Download Bootstrap Tests for Regression Models by L. Godfrey PDF

By L. Godfrey

An obtainable dialogue analyzing computationally-intensive strategies and bootstrap tools, delivering how you can enhance the finite-sample functionality of famous asymptotic assessments for regression types. This booklet makes use of the linear regression version as a framework for introducing simulation-based checks to aid practice econometric analyses.

Show description

Read or Download Bootstrap Tests for Regression Models PDF

Best econometrics books

The Years of High Econometrics: A Short History of the Generation that Reinvented Economics

A desirable and entire background, this ebook explores an important transformation in 20th century economics: the production of econometrics. Containing clean archival fabric that has no longer been released earlier than and taking Ragnar Frisch because the narrator, Francisco Louca discusses either the keys occasions - the institution of the Econometric Society, the Cowles fee and the magazine Econometrica - and the main avid gamers - economists like Wesley Mitchell, mathematicians like John von Neumann and statisticians like Karl Pearson - in background that formed the improvement of econometrics.

Non-standard Spatial Statistics and Spatial Econometrics

Regardless of spatial records and spatial econometrics either being contemporary sprouts of the final tree "spatial research with measurement"—some might take into account the controversy after WWII approximately "theory with out dimension" as opposed to "measurement with no theory"—several normal topics have emerged within the pertaining literature.

Quantitative Models for Performance Evaluation and Benchmarking: Data Envelopment Analysis with Spreadsheets

It's tricky to guage an organization’s functionality while there are a number of inputs and a number of outputs to the approach. The problems are extra more suitable while the relationships among the inputs and the outputs are complicated and contain unknown tradeoffs. This ebook introduces DEA as a multiple-measure functionality overview and benchmarking instrument.

Applied Nonparametric Econometrics

Nearly all of empirical examine in economics ignores the aptitude merits of nonparametric equipment, whereas nearly all of advances in nonparametric conception ignores the issues confronted in utilized econometrics. This e-book is helping bridge this hole among utilized economists and theoretical nonparametric econometricians.

Additional resources for Bootstrap Tests for Regression Models

Example text

Instead such relationships are simply artificial (auxiliary) regressions designed to allow the convenient computation of a diagnostic check. The issue of whether to use restricted or unrestricted residuals when calculating robust tests will be discussed in the next section and will be considered in Chapter 3 in the context of simulation-based tests. 5. Finite-sample properties of asymptotic tests It has been pointed out that there is rarely precise information available about the form of the error distribution and, in particular, there seems little reason to believe that the errors are Normally distributed.

7). If the errors are assumed to be independent and heteroskedastic, a Heteroskedasticity-Consistent Covariance Matrix Estimator (usually denoted by HCCME) is required. If the errors are heteroskedastic and autocorrelated, a Heteroskedasticity and Autocorrelation Consistent (usually denoted by HAC) estimator is needed. The former provides standard errors that are heteroskedasticityrobust. The latter provides standard errors that are heteroskedasticity and autocorrelation robust. Many computer programs offer users the chance to use robust standard errors from either some HCCME or some HAC estimate, rather than relying on the traditional IID-valid standard errors given by the matrix s2 (X X)−1 .

However, the traditional standard errors are often provided as the default and this approach has been criticized. Stock and Watson remark that In econometric applications, there is rarely a reason to believe that the errors are homoskedastic and normally distributed. Because sample sizes are typically large, however, inference can proceed . . by first computing the heteroskedasticity-robust standard errors. (Stock and Watson, 2007, p. 171) Similarly, it is argued in Hansen (1999) that a modern approach should involve the use of test statistics that are valid under heteroskedasticity and do not require the assumption of Normality.

Download PDF sample

Rated 4.45 of 5 – based on 27 votes