Anda di halaman 1dari 2

Applied Regression Analysis, Assignment 1

DEPARTMENT OF ENGINEERING MATHEMATICS


APPLIED REGRESSION ANALYSIS
ENGM 6671
ASSIGNMENT # 1
1. A car manufacturer plans on using their current minivan engine in a new line of Sport Utility
Vehicles they are developing. The mean fuel consumption rate is dependent on, amongst other
things, the vehicle mass, rolling resistance, wind resistance, and driver agressiveness. In an
effort to determine the mean highway fuel consumption rate for this vehicle, an engineer sends
25 of the vehicles out on separate highway test routes. The sample mean fuel consumption
rate is determined to be 7.8 /100 km with a standard deviation of 0.9/100 km.
(a) Find a 95% confidence interval for the mean highway feul consumption rate for the vehicle.
(b) Suppose that the engineer wants to be 95% confident that the estimated mean highway
fuel consumption rate is within 0.1 /100 km of the true mean highway consumption rate.
How many vehicles should be sent out?
2. A study of the electromechanical protection devices used in electrical power systems showed
that of 193 devices that failed when tested, 75 were due to mechanical parts failures. (Reliability of Protection Equipment in Operation, H. Hubensteiner, Brown Boveri Review, February,
1983, pp. 111-114.)
(a) Find a point estimate of p, the proportion of failures that are due to mechanical failures.
(b) Find a 95% confidence interval for p.
(c) How large a sample is required to estimate p to within .03 with 95% confidence?
3. A machine is producing cylindrical shafts. The specifications for the shafts call for a nominal
diameter of 5 cm and the standard deviation of diameter is to be at most 0.1 cm. A random
sample of 10 shafts have diameters as follows: 5.263, 5.079, 5.003, 4.811, 5.048, 4.945, 5.244,
5.055, 5.253, 5.011 (see the file diameter.txt).
(a) Compute a 95% confidence interval for the mean diameter of the shafts. Do you think
the machine is producing shafts which meet the specifications as to nominal diameter ?
(b) Test the hypothesis that the machine is producing shafts which meet the specifications
as to nominal diameter. What is the p-value of the test?
(c) Compute a 95% confidence interval for the variance of the diameter of the shafts. Do
you think the machine is producing shafts which meet the specifications as to standard
deviation?
(d) Test the hypothesis that the machine is producing shafts which meet the specifications
as to standard deviation. What is the pvalue of the test?
4. A chemical engineer is attempting to assess the concentration of lead remaining unabsorbed
from a gas after passing it over a catalyst. This will be done by measuring the remaining lead
content in the gas, in parts per million. Eight measurements of the lead content in the gas
after passing it over the catalyst are stored in the file lead.txt.
(a) Assuming that the unabsorbed lead content is (at least approximately) normally distributed, construct a 95% confidence interval for the mean unabsorbed lead content.
(b) The engineer is hoping that the catalyst will reduce the mean unabsorbed lead content to
0.830 parts per million (which is what the competitor is claiming their catalyst achieves).
Does it seem likely that the catalyst is achieving this goal? Explain your answer by
refering to the confidence interval found above.

Applied Regression Analysis, Assignment 1

(c) In the situation described the engineer is interested in the lower limit of the lead content.
Test the hypotheis: H0 : = 0.830 versus the one sided alternative H1 : > 0.830 using
the 5% level of significance.
(d) How can you reconcile your answers to b) and c)?
5. Epidemiologists have theorized that the risk of coronary heart disease can be reduced by an
increased consumption of fish. One study, begun in 1980, monitored the diet and health
of a random sample of middle-aged men. The men were divided into groups according to
the number of grams of fish consumed per day. Twenty years later, the level of HDL (good)
cholesterol present in each was recorded. A subset of the results are summarized in the following
table

Sample Size
Sample Mean
Sample Stdev

No Fish Consumption
0 grams/day
29
1.10
0.66

High Fish Consumption


45 grams/day
21
1.58
0.75

(a) Find 95% confidence intervals for the mean and standard deviation of each group.
(b) Based on the confidence intervals in a) can we say that fish consumption changes the
mean HDL cholesterol level?
(c) Use the 2sample t test with equal variances to test the hypotheses H0 : f ish nof ish = 0
versus H1 : f ish nof ish 6= 0. What is the pvalue of the test?
(d) How can you reconcile your answers to b) and c)?
(e) It would seem that a one sided test of hypotheses would be appropriate for this situation.
Test the hypotheses H0 : f ish nof ish = 0 versus H1 : f ish nof ish > 0. What is
the pvalue of the test?
6. A new coal liquefaction process is being studied. It is claimed that the new process results in
higher yield of distillate synthetic fuel than the current process. The observations, stored in
the file fuel.txt, were obtained on the number of kilograms of distillate synthetic fuel produced
per kilogram of hydrogen consumed in the process. (Liquefaction Process Promised Better
Efficiency, Modern Power Systems, May 1983, p. 13.)
(a) Assuming that these two random variables have the same standard deviation, find the
pooled standard deviation for the two data sets.
(b) Find a 95% confindence interval for the difference of mean distillate.
(c) Test the hypothesis that the new process results in higher yield. What is the p-value of
the test?
(d) Would you recommend the new process?
7. A study was conducted to decide whether a new statistical package has lower cost than the
one currently in use. To do so, 15 data sets are used. Each is analyzed by each package and
the cost of the analysis is recorded. The observations are stored in the file cost.txt.
(a) Find 95% confidence intervals for the costs when using the new and old packages. Can
we determine whether the new package has lower cost than the old one based on these
intervals?
(b) Find a 95% confidence interval of the difference of costs. Can we determine whether the
new package has lower cost than the old one based on this interval?
(c) Carry out a test of hypotheses to determine whether the new package has lower cost than
the old one. What is the pvalue?

Anda mungkin juga menyukai