No Slide Title
Download
Report
Transcript No Slide Title
EXAM REVIEW
Copyright © 2008 Pearson Education, Inc. Publishing as Pearson
Addison-Wesley
MODELING THE DISTRIBUTION OF
SAMPLE PROPORTIONS (CONT.)
A picture of what we just discussed is as follows:
EXAMPLE
A candy company claims that its jelly bean mix
contains 45% red jelly beans.
Suppose that the candies are packaged at
random with 100 jelly beans each.
What’s the probability that a bag will contain
more than 50% red jelly beans?
Slide
1- 3
PROPORTIONS: ONE-SAMPLE - SE AND Z*
HYPOTHESIS TESTING
The conditions for the one-proportion z-test are the
same as for the one proportion z-interval. We test the
hypothesis
H 0: p = p 0
using the statistic
pˆ p0
z
SD pˆ
where
SD pˆ
p0 q0
n
When the conditions are met and the null hypothesis is
true, this statistic follows the standard Normal model,
so we can use that model to obtain a P-value.
Slide
1- 4
PROPORTIONS: ONE-SAMPLE CONFIDENCE INTERVALS
When the conditions are met, we are ready to find the
confidence interval for the population proportion, p.
The confidence interval is
pˆ z SE pˆ
where
ˆˆ
SE( pˆ ) pq
n
The critical value, z*, depends on the particular
confidence level, C, that you specify.
Slide
1- 5
PROPORTIONS: ONE-SAMPLE - EXAMPLE
A state university wants to increase its retention
rate of 10% for graduating students from the
previous year.
After implementing several new programs to
increase retention during the last two years, the
university re-evaluated its retention rate using a
random sample of 352 students.
The new retention rate was 12%.
Slide
1- 6
PROPORTIONS: ONE-SAMPLE - EXAMPLE
Test the hypothesis that the retention rate had
increased and state your conclusion with a 98%
confidence interval.
Also test the hypothesis with a z-test using a
significance level of 0.01
Slide
1- 7
MEANS: ONE-SAMPLE – T-TESTING
A practical sampling distribution model for
means
When the conditions are met, the standardized sample
mean
y
t
SE y
follows a Student’s t-model with n – 1 degrees of
freedom.
s
We estimate the standard error with SE y
n
Slide
1- 8
MEANS: ONE-SAMPLE - HYPOTHESIS
One-sided alternatives
Ha: μ>hypothesized value
Ha: μ <hypothesized value
Two-sided alternatives
Ha: μ ≠ hypothesized value
Slide
1- 9
MEANS: ONE-SAMPLE – CONFIDENCE INTERVALS
One-sample t-interval for the mean
When the conditions are met, we are ready to find the
confidence interval for the population mean, μ.
The confidence interval is
SE y
n1
where the standard error of the mean is
y t
s
SE y
n
*
The critical value tn1depends on the particular
confidence level, C, that you specify and on the number
of degrees of freedom, n – 1, which we get from the
sample size.
Slide
1- 10
MEANS: ONE-SAMPLE – EXAMPLE
A sociologist develops a test to measure attitudes
about public transportation, and 50 randomly
selected subjects are given the test.
Their mean score is 85 and their standard
deviation is 15.
Construct a 95% confidence interval for the mean
score of all such subjects.
Slide
1- 11
PROPORTIONS: TWO-SAMPLE - HYPOTHESIS
The typical hypothesis test for the difference in
two proportions is the one of no difference. In
symbols, H0: p1 – p2 = 0.
The alternatives:
Ha: p1 –p2 > 0
Ha: p1 –p2 < 0
Ha: p1 –p2 ≠ 0
Slide
1- 12
PROPORTIONS: TWO-SAMPLE - SE AND Z*
HYPOTHESIS TESTING
We use the pooled value to estimate the standard error:
pˆ pooled qˆ pooled pˆ pooled qˆ pooled
SE pooled pˆ1 pˆ 2
n1
n2
Now we find the test statistic:
pˆ1 pˆ 2 0
z
SE pooled pˆ1 pˆ 2
When the conditions are met and the null hypothesis is
true, this statistic follows the standard Normal model,
so we can use that model to obtain a P-value.
Slide
1- 13
CALCULATING THE POOLED PROPORTION
The pooled proportion is
pˆ pooled
where
Success1 Success2
n1 n2
Success1 n1 pˆ1
and
Success2 n2 pˆ 2
If the numbers of successes are not whole numbers, round
them first. (This is the only time you should round values in
the middle of a calculation.)
Slide
1- 14
PROPORTIONS: TWO-SAMPLE CONFIDENCE INTERVALS
When the conditions are met, we are ready to find the
confidence interval for the difference of two proportions:
The confidence interval is
pˆ1 pˆ 2 z
SE pˆ1 pˆ 2
where
SE pˆ1 pˆ 2
pˆ1qˆ1 pˆ 2 qˆ2
n1
n2
The critical value z* depends on the particular confidence
level, C, that you specify.
Slide
1- 15
PROPORTIONS: TWO-SAMPLE - EXAMPLE
A survey of randomly selected college students
found that 50 of the 100 freshman and 60 of the
125 sophomores surveyed had purchased used
textbooks in the past year.
Test for a difference between the two student
groups using a significance level of 0.05.
Slide
1- 16
MEANS: TWO-SAMPLE - HYPOTHESIS
One-sided alternatives
Ha: μ1 – μ2 >0
Ha: μ1 – μ2 <0
Two-sided alternatives
Ha: μ1 – μ2 ≠ 0
Slide
1- 17
MEANS: TWO-SAMPLE – T-TESTING
When the conditions are met, the standardized sample
difference between the means of two independent groups
y1 y2 1 2
t
SE y1 y2
can be modeled by a Student’s t-model with a number of
degrees of freedom found with a special formula.
We estimate the standard error with
SE y1 y2
s12 s22
n1 n2
MEANS: TWO-SAMPLE – DEGREES OF
FREEDOM
The special formula for the degrees of freedom for
our t critical value is a bear:
2
s12 s22
n1 n2
df
2
2
1 s12
1 s22
n1 1 n1 n2 1 n2
Because of this, we will let technology calculate
degrees of freedom for us!
Slide
1- 19
MEANS: TWO-SAMPLE – CONFIDENCE INTERVAL
When the conditions are met, we are ready to find the confidence
interval for the difference between means of two independent
groups.
The confidence interval is
y1 y2 t
df
SE y1 y2
where the standard error of the difference of the means is
SE y1 y2
s12 s22
n1 n2
The critical value depends on the particular confidence level, C, that you
specify and on the number of degrees of freedom, which we get from the
Slide
sample sizes and a special formula.
1- 20
MEANS: TWO-SAMPLE – EXAMPLE
Two types of cereal brands are being tested for
sugar content
Brand Yummy – n=100, Ӯ=5, s=2
Brand Yuck – n=150, Ӯ=4.5, s=2
Construct a 95% confidence interval for the
difference between the two brands.
Slide
1- 21
UPCOMING IN CLASS
Exam #2 Wednesday
Data Project Due by 5pm Thursday December 5th
via email or my department mailbox.
Finals (optional)
Wednesday December 11th
1-3pm