ME614 PROBABILITY & STATISTICAL CONCEPTS IN ENGINEERING DESIGN
CONFIDENCE INTERVAL ESTIMATION
Two Types of Estimates
Point Estimate – the value of a single sample statistic
Confidence Interval Estimate – range of numbers, called an interval, constructed around the point estimate. It
is constructed such that the probability that population parameter is located somewhere within the interval is
known.
Terminologies:
Level of Confidence – symbolized by (1 – α) x 100%, where α is the proportion in the tails of the distribution
that is outside the confidence interval.
Critical Value – value of Z (or t) needed for constructing a confidence interval.
Lesson 1: Confidence Interval Estimate for the Mean (σ known)
z
CI x
n
#1 Interpolation methods are used to estimate heights above sea level for locations where direct measurements are
unavailable. In the article “Transformation of Ellipsoid Heights to Local Leveling Heights” (M. Yanalak and O. Baykal,
Journal of Surveying Engineering, 2001: 90-103), a second-order polynomial method of interpolation for estimating
heights from GPS measurements is evaluated. In a sample of 74 locations, the errors made by the method averaged 3.8
cm, with a standard deviation of 4.8 cm.
a. Find a 95% confidence interval for the mean error made by this method.
b. Find a 99% confidence interval for the mean error made by this method.
c. A surveyor claims that the mean error is between 3.2 and 4.4 cm. With what level of confidence can this
statement be made?
d. Approximately how many locations must be sampled so that a 95% confidence interval will specify the
mean to within ± 0.7 cm?
#2 Resistance measurements were made on a sample of 81 wires of a certain type. The sample mean resistance was
17.3 mΩ, and the standard deviation was 1.2 mΩ.
a. Find a 95% confidence interval for the mean resistance of this type of wire.
b. What is the level of confidence interval (17.1, 17.5)?
c. How many wires must be sampled so that a 95% confidence interval will specify the mean to within ±0.1
mΩ?
Lesson 2: Confidence Interval Estimate for the Mean (σ unknown)
ts
CI x
n
William S. Gosset wanted to make inferences about the mean when σ was unknown. Because employees of Guinness
Breweries in Ireland were not permitted to publish research work under their own names, Gosset adopted the
pseudonym “Student.” The distribution that he developed is known Student’s t distribution. The t-statistic with n-1 as
degrees of freedom is computed as
(x )
t
s/ n
Degrees of Freedom (n – 1)
To compute for the sample variance S 2, you must know x . Therefore, only n – 1 sample values are free to vary. This
means that you have n – 1 degrees of freedom.
Central Limit Theorem
If X1, X2, …, Xn is a random sample of size n taken from a population (either finite or infinite) with mean µ and finite
variance σ2, and if x is the sample mean, the limiting form of the distribution of
(x )
Z
/ n
as n , is the standard normal distribution.
#1 The solar energy consumed (in trillion BTU) in the U.S. by year from 1989-2004 (source: US Department of Energy
web site http://www.eia.doe.gov/emeu) is shown in the table below. Read down, then right for year.
55.291 66.458 70.237 65.454
59.718 68.548 69.787 64.391
62.688 69.857 68.793 63.620
63.886 70.833 66.388 63.287
Construct the 95% confidence interval for the mean solar energy consumed.
#2 A machine produces metal rods used in an automobile suspension system. A random sample of 15 rods is selected
and the diameter is measured. The resulting data (in millimeters) are as follows:
8.24 8.25 8.20 8.23 8.24
8.21 8.26 8.26 8.20 8.25
8.23 8.23 8.19 8.28 8.24
Calculate the 95% confidence interval on mean diameter.
Lesson 3: Confidence Interval Estimate for the Proportion
p (1 p )
CI p z
n
#1 The fraction of defective integrated circuits produced in a photolithography process is being studied.
A random sample of 300 circuits is tested, revealing 13 defectives. Calculate the 90% confidence interval on the fraction
of defective circuits produced by this particular tool.
#2 A random sample of 50 suspension helmets used by motorcycle riders and automobile race-car drivers was
subjected to an impact test, and on 18 of these helmets some damage was observed.
a. Find the 95% confidence interval on the true proportion of helmets of this type that would show damage from
this test.
b. Using the point estimate of p obtained from the preliminary sample of 50 helmets, how many helmets must be
tested to be 95% confident that the error in estimating the true value of p is less than 0.02?
