Computer Intensive Methods
Three Computer Intensive Methods
Simulation- testing a data set against a
parameter, no null distribution
Randomization
Bootstrapping- not for hypothesis test,
creating CI for statistics that dont have the
sampling distributi
Multiple Explanatory Variables
General Linear Models
A model is a mathematical representation
of the relationship between a response
variable and one or more explanatory
variables
GLMs are designed to represent Y with a
linear model plus random error
GL
Regression
Model building
Regression is a method that predicts the
value of one numerical variable (response)
from that of another (explanatory)
Linear Regression
Draws a straight line through a scatter plot
to predict the response variable (Y or
vertica
Name:
Biostatistics Fall 2013
Assignment 5 Fitting Probability Models and Contingency Analysis
Due Oct. 9 *BRING TWO COPIES TO RECITATION
Please answer the following questions from Whitlock and Schluter in the space below and show your
work where appropri
Chapter 12 Question 16 (Use R for this question and hand in your script)
Welch test is the best, since the sample sizes are approximately equal and they are less than 30.
Also since there is a more than 3-fold difference in variance, which violates the as
Name: Brinda Kamalia
Biostatistics Fall 2013
Assignment 2 Describing and Displaying Data in R
Due Sept 18, 2013
Please answer the following questions from Whitlock and Schulter in the space below. The
assignment should be printed and handed in at the begi
Name:
Biostatistics Fall 2013
Assignment 10 Multiple Explanatory Variables
Due - December 4, 2013
Type or write your answers to the following questions in the space below
Chapter 18 Question 11. Answer sections a-c in the text. Additionally, write a well
Name:
Biostatistics Fall 2013
Assignment 9 Correlation and Regression
Due November 20, 2013
Please answer the following questions from your textbook.
Chapter 16 Questions 12 (Answer this question by hand and show your work)
Chapter 16 Question 13 (Answer