BITS Pilani

Pilani Campus

Inferences on the Mean and Variance

Chapter 8
BITS Pilani, Pilani Campus
Interval estimation of Variability

Recall: S2 is an unbiased estimator for 2.

Theorem 8.1.1: Let X1, …, Xn be the random sample
of size n from a normal population
n
with mean  and
s.d. . Then (n  1) S2 /  2   ( X i  X )2 /  2
i 1
has chi-squared distribution with (n-1) degrees of
freedom
Proof: will be discussed in class!
Recall: Chi-squared dist with (n-1) degrees of freedom
is gamma dist with  = (n-1)/2,  = 2.
BITS Pilani, Pilani Campus
Recall Chi-squared Distribution

3
BITS Pilani, Pilani Campus
 Interval estimation of Variability
Theorem 8.1.2: Let X1, …, Xn be the random sample of size
n from a normal population with mean  and s.d. . Then,
100(1-)% confidence interval for 2 is given by
(n  1) S 2 (n  1) S 2
 2 
2 /2 12 /2
Proof:
(n  1) S 2
P(  df ,1  
2
 2
)  1
 2  2 df ,
2

(n  1) S 2 (n  1) S 2
2 
 df2 ,  df2 ,1 
2  2
BITS Pilani, Pilani Campus
 Interval estimation for σ2


Recall that 100  1    % CI for  is 
 n  1 S 2
 n  1 S 2

 .
2
,
 2 2 1 2 
2


Ex.8.1. A sample of size 9 from a normal population is given below. Find

the 90% CI for the mean  of the population. Also find 90% CI for the
variance  2 of the population. Sample: 0, 1, -1, 1, 1, 0, -1, - 2, 3.
5
BITS Pilani, Pilani Campus
 Interval estimation for σ2
Sol. Here n  9, s 2 =0.24 (check), and   0.10.
To find the CI for the population variance  2 , first let us
calculate 12 2, n 1 =  0.95,8
2
 2.733 and 2 2, n 1 =  0.2 05,8  15.507. Thus,
  n  1 S 2  n  1 S 2 
the 90% CI is 
 2 2
,    0.1238, 0.7025 
 1 2 
2

The length of the CI is  0.7025  0.1238  =0.5787

HW.8.1. The heights in inches of 8 students of a college, chosen at random,

were as follows: 62.2, 62.4, 63.1, 63.2, 65.5, 66.2, 66.3, 66.5. Compute 90%
and 95% CI for the variance of the population of heights, assuming
it be to be normal. Also, find the length of the interval in each case.
6 BITS Pilani, Pilani Campus
 α/2
Degree of 1-α/2
freedom Chi-sq(α/2)

Cumulative Probability

Similar to the
table (page
695) of your
book
7 BITS Pilani, Pilani Campus
 Interval estimation for σ2
HW 8.2 (Q7). Recent research indicates that heating
and cooling commercial buildings with ground water
source heat pumps is economically sound. The crucial
random variable being studied is the water
temperature. A sample of 15 wells in the state of
California yields a sample standard deviation of 7.5o F.
Find a 95% confidence interval on the standard
deviation in temperature of wells in California.
Assume normality of population.

BITS Pilani, Pilani Campus

 One Sided CI on σ2
Ex 8.2 (Q.4, Page 289): X is actual length of 63 mm nails.
Use the given data to find a 95% one side confidence
interval on the variance in length
63.0 63.1 63.0 63.0 62.9 63.0 63.0
63.1 62.8 63.1 63.1 63.0 62.9 63.2
The manufacturer wants to check to be sure that the
population variance of the length of nails being produced
does not exceed 0.03. Assuming normality of population,
does this sample indicate that this is in the case? Explain.
Sol:

BITS Pilani, Pilani Campus

 Interval estimation for σ2
14 14

 X i  882.2;
i 1
 i
X 2

i 1
 55591.34; S 2
 0.0105

We want L, such that P   L   0.95

 2

(n  1) S 2 13  0.0105
 L 2   0.023
1 ,13 5.89

Yes, we are 95% confident that the true population

variance of length 63 mm nails is less then 0.023,
therefore less then 0.03.
BITS Pilani, Pilani Campus
 Summary So Far…
• What is CI? What is the level of confidence?

• CI for population mean, when population s.d. is

known – do we need normality of population?

• CI for population variance – do we need normality

of population?

• Two sided and one sided CI

11 BITS Pilani, Pilani Campus
Interval estimation for µ: σ unknown

Interval Estimate for µ = X + /  Margin of Error

• If the population standard deviation σ is not known, then?

 We can use estimator, i.e., sample standard deviation S to
calculate the margin of error.
• What is the distribution of RV obtained by replacing σ by S?
 In this case, we would obtain a T-distribution – HOW??.
• Here also, we shall assume that our population is
normally distributed, WHY??
12 BITS Pilani, Pilani Campus
 Interval estimation for µ: σ unknown

Definition :
Let Z N  0,1 and 2 be an independent chi-squared RV with  degrees
Z
of freedom, then T  follows a T -distribution with  dof.
 
2

Theorem 8.2.1 :
Let X 1 , X 2 , , X n be a random sample of size n from a normal distribution
X 
with mean  and variance  . Then,2
Tn 1
S n

13 BITS Pilani, Pilani Campus

 T-Distribution
• Random variable T with  degrees of freedom (called
parameter) is a continuous r.v. with density
 ( 1) / 2
(  1) / 2  t 2

f (t )  1   ;  t  .
( / 2)    

• Density plot is bell shaped, symmetric about 0.

• Variance of T decreases as  increases. In fact T
approximates standard normal for large .

BITS Pilani, Pilani Campus

T-Distribution

BITS Pilani, Pilani Campus

 CDF of T-distribution
• Cdf is given by tables (Table VI, p.699).
• Like chi-squared table,  is the label of rows,
F(t) is label of columns, at their intersection is
the value of t for γ degrees of freedom.
• By tr we denote the value of the t-variable
such that area under its density to its right is r.
(degrees of freedom must be mentioned
separately).

BITS Pilani, Pilani Campus

 Problem Solving: Ex. 9 (pp. 290)

(g) Using T-table find t such that P[-t  T25  t] =0.90.

By symmetry, need to find t0.05 for 25 degrees of
freedom. Equivalently, F(t0.05)=0.95 for 25 degrees
of freedom.
Look in T-table in row corresponding to and
column corresponding to 0.95 .
This gives t = 1.708.
(k) P[T16  - t] =0.05 implies P[T16 ≥ t] = 0.05, i.e.
P[T16  t] =0.95, from table, t = 1.746.

BITS Pilani, Pilani Campus

 Interval estimation for µ: σ unknown

The T -distribution is symmetric, and becomes approx. std. normal for large n.
Taking two points  t 2 symmetrically about the origin, we get
 
 X  
P  t 2   t 2   1  
 S 
 
 n 
Here 1    is known as confidence level, and  is the level of significance. So,
 S S 
P X  t 2    X  t 2   1  
 n n 
Hence, the CI for   unknown  having confidence level 100  1    %
 S S 
is given as  x  t 2,n 1 , x  t 2, n 1  .
 n n 
18 BITS Pilani, Pilani Campus
Interval estimation for µ: σ unknown

BITS Pilani, Pilani Campus

 Examples
Ex.8.3. Seven laboratory experiments of the value of g (accelearation due
to gravity that follows normal distribution) at Pilani gave a mean 977.51 cm/s 2
and a s.d. 4.42 cm/s 2 . Find 95% CI for the true value of g (i.e., population mean).
Sol.
Step 1: Here n  7, x  977.51, s =4.42, and   0.05. We need CI for .
Population is also known to be normal dist.
Step 2: As   0.05, we need to find t 2 from t-distribution with (n -1)  6

 
degree of freedom, such that P T  t 2 =0.975.
From t-distribution table, we see t0.025,6  2.447.
 s s 
Step 3: The CI for   unknown  is  x  t0.025 , x  t0.025 
 n n 
  973.09,981.93
20 BITS Pilani, Pilani Campus
 α/2
1-α/2
Degree of
freedom tα/2

Cumulative Probability

Similar to
the table
(page 699)
of your
book

BAZC413: Introduction to Statistical Methods 21 BITS Pilani, Pilani Campus

 Examples
HW.8.3. Seven laboratory experiments of the value of g (accelearation due
to gravity that follows normal distribution) at Pilani gave a mean 977.51 cm/s 2
and a s.d. 4.42 cm/s 2 . Find 80%, 90% and 95% CIs for the population mean.
Can you find out the respective margin of errors?

HW.8.4. A random sample of size 49 taken from a N   ,  2  has a mean 40,

and s.d. 5. Obtain a 95% CI for  2  3 .

HW.8.5. A sample of size 15 taken from a larger population (normally dist.)

has a sample mean 12, and sample variance 25. Construct 95% CI for population
mean and s.d. 5. Obtain a 95% CI for population mean  . What is the length of CI?

22 BITS Pilani, Pilani Campus

 Examples
HW 8.6 (Ex.14, page 292) To estimate the average number of pounds of copper
recovered per ton of ore mine, a sample of 150 tons of ores is monitored.A sample
mean of 11 pounds with a sample s.d. of 3 pounds was obtained.Construct a 95%
confidence interval on the mean number of pounds of copper recovered per ton of
ore mined. Assume normaility of the population.
Sol.
Step 1: Here n  150 (large), x  11, s =3, and   0.05.
We need CI for   unknown  .
Step 2: As   0.05, we need to find t0.025,149 . However, for large degrees of
freedom, we can approximate T-distribution value t0.025,149 by Z-dist value z0.025 .
 s s 
Step 3: The CI for   unknown  is  x  z0.025 , x  z0.025 
 n n 
 10.52,11.48 
23 BITS Pilani, Pilani Campus
 Examples
HW 8.7. A credit card company wants to determine the mean income of its card
holders. It also wants to find if there are any differences in mean income between
males and females. A random sample of 225 male card holders and 190 female
card holders was drawn. The mean and s.d. for male card holders are Rs. 16450
and Rs. 3675, resp. For female card holders, these are Rs. 13220 and Rs. 3050.
(i) Calculate 90% and 95% CIs for mean income of male and female card holders.
(ii) Do you think that, on average, males' and females' incomes differ?

HW.8.8. A reporter for a student newspaper is writing an article on the cost of

off-campus housing. A sample of 16 one-bedroom apartments within a half-mile
of campus resulted in a sample mean rent (in rs.) of 750 per month and a sample
s.d. of 55. Find 70%, 95%, and 99% CIs estimate of the mean rent per month for
the population. Assume that the population is normally distributed.
What would happen if sample size is 160, with same sample mean and sample s.d ?
24 BITS Pilani, Pilani Campus
 One Sided CI
HW 8.9 (Q17). One sided confidence interval can be
used to approximate the maximum and minimum
value of the population mean.
An interval (-∞,L1] such that P(μ ≤ L1) = 1-α allows us
to place bounds on the maximum value of population
mean L  X t S / n
1 

An interval [L2, ∞) such that P(μ ≥ L2) = 1-α allows us

to place bounds on the minimum value of population
mean L  X t S / n
2 
BITS Pilani, Pilani Campus
 One Sided CI
Use the following data on X, the time that a
commercial airliner stays at the gate during a through
flight, to find a 95% one sided confidence interval that
puts a bound on the minimum time in minutes for μ:
25 29 32 37 40 27 30 35 38 41 42 45 45 47
49 50 55 53 60
Solution : x  41.05, s 2  98.61, s  9.93
s 9.93
Lobs  x  t / 2  41.05  1.734   37.10
n 19
CI  [37.10, )
BITS Pilani, Pilani Campus
Sample size required to estimate
μ with specified error
HW 8.10 (Q 19 a). (A consequence of central limit
theorem) We can assert with 100(1-)% confidence
that sample mean from a sample of size n differs from
population mean by at most d if the sample size
n ≥ (z/2)2 2/d2, i.e., using the sample mean of any
sample of such size, we can estimate the population
mean within d units with 100(1-)% confidence.

BITS Pilani, Pilani Campus

Sample size required to estimate
μ with specified error
Q 19. (b) A preliminary pilot study is run, and an
estimated standard deviation of 500 units is obtained.
How large a sample is needed to estimate μ to within
50 units with 95% confidence?

(c) If rough estimate of σ is 0.75. How large a sample

is needed to estimate μ to within 0.1 with 90%
confidence?

BITS Pilani, Pilani Campus

Summary of Interval Estimation for µ

29 BITS Pilani, Pilani Campus