The Size and Power of the Bias-Corrected Bootstrap Test for Regression Models with Autocorrelated Errors

Abstract

This paper is concerned with statistical inference for the coefficient of the linear regression model when the error term follows an autoregressive (AR) model. Past studies have reported severe size distortions, when the data are trending and autocorrelation of the error term is high. In this paper, we consider a test based on the bias-corrected bootstrap, where bias-corrected parameter estimators for the AR and regression coefficients are used. For bias-correction, the jackknife and bootstrap methods are employed. Monte Carlo simulations are conducted to compare size and power properties of the bias-corrected bootstrap test. It is found that the bias-corrected bootstrap test shows substantially improved size properties and exhibits excellent power for most of cases considered. It also appears that bootstrap bias-correction leads to better size and higher power values than jackknife bias-correction. These results are found to be robust to the choice of parameter estimation methods. Copyright Springer Science + Business Media, Inc. 2005bias-correction, bootstrap, jackknife, statistical inference,

Improving the Estimation of Influenza-Related Mortality Over a Seasonal Baseline

“…This suggests that some bias is possible if the autoregressive-moving-average structure for the noise is assumed in the model inference for our sample (538 observations and 33 regression coefficients). 29 Correspondingly, (the unbiased) ordinary-least-squares inference was used for our estimates of the contribution of influenza to mortality, and bootstrapping based on the autoregressive AR(1) model for the noise 21,22 was used to estimate the confidence bounds.…”

“…This estimate is generally biased downward. 21 The bias is estimated through the following bootstrap procedure:…”

“…$M_c^{\ast }(t)$ is then regressed against the covariates, and an estimate ρ * of the autoregressive parameter is generated as the correlation between the residuals in the above linear regression. The bias is estimated as $\mathit{Bias}=\overline{{\rho }^{\ast }}-\widehat{\rho }$, 21 , where $\overline{{\rho }^{\ast }}$ is the mean of the bootstrapped sample ( ρ * ). The bootstrapped estimate of the autoregressive parameter is ρ boot = ρ̂ − Bias …”

Well-Being and Professional Efficacy Among Health Care Professionals: The Role of Resilience Through the Mediation of Ethical Vision of Patient Care and the Moderation of Managerial Support

“…This method provides non-symmetric confidence intervals, which is particularly suitable for the indirect effects. Additionally, the bias-corrected bootstrap analysis has the highest statistical power among the bootstrap methods (Kim & Yeasmin, 2005). Model fits were examined by analyzing the magnitude and statistical significance of factor loadings, the root mean square error of approximation (RMSEA), the standardized root mean square residual (SRMR), the comparative fit index (CFI) and the Tucker-Lewis index (TLI).…”

Helping others not always helps ourselves: the relationship between victim incivility and emergency workers' burnout through work-to-family conflict

“…This method offers non-symmetric confidence intervals, which is particularly recommended to estimate indirect effects (Maffoni et al, 2021). Furthermore, the bias-corrected bootstrap analysis is less biased and more powerful than other bootstrap methods (Kim and Yeasmin, 2005). To test whether the strength of the relationship between victim incivility and burnout symptoms through work-family conflict was conditional on the values of social support seeking, a moderated mediation model was run using Mplus Version 8 (Muth en and Muth en, 2017) while controlling for gender, profession, job tenure and resilience levels.…”