Scribd
Upload
44 views

Sampling Distributions and Confidence Intervals: by Simulation and by Bootstrap

This document discusses sampling distributions and confidence intervals using simulation and the bootstrap method. It provides the names of 4 students in a normal lab group, including Maria Fernanda Herreño, Viviana Guayara, and Nicol Vergara. The document also contains an answer sheet with parts asking about sampling distributions, confidence intervals, and evaluating estimates using the bootstrap.

Uploaded by

Viviana Robayo
Copyright
© © All Rights Reserved
Available Formats
Download as DOCX, PDF, TXT or read online on Scribd
44 views

Sampling Distributions and Confidence Intervals: by Simulation and by Bootstrap

This document discusses sampling distributions and confidence intervals using simulation and the bootstrap method. It provides the names of 4 students in a normal lab group, including Maria Fernanda Herreño, Viviana Guayara, and Nicol Vergara. The document also contains an answer sheet with parts asking about sampling distributions, confidence intervals, and evaluating estimates using the bootstrap.

Uploaded by

Viviana Robayo
Copyright
© © All Rights Reserved
Available Formats
Download as DOCX, PDF, TXT or read online on Scribd
You are on page 1/ 3

Sampling distributions and confidence intervals:

By simulation and by bootstrap.

Module 3, Week 6, 2021-10


Bioestadística BIOL 2205, Ciencias Biológicas, Universidad de los Andes
Professor Andrew J. Crawford

Names of the members of usual normal Lab group: 4


1. Maria Fernanda Herreño
2. Viviana Guayara
3. Nicol Vergara

ANSWER SHEET

Part 1

Paste below a small copy of your original sample from the Gamma(1, 10) distribution:

Paste below a small copy of one of your sampling distributions of the possible means from a
draw of size n (labeled, so we know which of the two this figure represents).

Results of Shapiro-Wilk tests of normality applied to sampling distributions estimated from


5,000 replicate samples of size n:

Test statistic, W P-value


Sample size n = 500
Sample size n = 20

Does the SE of the mean of one sample provide a ‘good’ estimate of the SD of the sampling
distribution of means, as our textbook claims?

se of sample sd of sampling distribution


Sample size n = 500
Sample size n = 20

Which gave us the larger sampling error (uncertainty in our estimated mean), n = 500 or n = 20?
Advanced question:
How might you next evaluate whether our estimate of the sd the sampling distribution show
signs of bias or not?

Part 2A

For the variable Tacc (or ‘jumps’) based on n = 9 green frogs, report the following:

Mean:

SD:

SE:

90% Confidence Interval based on the t distribution:

Part 2B

Paste a histogram for the sampling distribution based on 1,000 bootstrap samples.

Paste a histogram for the sampling distribution based on 10,000 bootstrap samples.

Which one looks more ‘normal’ and why? Which one will we use to calculate the bootstrap
confidence interval?

Part 3

For your bootstrap sampling distribution, report the following:

Mean:

SD:

Compare the above values with your single-sample estimates of the:


Mean:

SE:

90% bootstrap confidence interval via the formula (CI90th):

90% bootstrap confidence interval via 5th & 95th quantiles (CI90quantiles):

90% confidence interval based on t distribution (from Part 2A, above):

Question:
Which of the above three methods for obtaining a Confidence Interval seems the most
conservative and why?

Question:
Which of the above three methods for obtaining a Confidence Interval do you prefer and why?

True / False question:


The most important assumption in this exercise is that the original data were sampled randomly
with respect to (and therefore representative of) the reference population?

Answer T or F, and provide a one sentence justification.

Question:
Write a concise and correct definition of the (1 – a)% Confidence Interval:

You might also like