the data. The exact way in which it does so is examined for several classes of models and tests. First, we analyze the power of tests of nonlinear regression models in regression directions. Next, we consider the power of heteroskedasticity-robust variants of these tests. Finally, we examine the power of very general tests in the context of a very general class of models.power, specification test, heteroskedasticity-robust test
Bootstrap Hypothesis Testing
No full textThis paper surveys bootstrap and Monte Carlo methods for testing hypotheses in econometrics. Several different ways of computing bootstrap P values are discussed, including the double bootstrap and the fast double bootstrap. It is emphasized that there are many different procedures for generating bootstrap samples for regression models and other types of model. As an illustration, a simulation experiment examines the performance of several methods of bootstrapping the supF test for structural change with an unknown break point.bootstrap test, supF test, wild bootstrap, pairs bootstrap, moving block bootstrap, residual bootstrap, bootstrap P value
Artificial Regressions
No full textAssociated with every popular nonlinear estimation method is at least one "artificial" linear regression. We define an artificial regression in terms of three conditions that it must satisfy. Then we show how artificial regressions can be useful for numerical optimization, testing hypotheses, and computing parameter estimates. Several existing artificial regressions are discussed and are shown to satisfy the defining conditions, and a new artificial regression for regression models with heteroskedasticity of unknown form is introduced.artificial regression, LM test, specification test, Gauss-Newton regression, one-step estimation, OPG regression, double-length regression, binary response model
Bootstrap Inference in a Linear Equation Estimated by Instrumental Variables
No full textWe study several tests for the coefficient of the single right-hand-side endogenous variable in a linear equation estimated by instrumental variables. We show that all the test statistics--Student's t, Anderson-Rubin, Kleibergen's K, and likelihood ratio (LR)--can be written as functions of six random quantities. This leads to a number of interesting results about the properties of the tests under weak-instrument asymptotics. We then propose several new procedures for bootstrapping the three non-exact test statistics and a conditional version of the LR test. These use more efficient estimates of the parameters of the reduced-form equation than existing procedures. When the best of these new procedures is used, K and conditional LR have excellent performance under the null, and LR also performs very well. However, power considerations suggest that the conditional LR test, bootstrapped using this new procedure when the sample size is not large, is probably the method of choice.bootstrap test, weak instruments, anderson-rubin test, conditional LR test, wald test, instrumental variables
Amphithyris parva MacKinnon, Hiller, Long & Marshall 2008
No full textAmphithyris parva MacKinnon, Hiller, Long & Marshall, 2008 (Fig. 4; Table 3) Amphithyris parva Logan, 2007: p. 3108 nomen nudum (erroneously cited original description as of 2007, citation pre-dates the original publication)— MacKinnon et al. 2008: p. 329, fig. 2 A–E; table 1 (original description) Type material. Holotype: NMNZ Br 1455, Paratypes: NMNZ Br 1069 Type locality. Cook Strait, near Wellington, New Zealand. Material examined. Paratypes (10): NMNZ Br 1069 Additional material. NZGS BR 2288 Diagnosis. Dorsibiconvex shell, weak capillae on dorsal valve exterior and ventral valve exterior. Dorsal median septum absent. Description. Very small species of Amphithyris. Shell dorsibiconvex, roundly subquadrate, transparent. Dorsal valve more inflated than ventral valve and without median septum (Fig. 4 A). Large amphithyrid foramen (Fig. 4 B), short socket ridges reach beyond posterior shell edge (Fig. 4 C). Exterior surface smooth with weak capillae (Fig. 4 C). More inflated than ventral valve (Fig. 4 C and F). Ventral valve with small emergent hinge teeth not parallel to long straight hinge line (Fig. 4 D). Very weak muscle impressions visible (Fig. 4 E). Exterior surface smooth with weak capillae (Fig. 4 F). Remarks. A. parva was described by MacKinnon et al. (2008) on the basis of apparent 'topotypic' material of A. buckmani. Both Foster (1974) and Campbell & Fleming (1981) were unable to locate Thomson's holotype of A. buckmani (NMNZ Br 80) for comparison with their new species (A. hallettensis Foster, 1974 and A. richardsonae Campbell & Fleming, 1981) and instead used what were assumed to be immature specimens collected later (NMNZ Br 1069) and "five additional juveniles attached to a concretionary boulder from the Outer Island Bay Bank (NZGS BR 2288)" (Campbell & Fleming 1981: p. 147–148). After their re-discovery of the holotype of A. buckmani at NMNZ, MacKinnon and co-workers came to the conclusion that the immature specimens assigned to A. buckmani by Foster (1974) and Campbell & Fleming (1981) belong to a new species, A. parva, which according to our study differs from the other species by its dorsibiconvex shells, the absence of a dorsal median septum and the presence of capillae on the outside of both dorsal and ventral valve. The material from Outer Island Bay Bank (NZGS BR 2288) was assigned to A. parva (MacKinnon et al. 2008), but not listed as type material.Published as part of Nauendorf, Alice, Wörheide, Gert & Lüter, Carsten, 2014, Revision of the brachiopod genus Amphithyris (Rhynchonelliformea: Platidiidae) with descriptions of two new species, pp. 221-240 in Zootaxa 3847 (2) on pages 228-229, DOI: 10.11646/zootaxa.3847.2.3, http://zenodo.org/record/22723
- …
