BCa_bootstrap

Computes the bias-corrected and accelerated bootstrap estimate with confidence intervals and 2-tailed p-values

Computes the bias-corrected and accelerated bootstrap estimate for confidence intervals and 2-tailed p-values of Efron, B., & Tibshirani, R. J. (1993). An Introduction to the Bootstrap, Chapman & Hall/CRC: New York.

Usage:

[p,CI] = BCa_bootstrap(data,loo,boot,null,confidence)

[p,CI] = BCa_bootstrap(data,loo,boot,null,confidence,adjustment)

p is the two-tailed p-value and CI is the confidence interval.

data - the statistic of interest calculated on the data

loo - a vector of length N of the leave-one-out values of the statistic

of interest.

boot - the bootstrapped values of the statistic of interest

null - (optional) the value of the statistic of interest under the null

hypothesis. Defaults to 0.

confidence - (optional) the % confidence for upper and lower bounds of

the confidence interval. Defaults to 95%

adjustment - (optional) NOT RECOMMENDED. This is included for a

specialized application. The integer value entered here will

include that many observations equal to the null value into

the bootstrap values AFTER calculating the bias and

acceleration.

Jared Van Snellenberg, PhD, 2009