Bootstrap tests: how many bootstraps?

By Russell Davidson, James G. MacKinnon. 2000

480 citations recorded by the source corpus.

Subfields: Artificial Intelligence, Economics and Econometrics, Management Science and Operations Research

View Paper: https://www.semanticscholar.org/paper/d41e0853b7b962dca28d8cbddd71c0348c59dee2

How researchers see this paper

The researchers whose trusted networks rank it highest, strongest advocate first. Every researcher gets a different order.

In Artificial Intelligence

In Economics and Econometrics

In Management Science and Operations Research

Citations in the corpus

Bootstrap Inference in Econometrics (2002, unknown): “In certain extreme cases, in which the ordinary bootstrap J test still overrejects noticeably, a more sophisticated bootstrap test proposed by Davidson and MacKinnon (2002b) greatly reduces the remaining overrejection.”
Bootstrap Methods in Econometrics (2006, unknown): “Other references on Monte Carlo tests include Dwass (1957), in which they were first proposed, Jöckel (1986) and Davidson and MacKinnon (2000), both of which discuss power loss, Dufour and Khalaf (2001), which is a valuable survey, and Dufour et al. (2004), which discusses Monte Carlo tests for…”
Cluster-Robust Bootstrap Inference in Quantile Regression Models (2014, methodology): “The value ofY ∗ given in the first step of the algorithm is similar to the one suggested by Belloni et al. (2011, algorithm A.4) (ii) Andrews and Buchinsky (2000) and Davidson and MacKinnon (2000) provided methods for determining an appropriate number of bootstrap simulationsm.”
Wild Bootstrap Tests for IV Regression (2010, background): “See Davidson and MacKinnon (2000) for more on how to choose B appropriately.”
Cluster-robust inference: A guide to empirical practice (2022, methodology): “Ideally, it should be reasonably large (Davidson and MacKinnon, 2000) and satisfy the condition that α(B + 1) is an integer for any α (the level of the test) that may be of interest (Racine and MacKinnon, 2007).”
Promises, Threats and Fairness (2004, unknown): “According to Davidson and MacKinnon (2000), this number of replications is high enough to guarantee a reasonable confidence in the estimated p–values, compared to the " ideal " bootstrap with infinitely many replications.”
Entrepreneurial Orientation and Growth of SMEs: A Causal Model (2008, unknown): “In line with the recommendations of Davidson and MacKinnon (2000) , we have applied a total of 1,000 resamples.”
Bootstrap Determination of the Co-Integration Rank in Heteroskedastic VAR Models (2014, methodology): “Remark 4 In practice, the cdf G r;T ( ) required in Step (iv) of Algorithm 1 will be unknown, but can be approximated in the usual way through numerical simulation; see, inter alia, Hansen (1996), Davidson and MacKinnon (2000) and Andrews and Buchinsky (2000).”
Faster permutation inference in brain imaging (2016, methodology): “Some of the methods can also be used with spatial test statistics, and for inferences corrected for the familywise error rate (FWER) using the distribution of the extre-mum statistic (see below).”
Reliable Inference for the Gini Index (2009, background): “For such a test, it is also desirable to choose B such that α(B + 1) is an integer; see, among other references, Davidson and MacKinnon (2000).”
How Many Bootstrap Replicates Are Necessary? (2009, unknown): “In the area of econometrics, Davidson and MacKinnon (2000) propose a two-step procedure to determine B for BS P-values based on the most powerful test.”
The Aversion to Lying (2009, methodology): “The mean payoff to the pair is compared between treatments using bootstrap techniques, as this variable was not normally distributed.12 The significance levels for all bootstrap tests have been obtained by generating 2,099 bootstrap replications (Davidson and MacKinnon 2000).”
The power of bootstrap and asymptotic tests (2006, unknown): “The eect of the estimation error can be made as small as desired by appropriate choice of B; see Davidson and MacKinnon (2000) .”
Thirty Years of Heteroskedasticity-Robust Inference (2013, methodology): “In practice, it would be better to use a larger number for B in order to obtain better power, but 399 is adequate in the context of a simulation experiment; see Davidson and MacKinnon (2000) .”