0). The null hypothesis is that the new method is not better than or equal to the old method (μ≤0). Hypothesis tests can have type I and type II errors. The significance level determines the probability of a type I error and the size of the rejection region. P-values indicate how significant sample evidence is in supporting the alternative hypothesis over the null hypothesis.">0). The null hypothesis is that the new method is not better than or equal to the old method (μ≤0). Hypothesis tests can have type I and type II errors. The significance level determines the probability of a type I error and the size of the rejection region. P-values indicate how significant sample evidence is in supporting the alternative hypothesis over the null hypothesis.">
STAT609 SP23 LCN Unit3
STAT609 SP23 LCN Unit3
STAT609 SP23 LCN Unit3
STATISTICAL INFERENCE
Chapter 9
Hypothesis Testing
© 2017 Cengage Learning. All Rights Reserved. May not be scanned, copied or duplicated, or posted to a publicly accessible website, in whole or in part.
9-1 Introduction
© 2017 Cengage Learning. All Rights Reserved. May not be scanned, copied or duplicated, or posted to a publicly accessible website, in whole or in part.
Example 9.1: A New Pizza Style at Pepperoni
Pizza Restaurant
© 2017 Cengage Learning. All Rights Reserved. May not be scanned, copied or duplicated, or posted to a publicly accessible website, in whole or in part.
9-2b One-Tailed versus Two-Tailed Tests
© 2017 Cengage Learning. All Rights Reserved. May not be scanned, copied or duplicated, or posted to a publicly accessible website, in whole or in part.
9-2c Types of Errors
© 2017 Cengage Learning. All Rights Reserved. May not be scanned, copied or duplicated, or posted to a publicly accessible website, in whole or in part.
9-2d Significance Level and Rejection Region
© 2017 Cengage Learning. All Rights Reserved. May not be scanned, copied or duplicated, or posted to a publicly accessible website, in whole or in part.
9-2e Significance from p-values
© 2017 Cengage Learning. All Rights Reserved. May not be scanned, copied or duplicated, or posted to a publicly accessible website, in whole or in part.
9-2g Hypothesis Tests and Confidence Intervals
© 2017 Cengage Learning. All Rights Reserved. May not be scanned, copied or duplicated, or posted to a publicly accessible website, in whole or in part.
9-2h Practical versus Statistical Significance
© 2017 Cengage Learning. All Rights Reserved. May not be scanned, copied or duplicated, or posted to a publicly accessible website, in whole or in part.
9-3 Hypothesis Tests for a Population Mean
P-Value:
▪ The p-value is also called the observed level of significance.
▪ The smallest α level at which 𝐻0 can be rejected.
▪ If p-value < α, reject 𝐻0
▪ If p-value ≥ α, do not reject 𝐻0
Ha P-value
𝜇 > 𝜇0 𝑃 𝑇>𝑡
𝜇 < 𝜇0 𝑃 𝑇<𝑡
𝜇 ≠ 𝜇0 2𝑃 𝑇 > 𝑡
© 2017 Cengage Learning. All Rights Reserved. May not be scanned, copied or duplicated, or posted to a publicly accessible website, in whole or in part.
Example 9.1: Undergraduate Study Habits slide
(1 of 2)
A recent study asserts that, over the past five decades, the number
of hours that the average college student studies each week has
been steadily dropping (The Boston Globe, July 4, 2010). In 1961,
students invested 24 hours per week in their academic pursuits,
whereas today’s students study an average of 14 hours per week.
The dean randomly selects 35 students and asks their average study
time per week (in hours). From their responses, she calculates a
sample mean of 16.3714 hours and a sample standard deviation
of 7.2155 hours. The dean would also like to test if the mean study
time of students at her university differs from today’s national
average of 14 hours per week. At the 5% significance level, what is
the conclusion to this test?
© 2017 Cengage Learning. All Rights Reserved. May not be scanned, copied or duplicated, or posted to a publicly accessible website, in whole or in part.
Example 9.1: Undergraduate Study Habits slide
(2 of 2)
© 2017 Cengage Learning. All Rights Reserved. May not be scanned, copied or duplicated, or posted to a publicly accessible website, in whole or in part.
Example 9.2: A New Pizza Style at Pepperoni
Pizza Restaurant (slide 1 of 5)
© 2017 Cengage Learning. All Rights Reserved. May not be scanned, copied or duplicated, or posted to a publicly accessible website, in whole or in part.
Example 9.2: A New Pizza Style at Pepperoni Pizza
Restaurant (slide 3 of 5)
SPSS Steps:
From the menus choose:
• Analyze Compare Means One-Sample T Test..
• Click Rating and move it onto the Test Variable(s) field.
• Click in the Test Value box and enter the value that you will compare to. In
© 2017 Cengage Learning. All Rights Reserved. May not be scanned, copied or duplicated, or posted to a publicly accessible website, in whole or in part.
Example 9.2: A New Pizza Style at Pepperoni
Pizza Restaurant (slide 4 of 5)
© 2017 Cengage Learning. All Rights Reserved. May not be scanned, copied or duplicated, or posted to a publicly accessible website, in whole or in part.
Example 9.2: A New Pizza Style at Pepperoni
Pizza Restaurant (slide 5 of 5)
© 2017 Cengage Learning. All Rights Reserved. May not be scanned, copied or duplicated, or posted to a publicly accessible website, in whole or in part.
9-4 Hypothesis Tests for Other Parameters
© 2017 Cengage Learning. All Rights Reserved. May not be scanned, copied or duplicated, or posted to a publicly accessible website, in whole or in part.
9-4a Hypothesis Tests for a Population Proportion
© 2017 Cengage Learning. All Rights Reserved. May not be scanned, copied or duplicated, or posted to a publicly accessible website, in whole or in part.
9-4a Hypothesis Tests for a Population Proportion
P-Value:
▪ If p-value < α, reject 𝐻0
▪ If p-value ≥ α, do not reject 𝐻0
Ha P-value
𝑝 > 𝑝0 𝑃 𝑍>𝑧
𝑝 < 𝑝0 𝑃 𝑍<𝑧
𝑝 ≠ 𝑝0 2𝑃 𝑍 > 𝑧
© 2017 Cengage Learning. All Rights Reserved. May not be scanned, copied or duplicated, or posted to a publicly accessible website, in whole or in part.
Example 9.3: Customer Complaints at Walpole
Appliance Company (slide 1 of 2)
© 2017 Cengage Learning. All Rights Reserved. May not be scanned, copied or duplicated, or posted to a publicly accessible website, in whole or in part.
9-4b Hypothesis Tests for Differences between
Population Means
© 2017 Cengage Learning. All Rights Reserved. May not be scanned, copied or duplicated, or posted to a publicly accessible website, in whole or in part.
9-4b Hypothesis Tests for Differences between
Population Means
© 2017 Cengage Learning. All Rights Reserved. May not be scanned, copied or duplicated, or posted to a publicly accessible website, in whole or in part.
Example 9.4: Measuring the Effects of Soft-Drink
Cans (slide 1 of 4)
© 2017 Cengage Learning. All Rights Reserved. May not be scanned, copied or duplicated, or posted to a publicly accessible website, in whole or in part.
Example 9.4: Measuring the Effects of Soft-Drink
Cans (slide 2 of 4)
Let
1: Attractiveness of the traditional-style can (AO)
2: Attractiveness of the new-style can (AN)
Test the rate the attractiveness of the new design is higher than
the attractiveness of the current design.
Let D = the mean difference in the two style can (D = 1- 2)
Hypotheses: H0: D = 0, Ha: D < 0.
Use 5% significance level (𝛼 = 0.05).
SPSS Steps:
From the menus choose:
Analyze Compare Means Paired-Samples T Test
Select AO as Variable 1 and AN as Variable 2
© 2017 Cengage Learning. All Rights Reserved. May not be scanned, copied or duplicated, or posted to a publicly accessible website, in whole or in part.
Example 9.4: Measuring the Effects of Soft-Drink
Cans (slide 3 of 4)
ഥ 0
𝐷−𝐷 −0.539−0
𝑡= = = −5.351,
𝑠𝐷 Τ 𝑛 1.351Τ 180
P-value < 0.001
P-value=0.00000013232, so reject H0.
We conclude that there is overwhelming
evidence that consumers, on average,
rate the attractiveness of the new
design higher than the attractiveness of
the current design.
© 2017 Cengage Learning. All Rights Reserved. May not be scanned, copied or duplicated, or posted to a publicly accessible website, in whole or in part.
Example 9.4: Measuring the Effects of Soft-Drink
Cans (slide 4 of 4)
SPSS Output:
A 95% confidence interval for the mean difference extends from -0.738
to -0.340. Note that this 95% confidence interval does not include the
hypothesized value 0, so we reject H0. This is consistent with the fact that
the two-tailed p-value is less than 0.05. (Recall the relationship between
confidence intervals and two-tailed hypothesis tests from Section 9-2g.)
© 2017 Cengage Learning. All Rights Reserved. May not be scanned, copied or duplicated, or posted to a publicly accessible website, in whole or in part.
9-4b Hypothesis Tests for Differences between
Population Means
If the samples are independent, the test is referred to as the t test for
difference between means from independent samples.
One- and Two-Sided Tests
𝐻0 : 𝜇1 − 𝜇2 = 𝐷0 𝜇1 − 𝜇2 ≤ 𝐷0 𝑣𝑠. 𝐻𝑎 : 𝜇1 − 𝜇2 > 𝐷0
𝐻0 : 𝜇1 − 𝜇2 = 𝐷0 𝜇1 − 𝜇2 ≥ 𝐷0 𝑣𝑠. 𝐻𝑎 : 𝜇1 − 𝜇2 < 𝐷0
𝐻0 : 𝜇1 − 𝜇2 = 𝐷0 𝑣𝑠. 𝐻𝑎 : 𝜇1 − 𝜇2 ≠ 𝐷0
❑ Test statistic for independent samples test of difference between means when
equal variances is assumed,= 12 = 22:
❑ The t-value follows a t distribution with degrees of freedom (df) equal to (n1 +
n2 – 2) where the pooled standard deviation sp is given by:
© 2017 Cengage Learning. All Rights Reserved. May not be scanned, copied or duplicated, or posted to a publicly accessible website, in whole or in part.
9-4b Hypothesis Tests for Differences between
Population Means
P-value
Ha P-value
𝜇1 − 𝜇2 > 𝐷0 𝑃 𝑇>𝑡
𝜇1 − 𝜇2 < 𝐷0 𝑃 𝑇<𝑡
𝜇1 − 𝜇2 ≠ 𝐷0 2𝑃 𝑇 > 𝑡
© 2017 Cengage Learning. All Rights Reserved. May not be scanned, copied or duplicated, or posted to a publicly accessible website, in whole or in part.
Example 9.5: Designing a promotional Web
page to increase online sales (slide 1 of 4)
© 2017 Cengage Learning. All Rights Reserved. May not be scanned, copied or duplicated, or posted to a publicly accessible website, in whole or in part.
Example 9.5: Designing a promotional Web
page to increase online sales (slide 2 of 4)
𝑃 − 𝑣𝑎𝑙𝑢𝑒 = 𝑃 𝑇 > 4.09 = 0.000056 (by SPSS). The p-value is much less
than 0.05, then reject 𝐻0 .
Conclusion:
We conclude that the new page generates statistically significantly higher sales
than the old page.
© 2017 Cengage Learning. All Rights Reserved. May not be scanned, copied or duplicated, or posted to a publicly accessible website, in whole or in part.
Example 9.5: Designing a promotional Web
page to increase online sales (slide 3 of 4)
SPSS Steps:
From the menus choose:
Analyze Compare Means Summary Independent- Samples
T Test.
Complete the dialog box as shown.
© 2017 Cengage Learning. All Rights Reserved. May not be scanned, copied or duplicated, or posted to a publicly accessible website, in whole or in part.
Example 9.5: Designing a promotional Web
page to increase online sales (slide 4 of 4)
SPSS Output:
© 2017 Cengage Learning. All Rights Reserved. May not be scanned, copied or duplicated, or posted to a publicly accessible website, in whole or in part.
Example 9.6: Productivity Due to Exercise at
Informatrix (slide 1 of 3)
© 2017 Cengage Learning. All Rights Reserved. May not be scanned, copied or duplicated, or posted to a publicly accessible website, in whole or in part.
Example 9.6: Productivity Due to Exercise at
Informatrix (slide 2 of 3)
SPSS Steps:
From the menus choose:
Analyze Compare Means Independent- Samples T Test.
Select Rating and move it onto the Test Variable(s) field.
Select Exerciser and move it onto the Grouping Variable field.
Click on the <Define Groups> button. In the window displayed,
enter Yes and No for Groups 1 and 2 respectively.
SPSS Output:
© 2017 Cengage Learning. All Rights Reserved. May not be scanned, copied or duplicated, or posted to a publicly accessible website, in whole or in part.
Example 9.6: Productivity Due to Exercise at
Informatrix (slide 3 of 3)
© 2017 Cengage Learning. All Rights Reserved. May not be scanned, copied or duplicated, or posted to a publicly accessible website, in whole or in part.
Example 9.6: Distribution of Metal Strip Widths
in Manufacturing (slide 2 of 5)
SPSS Steps:
From the menus choose:
Analyze Descriptive Statistics Explore.
Select Width and move it onto the Dependent List field.
Click Plots and choose Histogram and Normality plots with tests.
The P-value of Lilliefors test is equal to 0.200 which is greater than
𝛼 = 0.05, then we fail to reject the null hypothesis that says the
width is normally distributed.
The histogram below confirms that the normal fit to the data
appears to be quite good.
© 2017 Cengage Learning. All Rights Reserved. May not be scanned, copied or duplicated, or posted to a publicly accessible website, in whole or in part.
Example 9.6: Distribution of Metal Strip Widths
in Manufacturing (slide 3 of 5)
SPSS Output:
© 2017 Cengage Learning. All Rights Reserved. May not be scanned, copied or duplicated, or posted to a publicly accessible website, in whole or in part.
Example 9.6: Distribution of Metal Strip Widths
in Manufacturing (slide 4 of 5)
© 2017 Cengage Learning. All Rights Reserved. May not be scanned, copied or duplicated, or posted to a publicly accessible website, in whole or in part.
Example 9.6: Distribution of Metal Strip Widths
in Manufacturing (slide 5 of 5)
Although the points in this Q-Q plot do not all lie exactly on a
45° line, they are about as close to doing so as can be
expected from real data. Therefore, there is no reason to
question the normal hypothesis for these data—the same
conclusion as from the Lilliefors test.
© 2017 Cengage Learning. All Rights Reserved. May not be scanned, copied or duplicated, or posted to a publicly accessible website, in whole or in part.