1
BUSINESS MANAGEMENT 2320
DECISION SCIENCES: STATISTICAL TECHNIQUES
Autumn 2013
INSTRUCTOR
Mrs. Bonnie Schroeder
OFFICE Fisher Hall 330
TBA
OFFICE HOURS Office Hours are run according to a FIFS system. You cannot schedule time during
office hours for an

Business Management 2320 Case #1
Instructions:
1. An overview of instructions that apply to all cases is posted on Carmen as General Case Requirements
and StatTools Introduction. If you have not yet read that document, do so before starting Case #1.
2. A

Business Management 2320 Case #2
Instructions:
1. An overview of instructions that apply to all cases is posted on Carmen as General Case Requirements
and StatTools Introduction. If you have not yet read that document, do so before starting Case #1.
2. A

First 5 letters of your last name
OSU dot Number
Student Name: _
last
first
MI
Instructor Name: _
Recitation Day and Start Time:
_
Instructions:
You must use pencil to write your exam.
Turn off all cell phones and pagers. You may not use cell phones or ot

7
Tentative Course Schedule Autumn 2013
Week
Dates
L
Th, 8/22
R
1
M, 8/26 and T, 8/27
2
Topic
Course Introduction
General Review and Data Modeling
Practice: Normal Distribution
Practice: Sampling Distributions ( X , p )
Introduce Case #1
Stat1430 Quiz
Rea

Business Management 2320 Case #4
Instructions:
1. Case #4 reports and peer evaluations are due at the start of your scheduled recitation on Monday, 12/2
or Tuesday, 10/3. There are no alternate submission opportunities for this case. Plan accordingly.
2.

Business Management 2320 Case #3
Instructions:
1. Case #3 reports and peer evaluations are due at the start of lecture on Thursday, 10/31. Happy
Halloween!
2. An Excel file identifying your group assignment for Case #3 will be posted on Carmen by your ins

First 5 letters of your last name
K
E
Y
OSU dot Number
Student Name: _
last
first
MI
Instructor Name: _
Recitation Day and Start Time:
_
Instructions:
You must use pencil to write your exam.
Turn off all cell phones and pagers. You may not use cell phones

BM2320 Videos - SLR Inference [Slope]
10/25/2012
Point Estimator of 1
Population Model: yi 0 1 xi i
Bivariate Regression
Part II
Inference About 1
Y / X 0 1 X
Sample Equation: yi b0 b1 xi
Sample Slope, b1: b1
cov( x, y )
2
sx
b1 r
Sample Data Sampling Er

BM2320 Videos - SLR Part II - CI and PI
10/25/2012
Sample Data Sampling Error
yi 51133 561.1xi
x 30 yi 34,300
Bivariate Regression
Part II
(1 )% CI for Y/x
(1 )% PI for y
yi 47238 517.6 xi
x 30 yi 31,710
(1 )% CI Estimate of Y/x
Point estimate m =
y t *

An insurance company wishes to examine the relationship between annual income and the
amount of life insurance held by heads of families. The company takes a simple random sample
of ten family heads and obtains the following results (in $1000s).
Family
Am

BM2320 Videos - SLR Part I Scatter Plots
10/22/2012
Scatterplot (Deterministic vs. Stochastic)
m
Bivariate Regression
Part I
y = mx + b
y = 2
b=1
Summarizing Relationship
Scatter Plots
y 2
2
x 1
x = 1
y = (mx + b) + e
y mx b
Scatterplot (Positive Slope v

Time Series
R esponse
M o.,Qtr.,
Y r.
Components
What are Components?
Specific types of patterns or movements that
might be observed in time series data:
Trend
Cycle
Season
Random
Time Series Plot
Is a scatter plot of Yt against t (time
period)
F
F
F
succ

Time Series
Forecasts for Stationary
Data
Nave Forecasting Methods
Stationary time series - ones that exhibit
very gradual or no Trend and no Seasonal
variation.
Methods:
Lag 1 earlier video
Simple Moving Average (MA) earlier
video
Weighted Moving Average

Time Series
Forecasts for Stationary
Data
Nave Forecasting Methods
Appropriate for time series data that exhibit
very gradual or no Trend and no Seasonal
variation.
Examples:
Lags
Simple Moving Average (MA)
Weighted Moving Average
Single (Simple) Exponent

Time Series Methods
Moving Average?
Exponential Smoothing?
Accuracy
Measures
Other?
Definition of Forecast Error
For a given time period, t the forecast error
is the difference between the actual value of
the time series and the value forecasted for
that

BM2320 Video Lectures
Astimegoesby.
t
Time Series Module
Definition
Time Series
a set of observations on a variable
measured at successive points in time or
successive periods of time at regular
intervals.
Intent
1.
Use clues provided by the past values

Time Series Data?
Cross-sectional
Data?
Cross-sectional Data
- data collected at the same, or
approximately the same, point in time, often
for multiple variables.
Snapshot Passage of
time NOT Relevant!
Example:
Print_Q Rough Strength Supplier
68.7
67.2
66

BM2320 Videos - Multiple Regression
Extension of SLR
11/7/2012
Regression Analysis
Multiple Regression
Extension of OLS Simple Regression
The Basics
A technique used to build an equation that can be used
to estimate or predict the value of one variable
th

BM2320 Videos - Multiple Regression Ext of
SLR Example
11/7/2012
Regression Steps
Estimate the Model
Multiple Regression
Extension of OLS Simple Regression
The Basics - Example
Estimate the Model
Choose relevant variables and collect data.
Summarize da

BM2320 Videos - SLR Part II Residual
Analysis
10/25/2012
Required Conditions,
Bivariate Regression
Part II
Residual Analysis
Verifying Required Data Conditions
Normality
The xi are considered to be constants in the model,
as are o and 1. Then, yi is a s

BM2320 Videos - SLR Part II - F-Test
10/25/2012
Partitioning the Sums of Squares
Bivariate Regression
Part II
SSR = ( yi y )2 66.06
Explained Variation in Y
SSE = ( yi yi )2 61.94
Testing Significance
F-Test
Unexplained Variation in Y
SST = ( yi yi )2 128

BM2320 Videos - SLR Part I OLS
10/23/2012
Example
X
Y
1
6
2
1
3
9
4
5
5
15
6
12
3.5
8
3.5
25.6
1.871 5.060
Bivariate Regression
Part I
Fitting the Regression Line
OLS - Ordinary Least Squares
avg
var
stdev
Cov(x,y)
Line of Best Fit
?
?
14
12
10
Y
8
6
4
2

Two Cases
Inference about the Proportion
Comparing Two Populations
Hypothesis Tests
Independent Samples
Case 1: Assume p1 = p2
H0: p1 p2 = 0. This implies that p1 = p2 = p.
Combine the two sample proportions to
calculate a pooled proportion estimate of th

Hypothesis Tests
One-sample Z
The Confidence Interval Method
Symmetric (1 )% CI Review
m = ( Z * )(stdev of statistic) (LCL, UCL) = statistic m
1
/2
0
2
4
-Z*
6
/2
8
Z *
mX
p*
Zm p
*
Zm
Parameter
Z*
10
12
Z
X
p
Statistic
Critical Z-score:
Z* = Z/2
Interp

BM2320 Video Lectures
H0
:
Ha
:
, p
Z*
Zob
s
Hypothesis Testing Module
(One-Sample Z)
Reject H0
Do Not Reject H0
Inference
Test
population
parameter
()
Sample
statistic
(X)
Population
Sample
Inference
Learning Objectives
Know how to formulate hypotheses a

Hypothesis Tests
The P-value Method
P-value = Probability
It measures with probability how well the
observed sample result and an initial assumption
about a population agree.
How likely was the sample result if H0 is true?
P-value = the probability that w

Hypothesis Tests
The Critical Value Method
Zobs, Z*
Role of the Significance Level
SupposeH0: =
50
SupposeHa:
we choose = 0.05.
50
We would Reject H0 only if P-value < 0.05.
That is, we would Reject H0 only with the
5% most extreme values of the test
sta