0% found this document useful (0 votes)
23 views7 pages

Statistical Hypothesis Testing Problems

1. The document provides instructions and examples for problems related to hypothesis testing using two-sample cases from chapters 9 and 10. Several examples are provided that demonstrate setting up null and alternative hypotheses and analyzing the results of t-tests and z-tests. 2. One example analyzes assets data from mutual funds that do and do not charge fees. The null hypothesis is that the mean assets are the same between the two groups, and the alternative is that the means are different. The analysis shows the data does not meet the normality assumption for the group with fees. However, the ANOVA p-value of 0.3817 is well above the significance level of 0.05, so the null hypothesis cannot be rejected

Uploaded by

Lily Tran
Copyright
© All Rights Reserved
We take content rights seriously. If you suspect this is your content, claim it here.
Available Formats
Download as DOC, PDF, TXT or read online on Scribd
0% found this document useful (0 votes)
23 views7 pages

Statistical Hypothesis Testing Problems

1. The document provides instructions and examples for problems related to hypothesis testing using two-sample cases from chapters 9 and 10. Several examples are provided that demonstrate setting up null and alternative hypotheses and analyzing the results of t-tests and z-tests. 2. One example analyzes assets data from mutual funds that do and do not charge fees. The null hypothesis is that the mean assets are the same between the two groups, and the alternative is that the means are different. The analysis shows the data does not meet the normality assumption for the group with fees. However, the ANOVA p-value of 0.3817 is well above the significance level of 0.05, so the null hypothesis cannot be rejected

Uploaded by

Lily Tran
Copyright
© All Rights Reserved
We take content rights seriously. If you suspect this is your content, claim it here.
Available Formats
Download as DOC, PDF, TXT or read online on Scribd

Problem Set 5 (with instructions)

Chapters 9 and 10 (2-sample case)

Chapter 9:

1. Problems 9.73 (a,b); (1 pt) Just read the problem and set up appropriate null and alternative
hypotheses (this will give you practice in recognizing problems. Do not work out the entire problem,
just set up hypotheses.
a. H0 : μ>= $100
mean reimbursement for office visits paid for by Medicare is equal to or more than
$100.
H1: μ < $100
mean reimbursement for office visits paid for by Medicare is less than $100.
b. H0 : π =< 0.10
Proportion of incorrect reimbursement in the population is equal to or less than10%.
H1 : π > 0.10
Proportion of incorrect reimbursement in the population is greater than 10%.

2. Problem 9.33 ( 2 points)

Data
Null Hypothesis = 0
Level of Significance 0.05
Sample Size 100
Sample Mean 0.000165
Sample Standard Deviation 0.001998806

Intermediate Calculations
Standard Error of the Mean 0.0002
Degrees of Freedom 99
t Test Statistic 0.8255

Two-Tail Test  
Lower Critical Value -1.9842
Upper Critical Value 1.9842
p-Value 0.4111
Do not reject the null hypothesis  

a. . H0: μ = 0.0 inches


The mean difference is equal to 0.0 inches.
H1 : μ ≠ 0.0 inches
The mean difference is different from 0.0 inches

P value of 0.4111 is greater than Level of significance of 0.05. So there is not enough evidence to
reject the null. We can only be 58.89% confident in H1.

Data
Sample Standard Deviation 0.001998806
Sample Mean 0.000165
Sample Size 100
Confidence Level 95%

Intermediate Calculations
Standard Error of the Mean 0.000199881
Degrees of Freedom 99
t Value 1.9842
Interval Half Width 0.0004
   
Confidence Interval
Interval Lower Limit 0.00
Interval Upper Limit 0.00
b. The confidence of 58.89% is lower than the required confidence of 95% so there is not
enough evidence to reject the null hypothesis.
c. In both problems a and b, the conclusions are the same. In problem b, the hypothesized
mean of 0 is between the confidence intervals of 0 and 0 so we are unable to reject the
null. In problem a, there is no evidence to reject the null since the P value is greater than
the level of significance.
d. Due to the sample size being 100, which is more than 30, we do not have to be concerned
about the normality assumption.

2
e. 3. Problem 9.57: (1 point) Use 1 sample test – for a proportion. Make sure you state
hypotheses and write up a conclusion stating how confident you are in H 1.

Data
Null Hypothesis = 0.55
Level of Significance 0.05
Number of Items of Interest 9
Sample Size 17

Intermediate Calculations
Sample Proportion 0.529411765
Standard Error 0.1207
Z Test Statistic -0.1706

Two-Tail Test  
Lower Critical Value -1.9600
Upper Critical Value 1.9600
p-Value 0.8645
Do not reject the null hypothesis  

H0 : π = 55%
The proportion of females in this position at the medical center is equal to 55%.
H1 : π ≠ 55%
The proportion of females in this position at the medical center is different from 55%.

Since p value of 0.8645 is greater than level of significance of 0.05, it does not reject the null
hypothesis. We can only be 13.55% confident in H1.

4. Problem 9.29 (1.5 points)

Data
Null Hypothesis = 45
Level of Significance 0.05
Sample Size 27
Sample Mean 43.88888889
Sample Standard Deviation 25.28352053

Intermediate Calculations
Standard Error of the Mean 4.8658
Degrees of Freedom 26
t Test Statistic -0.2284

Two-Tail Test  
Lower Critical Value -2.0555
Upper Critical Value 2.0555

3
p-Value 0.8212
Do not reject the null hypothesis  

a. H0 : μ = 45 days
The mean processing time is equal to 45 days.
H1 : μ ≠ 45 days
The mean processing time is different from 45 days.

Because the p-value of 0.8212 is greater than level of significance of 0.05, there is not enough
evidence to reject the null hypothesis. We can only be 17.88% confident in H 1.
b. The population distribution must be assumed to be normal in order to conduct the t
test.

Time
Mean 43.88888889
Median 45
Mode 17
Minimum 16
Maximum 92
Range 76
Variance 639.2564
Standard Deviation 25.2835
Coeff. of Variation 57.61%
Skewness 0.5173
Kurtosis -0.9041

4
Count 27
Standard Error 4.8658

c. The skewness is 0.5173 and the kurtosis is -0.9041 which is between -1 and 1
therefore it is normally distributed.
d. Without checking for the skewness and kurtosis to determine normality, since the
sample size is only 27, the assumption would not be valid since it is does not have 30
values.
Chapter 10

5. File: MUTUALFUNDS_S13 in Data for problem sets folder (you may do this as a group).
(4.5 points)

Assets Report

1. H0 = mean assets for fee funds is the same as no-fee funds. μ fee = μno-fee
H1 = mean assets for fee funds is different than no-fee funds. μ fee ≠ μno-fee
Assets

No Yes
Mean 2082.602128 3352.822222
Median 443.15 242
Mode #N/A #N/A
Minimum 5 1.6
Maximum 35008 45421.1
Range 35003 45419.5
Variance 29970409.4247 93698300.6064
Standard Deviation 5474.5237 9679.7883
Coeff. of Variation 262.87% 288.71%
Skewness 4.3029 3.8475
Kurtosis 19.6925 15.1345
Count 94 27
Standard Error 564.6540 1862.8761

2. RoT for Normality assumption satisfies for No fee because there are more than
30 data points (94). But RoT for Normality assumption does not satisfy for
Yes(fee) because both skewness and kurtosis are over 1 and also it has only 27
data points. So we cannot be confident in the accuracy of p-value.
Assets

SUMMARY
Groups Count Sum Average Variance
No 94 195764.6 2082.602128 29970409.4247
Yes 27 90526.2 3352.822222 93698300.6064

ANOVA

5
P-
Source of Variation SS df MS F value F crit
Between Groups 33842637.7433 1 33842637.7433 0.7710 0.3817 3.9208
Within Groups 5223403892.2662 119 43894150.3552

Total 5257246530.0096 120


Level of significance 0.05

3. P-value of 0.3817 is greater than level of significance of 0.05. So we cannot


reject the null hypothesis. Although the Normality assumption does not satisfy for
the Yes (fee) data, we can be confident in this conclusion because the p-value is
very much larger than .05(level of significance) and any inaccuracy is unlikely to
affect this result.
4. We have about 61.83% confidence implying that there is insufficient evidence to
conclude that the average assets for Fee funds are different than the average
assets for No-fee funds. Because the p-value is much larger than the required
level of .05 we can be confident in this result even though the Normality
assumption does not satisfy for the Yes (Fee funds) data.

Returns Report

1. H0 = mean refund for fee funds is the same as no-fee funds. μ fee = μno-fee
H1 = mean refund for fee funds is different than no-fee funds. μ fee ≠ μno-fee

Returns

No Yes
Mean 42.59468085 42.71481481
Median 40.7 42.3
Mode 37.5 29.6
Minimum 14.9 27.2
Maximum 78 64.3
Range 63.1 37.1
Variance 185.9734 119.2105
Standard Deviation 13.6372 10.9184
Coeff. of Variation 32.02% 25.56%
Skewness 0.3574 0.3048
Kurtosis -0.2789 -0.9522
Count 94 27
Standard Error 1.4066 2.1012

2. RoT for Normality assumption satisfies for No-fee, because both skewness and
kurtosis are between -1 and 1, also it has more than 30 data points. RoT for
Normality assumption also satisfies for Yes (fee funds), because both skewness
and kurtosis are between -1 and 1. We can be confident in the accuracy of p-
value.

6
Returns Calculations
c 2
SUMMARY n 121
Groups Count Sum Average Variance
No 94 4003.9 42.59468085 185.9734
Yes 27 1153.3 42.71481481 119.2105

ANOVA
P-
Source of Variation SS df MS F value F crit
Between Groups 0.3027 1 0.3027 0.0018 0.9665 3.9208
Within Groups 20395.0014 119 171.3866

Total 20395.3041 120


Level of significance 0.05

3. P-value of 0.9665 is greater than level of significance 0.05. We cannot reject the
null hypothesis. We can conclude that mean return for fee funds is same with
mean return for no-fee funds.

4. We have only about 3.35% confidence implying that there is insufficient evidence
to conclude that the average return for Fee funds is different than the average
return for No-fee funds.

You might also like