bootstrapping

Question 1

parameter (N units)

Answer 1

numerical summary for the population (population slope B1)

Question 2

statistic (n units)

Answer 2

numerical summary calculated from sample data (estimated slope in sample B^1)

Question 3

good samples

Answer 3

drawn at random, unbiased, and representative of of population
should be good estimates of population parameter without need for census

Question 4

how to get a bootstrap sample

Answer 4

choose with replacement from the existing sample, using the same sample size

Question 5

bootstrap statistic

Answer 5

statistic computed for each bootstrap sample

Question 6

bootstrap distribution

Answer 6

collection of bootstrap statistics from many bootstrap sample

Question 7

how to get bootstrap distribution

Answer 7

start with sample size n, take k resamples, calculate statistic on each as k -> infinity, distribution of k resample statistics approximates sampling distribution

Question 8

standard error

Answer 8

standard deviation of sample distribution (measure of sampling variability)

Question 9

bootstrap standard error

Answer 9

the standard deviation of the bootstrap distribution

Question 10

what does Bootstrap assess

Answer 10

if sample results are statistically significant and draw inferences from the regression model to the population
based on sampling repeatedly with replacement from data at hand and computing regression coefficients from each re-sample