06hypothesis Testing v2 PDF
06hypothesis Testing v2 PDF
Faculty of Engineering
Semester 1 (2010/2011)
Introduction
Testing
variance
g a population
p p
KNF2053
Numerical Methods and Statistics
Statistical hypothesis
Null hypothesis, H0
A claim
l i ((or statement)) about
b
l i
a population
parameter that is assumed to be true until
it is declared false
KNF2053
Numerical Methods and Statistics
Alternative hypothesis, H1
The hypothesis that we will accept if we
decide to reject the null hypothesis
Test statistic
A value computed from sample data
Rejection area (critical area)
Indicates the value of the test statistic that
will imply rejection of null hypothesis
KNF2053
Numerical Methods and Statistics
Acceptance region
The range that H0 will not be rejected if the
value of test statistic falls into it.
Critical point
The border of the rejection region
Decision Rule
Test statistics
Null hypothesis or alternative hypothesis
Acceptance region or critical region
KNF2053
Numerical Methods and Statistics
Type II error
Acceptance of the
null hypothesis when
it is false
Al
Also k
known as b
beta
error
yp II error)) =
P ((type
KNF2053
Numerical Methods and Statistics
Power
The probability of rejecting the null
hypothesis, H0 when in fact H1 is false.
Given by (1-)
Summary of error
A
t H0
Accept
Reject H0
H0 is true
C
t
Correct
decision
Type I error
H0 is false
T
Type II error
Correct decision
KNF2053
Numerical Methods and Statistics
H1:
>3o
or H1:
<3
Test
est bot
both pos
positive
t e
and negative effect
simultaneously
H1 : 3
KNF2053
Numerical Methods and Statistics
Procedure:
KNF2053
Numerical Methods and Statistics
acceptance
region
critical
region
1.5
KNF2053
Numerical Methods and Statistics
Introduction
Normal and
large sample
Small sample
Testing
variance
g a population
p p
KNF2053
Numerical Methods and Statistics
One-Tailed Test
H 0 : 0
H1 : 0 ( H1 : 0 )
Test statistic:
X 0
/ n
X 0
S/ n
C i i l region:
Critical
i
Z > Z (or
(
Z < Z
Z )
KNF2053
Numerical Methods and Statistics
KNF2053
Numerical Methods and Statistics
n 100
x RM 13.35
H 0 : RM 13.25
H1 : RM 13.25
RM 13.25
KNF2053
Numerical Methods and Statistics
rejection
region
Z > 2.3236
X 0
Compute test statistic value by using Z
/ n
2.3236
X 13.35 13.25
Z
2
/ n
0.5 / 100
KNF2053
Numerical Methods and Statistics
Z (test statistic
i i )2
acceptance
region
rejection
region
Z > 2.3236
2 2.3236
Two-Tailed Test
H 0 : 0
H1 : 0
Test statistic:
X 0
/ n
X 0
S/ n
C i i l region:
Critical
i
|Z| > Z/2
KNF2053
Numerical Methods and Statistics
KNF2053
Numerical Methods and Statistics
n 50
x 7.8kgg
if mean = 8kg
H 0 : 8kg
H1 : 8kg
8kg
KNF2053
Numerical Methods and Statistics
-2.575
2.575
KNF2053
Numerical Methods and Statistics
X
7.8 8
2.83
/ n 0.5 / 50
X 0
/ n
Decision?
Reject
j
H0 and accept
p H1
rejection region
Z< 2 575 and Z>2
Z<-2.575
Z>2.575
575
-2.575 2.575
-2.83 ( from test statistic
calculation)
KNF2053
Numerical Methods and Statistics
(n<30)
One-Tailed Test
H 0 : 0
H1 : 0 ( H1 : 0 )
Test statistic:
X
t
S/ n
Critical region:
t > t (or
(
t < tt )
X
t
S/ n
Critical region:
|t| > t/2
KNF2053
Numerical Methods and Statistics
n 10
and 0.245
S
h null
ll h
h i and
d alternative
l
i
State the
hypothesis
volume 10 litre
hypothesis
H 0 : 10 litre
volume
l
10 litre
li
H1 : 10 litre
KNF2053
Numerical Methods and Statistics
10 litre
t 2.8
region
rejection
j
g
t 2.8
2.8
0.77
/ n 0.245 / 10
X
t
/ n
KNF2053
Numerical Methods and Statistics
Decision?
t 2.8
rejection region
t 2.8
0.7
2.8
KNF2053
Numerical Methods and Statistics
n 10
and 0.245
H1 : 10 litre
lit
volume 10 litre
(reject )
10 litre
(only accept this!)
KNF2053
Numerical Methods and Statistics
rejection
j
region
g
rejection region
t -2.8
t 2.8
2.8
2.8
0.77
/ n 0.245 / 10
X
t
/ n
KNF2053
Numerical Methods and Statistics
acceptance region
2.8 t 2.8
Decision?
Accept H0 and reject H1
rejection
j
region
g
t -2.8
rejection region
t 2.8
KNF2053
Numerical Methods and Statistics
One-Tailed Test
HH0 0: :pp0 0
H 0 : p p0
H 1 : p p0
Test statistic:
p p0
z
p0 (1 p0 )
n
Critical region:
z > z (or
(
z < z )
Test statistic:
p p0
z
p0 (1 p0 )
n
Critical region:
| | > z/2
|z|
KNF2053
Numerical Methods and Statistics
p : sample proportion
KNF2053
Numerical Methods and Statistics
One-Tailed Test
H0 : 2 0
H1 : 2 0 (or H1 : 2 0 )
2
H1 : 2 0
Test statistic:
(n 1) s
Test statistic:
Critical region:
2>2 (or
( 2< 2 )
(n 1) s 2
Critical region:
| 2 | > 2 /2
2
s : sample variance
KNF2053
Numerical Methods and Statistics
KNF2053
Numerical Methods and Statistics