Bootstrap Student's t-test for 2 independent samples
Bootstrap Student's t-test for 2 independent samples.
boot.student2(x, y, B = 999)
x |
A numerical vector with the data. |
y |
A numerical vector with the data. |
B |
The number of bootstrap samples to use. |
We bootstrap Student's (Gosset's) t-test statistic and not the Welch t-test statistic. For the latter case see the "boot.ttest2" function in Rfast. The difference is that Gosset's test statistic assumes equaility of the variances, which if violated leads to inlfated type I errors. Bootstrap calibration though takes care of this issue. As for the bootstrap calibration, instead of sampling B times from each sample, we sample sqrt{B} from each of them and then take all pairs. Each bootstrap sample is independent of each other, hence there is no violation of the theory (Chatzipantsiou et al., 2019).
A vector with the test statistic and the bootstrap p-value.
Michail Tsagris
R implementation and documentation: Michail Tsagris mtsagris@uoc.gr.
Efron Bradley and Robert J. Tibshirani (1993). An introduction to the bootstrap. New York: Chapman \& Hall/CRC.
Chatzipantsiou C., Dimitriadis M., Papadakis M. and Tsagris M. (2019). Extremely efficient permutation and bootstrap hypothesis tests using R. To appear in the Journal of Modern Applied Statistical Methods.
x <- rexp(40, 4) y <- rbeta(50, 2.5, 7.5) system.time(t.test(x, y, var.equal = TRUE) ) system.time( a <- boot.student2(x, y, 9999) ) a
Please choose more modern alternatives, such as Google Chrome or Mozilla Firefox.