Jackknife, Bootstrap and Other Resampling Methods in Regression Analysis

Wu, C. F. J.

doi:10.1214/aos/1176350142

Cited by 1,500 publications

(962 citation statements)

References 30 publications

Supporting

Mentioning

926

Contrasting

Unclassified

Order By: Relevance

“…Although asymptotically one only needs the first term of (4) sinceγ = O(n −1/2 ), the inclusion of the second term is likely to improve the approximation to the distribution of (3) in finite sample. This perturbation method is similar to the so-called wild bootstrap (Wu, 1986;Härdle, 1990;Mammen, 1992) and has been used successfully for a number of interesting problems in survival analysis (Jin et al, 2001;Park & Wei, 2003;Cai et al, 2005). The distribution of (4) can be easily approximated by generating a large number M of realizations of .…”

Section: Pointwise and Simultaneous Interval Estimation Procedures Fomentioning

confidence: 99%

Calibrating parametric subject-specific risk estimation

Cai¹,

Tian²,

Uno³

et al. 2010

Biometrika

View full text Add to dashboard Cite

SUMMARYFor modern evidence-based medicine, decisions on disease prevention or management strategies are often guided by a risk index system. For each individual, the system uses his/her baseline information to estimate the risk of experiencing a future disease-related clinical event. Such a risk scoring scheme is usually derived from an overly simplified parametric model. To validate a model-based procedure, one may perform a standard global evaluation via, for instance, a receiver operating characteristic analysis. In this article, we propose a method to calibrate the risk index system at a subject level. Specifically, we developed point and interval estimation procedures for t-year mortality rates conditional on the estimated parametric risk score. The proposals are illustrated with a dataset from a large clinical trial with post-myocardial infarction patients.

show abstract

Section: Pointwise and Simultaneous Interval Estimation Procedures Fomentioning

confidence: 99%

Calibrating parametric subject-specific risk estimation

Cai¹,

Tian²,

Uno³

et al. 2010

Biometrika

View full text Add to dashboard Cite

show abstract

“…Both axes use the log scale. Also shown are 90% bootstrapped confidence intervals obtained using the "wild" bootstrap procedure (Wu 1986) which allows for heteroskedastic errors. The forty-five degree line in the figures shows reported self-employment income.…”

Section: Resultsmentioning

confidence: 99%

Estimating the income reporting function for the self-employed

Tedds

2009

Empir Econ

View full text Add to dashboard Cite

There is considerable interest in measuring the underground economy using microeconomic data. One such method estimates income under-reporting by households by assuming a known, parametric form of the Engel curve and making the further parametric assumption that households under-report their income by a constant fraction, independent of income. This paper proposes a nonparametric approach which avoids functional form restrictions and enables the reporting function to vary across income levels and household characteristics. I illustrate by estimating the effect of the Canadian Goods and Services Tax on income under-reporting.

show abstract

“…De fato, estimativas de bootstrap obtidas como descrito acima não são nem consistentes nem assintoticamente não-viesadas quando os dados provêm de um mecanismo gerador heterocedástico; (Wu, 1986). Um estimador de bootstrap robustoà presença de heterocedasticidade foi proposto por Wu (1986) e pode ser descrito da seguinte forma: a) para cada i, i = 1, .…”

Section: O Modelo E Estimadoresunclassified

“…O objetivo deste artigoé investigar, através de simulações de Monte Carlo, o desempenho deste estimador em amostras de tamanho pequeno a moderado. Consideramos também duas alternativas ao estimador de White, a saber: a seqüência de estimadores de White corrigidos, proposta por Cribari-Neto, Ferrari e Cordeiro (2000), e o estimador de bootstrap ponderado, proposto por Wu (1986). São, ainda, apresentados resultados para três estimadores propostos por MacKinnon e White (1985), denotados HC1, HC2 e HC3.…”

Section: Introdução E Motivaçãounclassified

Uma análise de Monte Carlo do desempenho de estimadores de matrizes de covariância sob heterocedasticidade de forma desconhecida

Cribari‐Neto¹,

Gois²

2002

Rev. Bras. Econ.

View full text Add to dashboard Cite

Este artigo analisa o desempenho em amostras finitas do estimador consistente de matrizes de covariância proposto por Halbert White. O comportamento deste estimadoré estudado tanto sob homocedasticidade quanto sob heterocedasticidade usando métodos de simulação Monte Carlo. Este estimador pode apresentar viés significativo para amostras de tamanho pequeno a moderado. O desempenho em amostras finitas de estimadores de bootstrap e analiticamente corrigidos tambémé analisado. Os resultados numéricos favorecem os estimadores corrigidos analiticamente em relação ao estimador obtido a partir de um esquema de reamostragem de bootstrap. Três estimadores alternativos que são construídos como variações do estimador originalmente proposto por White são também analisados. Os resultados revelam, ainda, a influência de pontos de alta alavancagem sobre o desempenho dos diversos estimadores. This paper analyzes the finite-sample performance of the consistent covariance matrix estimator proposed by Halbert White under both homoskedasticity and heteroskedasticity using Monte Carlo simulation methods. It showns that this estimator can be quite biased in samples of small to moderate sizes, thus leading to associated quasit tests with large size distortions. The finite-sample performance of bootstrap and bias-corrected estimators is also investigated. The numerical results favor the analitically corrected estimators over the one obtained from a weighted bootstrapping scheme. The paper analyzes the finite-sample of three alternative estimators, which are defined as small variations of the White estimator. Finally, the results also show that the existence of points of high leverage in the regression matrix has a substantial impact on the finite-sample performance of the different covariance matrix estimators.*Artigo recebido em out. 2000 e aprovado em fev. 2001. Os autores agradecem o apoio financeiro parcial do CNPq, bem como os comentários, críticas e sugestões de dois pareceristas anônimos.

show abstract

Jackknife, Bootstrap and Other Resampling Methods in Regression Analysis

Cited by 1,500 publications

References 30 publications

Calibrating parametric subject-specific risk estimation

Calibrating parametric subject-specific risk estimation

Estimating the income reporting function for the self-employed

Uma análise de Monte Carlo do desempenho de estimadores de matrizes de covariância sob heterocedasticidade de forma desconhecida

Contact Info

Product

Resources

About