Session 9: Two-Sample

Hypothesis Tests
Statistics for Business
Dr. Le Anh Tuan

3
Contents
►Two-sample tests

►Two independent samples

►σ12 and σ22 Known

►σ12 and σ22 Unknown and Assumed

Equal

►σ12 and σ22 Unknown and Assumed

Unequal

4
Two-sample tests
► What Is a Two-Sample Test?
► A two-sample test compares two sample estimates with
each other
► Mean, median, variance, proportion.

Two Sample Tests

Population Means, Population Means,

Population
Dependent Independent
Proportions
Samples Samples

Same group Group 1 vs. Proportion 1 vs.

before vs. after independent Proportion 2
treatment Group 2

5
Dependent vs Independent

►Two samples are independent if the sample

values from one population are not related to or
somehow naturally paired or matched with the
sample values from the other population.

►Two samples are dependent (or consist of

matched pairs) if the sample values are
somehow matched where the matching is based
on some inherent relationship.

6
 Two independent samples

7
Two independent samples

Population means,
independent
samples

!" # $%& !# #
σx2 and σy2 known Test statistic is a z value
$'( )%*+%
! # $%& !# #
σx2 and" σy2 unknown
$'( ,%)%*+%
! # $%& ! #
σ"x2 and σy 2 #
$--,.(&
assumed (/,$0
equal Test statistic is a a value from the
2#and # Student’s t distribution
σ!x" $%&
σy !#
2
assumed unequal
$--,.(& ,%(/,$0

8
σ12 and σ22 Known
►Assumptions:
►Samples are randomly and independently drawn
►Both population distributions are normal
►Population variances/SD are known or each sample
size must be at least 30 (CLT).

►For the population 1, we let:

►!1 = $%$&'()*%+ ,-(+
► /̅ 1 = 0(,$'- 1 ,-(+
►12 = $%$&'()*%+ 1 34
►02 = 0(,$'- 1 34
►+2 = 0*5- %6 0(,$'- 1

►The corresponding notations !2, /̅ 2 , 18 , 08 , +8 , apply to

population 2.
9
 Hypothesis

10
 σ12 and σ22 Known
►When σ12 and σ22 are known and both populations are
normal, the standard deviation of sample (1) – (2) is

►And z-statistic:

11
 σ12 and σ22 Known

Two Population Means, Independent

Samples, Variances Known
Lower-tail test: Upper-tail test: Two-tail test:
H0: μ1 – μ2 ³ 0 H0: μ1 – μ2 ≤ 0 H0: μ1 – μ2 = 0
H1: μ1 – μ2 < 0 H1: μ1 – μ2 > 0 H1: μ1 – μ2 ≠ 0

a a a/2 a/2

-za za -za/2 za/2

Reject H0 if z < -za Reject H0 if z > za Reject H0 if z < -za/2
or z > za/2

|z|>!"/$
12
 Example
► A high school teacher claims that students in class A will score
higher on the score of class B in math. The mean math score for
49 students in class A is 22.1 and the population standard
deviation is 4.8. The mean math score for 44 students of class B
is 19.8 and the population standard deviation is 5.4. At α = 0.05,
can the teacher’s claim be supported?
Solutions
► Hypothesis:
H0: μ1 – μ2 ≤0 (μ1 – μ2 =0)
H1: μ1 – μ2 >0

► At α = 0.05, critical value z=1.645

► P(z>a)=0.05=> P(z<a)=0.95 => C2: z=1.645

13
 Example

► Z=2.16
► Because z>critical value z0.05, we reject H0, thus μ1 – μ2 > 0
► There is enough evidence at the 5% level to support the teacher’s
claim that the score of class A is better on the score of class B.

14
 σ12 and σ22 Unknown,
Assumed Equal

15
 σ12 and σ22 Unknown, Assumed Equal
►If samples of size less than 30 are taken from normally-
distributed populations, a t-test (Student’s distribution)
may be used to test the difference between the
population means µ1 and µ2

►Three conditions are necessary to use a t-test for small

independent samples.
►Samples are randomly and independently drawn
► Populations are normally distributed
► Population variances/SDs are unknown but
assumed equal.

16
 σ12 and σ22 Unknown, Assumed Equal

17
 σ12 and σ22 Unknown, Assumed Equal
Two Population Means, Independent
Samples, Variances Unknown
Lower-tail test: Upper-tail test: Two-tail test:
H0: μ1 – μ2 ³ 0 H0: μ1 – μ2 ≤ 0 H0: μ1 – μ2 = 0
H1: μ1 – μ2 < 0 H1: μ1 – μ2 > 0 H1: μ1 – μ2 ≠ 0

a a a/2 a/2

-ta ta -ta/2 ta/2

Reject H0 if Reject H0 if Reject H0 if
t < -t (n1+n2 – 2), a t > t (n1+n2 – 2), a t < -t (n1+n2 – 2), a/2 or
t > t (n1+n2 – 2), a/2

|t|>!"/$
18
 Example
►You are a financial analyst for a brokerage firm. Is there a
difference in dividend yield between industries A and B?
You collect the following data:
A (1) B (2)
Num of obs (firms) 21 25
Sample mean 3.27 2.53
Sample std dev 1.30 1.16
a. Assuming both populations are approximately
normal with equal variances, is there a difference in
average yield (a = 0.05)?
b. Significance level =1%???

19
 Example
►Hypothesis:
H0: μ1 – μ2 = 0
H1: μ1 – μ2 ≠ 0

► Because sample is small and population SD is unknown, we use

Student’s distribution.
► At α = 0.05, and degree of freedom n1+n2-2=21 + 25 − 2 = 44
=> critical t-value (5%)= ± 2.015
► At α = 0.01, and degree of freedom =44
=> critical value at 1%=2.692
► t=2.040 < critical value 1% => fail to reject H0 => μ1 = μ2 => no
difference in dividends ratios between A and B.

20
 Example
►Hypothesis:
H0: μ1 – μ2 = 0
H1: μ1 – μ2 ≠ 0
► Because sample is small and population SD is unknown, we use
Student distribution. At α = 0.05, and degree of freedom n1+n2-
2=21 + 25 − 2 = 44 => critical t-value= ± 2.015
► Test statistic
Theis:
test statistic is:

t=
(X 1 )
- X 2 - (μ1 - μ2 )
=
(3.27 - 2.53 ) - 0 = 2.040
æ1 1ö æ 1 1 ö
S çç + ÷÷
2
p
1.5021ç + ÷
è n1 n2 ø è 21 25 ø

Sp2 =
(n1 - 1)S12 + (n2 - 1)S2 2 = (21 - 1)1.30 2 + (25 - 1)1.16 2 = 1.5021
(n1 - 1) + (n2 - 1) (21 - 1) + (25 - 1)

21
 Example
►Because |t|>critical value t, we reject H0 at 5%
significance level , thus μ1 – μ2 ≠ 0
►There is evidence of a difference in means.
Reject H0 Reject H0

.025 .025

-2.015 0 2.015

2.040

22
 σ12 and σ22 Unknown,
Assumed Unequal

23
 σ12 and σ22 Unknown, Assumed Unequal
►If samples of size less than 30 are taken from normally-
distributed populations, a t-test (Student’s distribution)
may be used to test the difference between the
population means µ1 and µ2

►Three conditions are necessary to use a t-test for small

independent samples.
►Samples are randomly and independently drawn
► Populations are normally distributed
► Population variances/ SDs are unknown but
assumed unequal

24
 σ12 and σ22 Unknown, Assumed Unequal
►Test statistic and degree of freedom are:

►A conservative quick rule for degrees of freedom is to

use
Minimum of n1 – 1 or n2 – 1
►Samples (n1=21, n2=25)
►df=min(21-1, 25-1)=min(20,24)=20 + significance level => t
critical value.
25
 σ12 and σ22 Unknown, Assumed Unequal

Minimum of
n1 – 1 or n2 – 1

26
 σ12 and σ22 Unknown, Assumed Unequal

27
28