Hi, the following codes provided me results of bootstrapped logistic regression, but boot.ci() only provide one CI, which I have no idea what it it. I wonder if there is a way to obtain the CI for each factor in the model. Any help is appreciated.

logistic_coef <-function(d, i){

d <- d[i,]

fit <- glm(bpd ~ birthwt + gestage + toxemia,

data = d,

family = "binomial")

return(coef(fit))

}

#apply the boot() function

#sytem.time estimate time used

set.seed(77)

system.time(boot_logistic <- boot(data = bpd,

statistic = logistic_coef,

R=5000))

#view results

boot_logistic

boot.ci(boot_logistic, type = "perc")

Call:

boot(data = bpd, statistic = logistic_coef, R = 5000)

Bootstrap Statistics :

original bias std. error

t1* 13.936082577 0.4182376970 2.7596107013

t2* -0.002643578 -0.0001080289 0.0009459613

t3* -0.388535681 -0.0107899365 0.1042887437

t4* -1.343786469 -0.1162307066 1.0181248939

BOOTSTRAP CONFIDENCE INTERVAL CALCULATIONS

Based on 5000 bootstrap replicates

CALL :

boot.ci(boot.out = boot_logistic, type = "perc")

Intervals :

Level Percentile

95% ( 9.32, 20.01 )

Calculations and Intervals on Original Scale