How to test the significance of, and report, a bootstrap of a GLM? So I'm looking at the number of lion-livestock predation incidents in one area. I created a loop which extracts 390 lion attack days and 390 random non-attack days from 2 data sets to see how climatic variables affect attacks. Obviously each time I run a binomial GLM I get different results based on the different non-attack days. So, I  created a bootstrap to run a binomial GLM 1000 times and extract the associated T-values and coefficient distributions for my independent variables. 
I can plot these to look at their significance. But is there any simple way to test and report the significance of my 1000 GLMs, and all the 1000 t values and coefficient distributions for my variables?
Cheers!
 A: Your key question is "is there any simple way to test and report the significance of my 1000 GLMs, and all the 1000 t values and coefficient distributions for my variables?"
What you want to report is the distribution of the estimated parameters and confidence intervals.
Example:
I went ahead and generated some data, and a simple t-test is performed. Say, I'm REALLY interested in seeing what the t-statistics look like. 
Given that there is a significant difference between the sample sizes, and you want to comparable sample sizes, you can sample n (400 in my example) from each group with replacement, 
as long as you boost the number of simulations. Anything above 10^4 is fine, but I use 10^5.
s1 <- rnorm(n = 400, mean = 3, sd = 2) ## I'll assume there's a higher rate here
s2 <- rnorm(n = 1000, mean = 2, sd = 2)


t_stats <- numeric(length = 10^5)

for(i in 1:10^5){
  boot_s1 <- s1[sample(x = 1:400, size = 400, replace = TRUE)]
  boot_s2 <- s2[sample(x = 1:1000, size = 400, replace = TRUE)]
  temp <- t.test(x = boot_s1, y = boot_s2, var.equal = TRUE)
  t_stats[i] <- temp$statistic
}

After producing my bootstrapped samples, I can perform a statistical test.
qt(p = 0.95, df = 399, lower.tail = TRUE)
## null hypothesis: 95% of bootstrapped t-statistics are not beyond 1.648682

summary(t_stats)

From a summary, we see that 100% of above 1.65, so you can iterate that in your report or whatevs.
Next, visualize the distribution
hist(x = t_stats, breaks = "scott", 
     probability = TRUE, main = "Bootstrapped t-Statistics", 
     xlim = c(-1, 15), xlab = "t-Statistics")
abline(v = 1.648682, lty = 2, col = 2)

And include the confidence intervals of your statistics. I'm lazy, so I use the quantiles.
## boot CI
quantile(x = t_stats, probs = c(0.025, 0.975))

That's pretty much it. The rest of the job is for you to provide some context as to what the analysis is showing with respect to your data.
Here's a reference for further reading/examples: http://www.statoo.com/en/publications/bootstrap_scgn_v131.pdf
Hope this helps
