Population 2 = Bay Health
o
Hypotheses
▪
H0: σ1 = σ2
▪
H1: σ1 ≠ σ2 (population variances are different)
●
Step 3: set a value for the significance level, ∞
o
Set ∞ = 0.05
●
Step 4: Calculate the Ftest statistic
o
Finding F
▪
F = s1 / s2
▪
= 0.8464 / 0.5041 = 1.679
●
Step 5: Determine the critical value (Fscore)
o
Degrees of Freedom
▪
D1 = (n1 – 1) = (20 – 1) = 19
▪
D2 = (n2 – 1) = (20 – 1) = 19
o
FScore
▪
Twotail test → ∞/2 = 0.025
▪
Area in right tail of distribution (0.025 column)
▪
D1 = 19 and D2 = 19
▪
→ 2.526
●
Step 6: Compare the test statistic (F) with the critical value (Fscore)
o
1.679 (F) < 2.526 (Fscore) → do not reject H0
●
Step 7: State your conclusions
o
Fail to reject null hypothesis
▪
Do not have evidence that population variances are diff for the two
hospitals
▪
CANT SAY variances are equal, but business says we probably
have no reason to investigate the diff in stay time for each of the
hospitals
●
Using PHStat2 to Compare Two Population Variances
o
Procedures
▪
Addins > TwoSample Tests (Summarized Data) > F Test for
Differences in Two Variances
▪
Fill in values where needed

Sample SD

Level of significance, ∞

Population 1 sample size and sample variance

Population 2 sample size and sample variance

Twotail test or uppertail test
o
Results
▪
Intermediate Calculations

F Test Statistic

Population 1 Sample DF

Population 2 Sample DF
▪
TwoTail Test

Upper Critical Value (FScore)

Pvalue

Whether or not to reject the null hypothesis
Chapter 14: Correlation and Simple Regression Analyses
●
Intro
o
Correlation & Investments
▪
Investors strive to achieve uncorrelated investments to reduce the
risk of one of their stocks affecting the other OR all increasing /
decreasing at one time
o
Techniques that find relationships b/w 2 factors
▪
Correlation Analysis

Determines strength and direction of the relationship b/w
two variables

Hypothesis test – det. if strength b/w 2 variables is strong
enough to be useful
▪
Simple regression

Describes the relationship b/w 2 variables using a linear
equation

Predict variable 1 given specific value of variable 2, and
vice versa

Hypothesis to det. if results are accurate enough to be
useful
o
Applications in Business
▪
Realtors want to est. relationship b/w living space and eventual
selling price in particular town
▪
Best Buy manager wants to know effect of dropping printer price
$10 will have on demand in next week
▪
Coke wants to predict extent to which running a 30second Super
Bowl ad will improve sales
➢
14.1 – DEPENDENT AND INDEPENDENT VARIABLES
o
Terms
▪
Independent Variable
(X) – explains the variation (change) in
another variable
▪
Dependent variable
(Y)
▪
X → Y
▪
Does NOT work in reverse
➢
14.2 – CORRELATION ANALYSIS
o
The Basics
▪
Measure strength / direction of linear relationship b/w 2 variables
▪
Calculate correlation coefficient (R)

Provides value describing relationship
▪
Hypothesis test to decide if relationship b/w 2 variables is strong
enough to be considered statistically significant
You've reached the end of your free preview.
Want to read all 133 pages?
 Fall '12
 Donnelly
 Normal Distribution, Null hypothesis, Hypothesis testing, Statistical hypothesis testing