SignificanceTesting
The assumption that what is being tested
has no significant relationship. The null
hypothesis attempts to show that no
variationexistsbetweenvariables,orthata
singlevariableisnodifferentthanzero.
Residuals:
Standarddeviationoftheresid
Solutions to Statistics 102 Midterm II Fall 2014
1. Predicted value is 2117.47.5*200+4.875*200= 1592.4: 2RMSE=2*35.7=71.4
Prediction interval is then from 1592.471.4=1521 to 1592.4+71.4=1663.8
2. Confidence interval for increase by $1 is 7.5 2*0.226 or
Stat 102 Solutions: Homework 1
1
A
a. P (0.0 Z 1.2) = 0.3849
c. P (0.0 Z 1.45) = 0.4265
g. P (Z 2.50) = 0.0062
h. P (Z 1.66) = 0.9515
B
i) X:= monthly demand,
X N (200, 502 ), Z =
P (X > 275) = P (
X 200
N (0, 1)
50
X 200
> 1.5) = P (Z > 1.5) = 0.0668
50
Statistics 102 Homework 1: Due Monday September 15th, 2014
Notes: 1. All data sets are available in the homework folder in Canvas.
2. Unless stated otherwise, all tests should be performed at an =.05 level and confidence
intervals should be 95%.
Textbook
Stat 102: Quiz 1
Name:
Section:
Record your answers here. Answers not recorded here will not be graded. There is one
answer to each question.
1.
2.
3.
4.
1. The daily receipts of a fast food franchise are normally distributed with a mean of 2000
and a sta
STAT 102 Problem Set 5, Spring 2015
Due: 5pm Tuesday, April 14 (NOTE: This problem set is due on a TUES)
Read:
Indicators, polynomial regression:
Dielman 7.17.2, 5.15.2
Model building:
Dielman 8.11.4
Finalcial timeseries
Dielman 3.6, 4.7, 7.3
Do:
Probl
Hyunseung Kang
STAT 102 Review Guide
Mar. 19, 2011
March 19, 2011: Thanks to Eliza and Jennie, xed typo on Sm and Sp in simple linear
regression as well as the typo on on the prediction interval for dierence in only ONE x
March 20, 2011: Thanks to Tianpu,
STAT 102
Homework 5
Due Date: Tuesday, March 23 before 5:00pm in 400 JMHH.
Note: Ignore the output from the text book. You should produce JMP output and use that
in your solution to all problems.
1. Problem 4.4 (p. 158)
2. Problem 4.8 (p. 169) from the te
Solutions for Statistics 102 Midterm #1
Professors Lawrence Brown & Linda Zhao
Solutions prepared by Adam Kapelner and Emil Pitkin
Some editing by J. Bleich many years later
100 points total
Feb 23, 2010
Problem 1
a
We conduct a matched pairs test. Let Di
Stat 102 Solutions: Homework 4
1
For pedagogical purposes, the answer to this question is very long, but your answer need not be.
There are two ways to structure the equation: one where SOFT COVER is your reference, which
is what JMP does if you use the c
Statistics 102 Homework 5: Monday November 24th , 2013
1. All data sets are available in the homework folder in HW5 .
2. All tests should be performed at an =.05 level; use 95% for confidence and prediction intervals.
1. A. If a supermarket product is fr
Stat 102 Solutions: Homework 5
1
A A Distributions of sales by groups are as follows. The standard deviations across four groups
do not vary much. So the assumption of equal variability is reasonable.
B We use Tukey HSD to see if there are dierences among
Administrative Sheet for Exam II for Statistics 102
A. Coverage: The exam primarily covers multiple regression (Chapters 4,6 and
7) but of course some aspects of simple regression that relate to multiple
regression are also fair game.
B. The exam will be
Statistics 102
Spring 2013
Midterm Exam II
Read these instructions carefully.
This is a closedbook exam. You are allowed to use a calculator and one page (8.5 by
11 inches or A4, both sides) of notes. No use of cellular telephones or other portable
elect
1
Department of Statistics
The Wharton School
University of Pennsylvania
Statistics 102
Fall 2014
Midterm Exam 2
Read these instructions carefully.
This is a closedbook exam. You are allowed to use a calculator and one page (8.5 11 inches
or A4, both sid
Solutions to Practice Problems for Exam 1
February 8, 2013
1
Question 1
Part A
1. The approximate condence interval for the mean is found by
x 2 s/ n.
The approximate comes from the fact that we use 2 instead of the Tdistribution
quantile (with df equal
Administrative Sheet for Exam II for Statistics 102
I. Coverage: The exam will cover multiple regression. Concepts from simple regression might also be
necessary in that multiple regression builds on simple regression. In other words, the exam will focus
Statistics 102 Homework 6: Wednesday December 10th , 2014
1. All data sets are available in the homework folder in HW6 .
2. All tests should be performed at an =.05 level; use 95% for confidence and prediction intervals.
1. A. Dielman Chapter 10 Page 390
Statistics 102 Overview Lecture 20
Purpose: Capstone on Regression and prepare for the exam
o Review of the material (see notes below)
o Go over practice problems (short answers and multiple
choice)

Review for Exam II
I. Looking at Data
A. Distribution
Overview Lecture 1 Statistics 102
o Administrative Issues
o Overview of Course
Theme: Predict a variable of interest (Y) using available
information (X variables predictors)
Questions:
How would we predict Y for a new instance and with what
accuracy?
W
Statistics 102 Overview Lecture 16
Review Multicollinearity
Model Improvement: Residuals Residual Residuals
o Diagnostics to Uncover Improvements
Residual Plot
Outliers in the Ydirection lonely vertical
Transformation of Y U or shape
Unequal varia
Statistics 102 Overview Lecture 22
Theme: Y(#s) versus Xs (Categories).
o One Way Analysis of Variance (ANOVA) Review
Question 1: How do we estimate the parameters?
Question 2: Are Y and X related (Ftest) ?
Question 3: Are the assumptions reasonable?
Statistics 102 Overview Lecture 24
Theme: Do the counts for categorical variables agree with some
model (null distribution)
o Application I Goodness of Fit
Establish Approach
o Application II Are two categorical variables related
Apply the goodness of
Statistics 102 Overview Lecture 26
Topic: Logistic regression predicting the probability
of a dichotomous outcome
o Review: Basic questions (Begin with one X)
What is the meaning of the slope? Is X significant?
What is the meaning of the intercept?
Ho
Statistics 102 Overview Lecture 25
o Contingency Table
Applications
1. Bayes Rule: Col % Count Row %
2. Understanding through simulation
Review Example
o Logistic regression predicting the probability of
a dichotomous outcome
CONFUSERS: 1.The meaning o
Statistics 102 Overview Lecture 23
Theme: Y(#s) versus Xs (Categories).
o One Way Analysis of Variance (ANOVA) Review
Question 1: Are Y and X related (Ftest)?
Question 2: Which levels are different (Tukey; Hsu)?
Question 3: Are the assumptions reasona
Statistics 102 Overview Lecture 17
Review: Modifying the model
Residual Plot
o Extreme observations in the Ydirection
o Fanning out unequal variability
o U or shape nonlinearity
Leverage Plot
o Extreme observations in the Xdirection
o U or shape non
Statistics 102 Overview Lecture 19
o Review: Building Models with Categorical Variables
Two kinds of models:
Additive model (no interactions) parallel lines
Interaction model: The effect of the categorical
variable on the response differs by value of a
Name:_
Check one:
Mon.Wed. Section:_
Tues.Thurs. Section:_
Statistics 102
Final Exam
Dec. 21, 2000
46pm
This exam is closed book.
You may have three pages of notes.
You may use a calculator.
You may write in pen or pencil.
Show all your work.
Statistic
Name:_
Check one:
Section 1 (TuesThurs. 10:30noon):_
Section 2 (TuesThurs. 1:303:00):_
Statistics 102 Midterm #1
October 11, 2001 68pm
This exam is closed book.
You may have one page of notes. You may use a calculator.
You must write the exam using p
Statistics 102 Fall 2013
Introduction to Business Statistics
Linda Zhao: [email protected], Office: 470JMHH.
Office Hours: 3:00 5:00 pm Mondays or by appointment
Textbook: Applied Regression Analysis for Business and Economics, Terry Dielman, Fourt
STAT 102 Spring 2014
Section 003
Course Syllabus
The Wharton School of the University of Pennsylvania
Administrative Details
Class Meets: JMHH F85  Tues. and Thurs. 3:00 4:30PM
PreRequisites: STAT 101 (or AP Stat)
Course Instructor: Justin Bleich
Offic