APPLIED ECONOMETRICS COURSE
Chapter III
Statistic Inference and
Hypothesis Testing
NGUYEN BA TRUNG 2016
I. THE DISTRIBUTION OF THE PARAMETERS
Assumpt 6. Ui ~ N(0,
2)
Theorem 4.1: Normal Distribution of the
parametors
ˆ −β
β
0
βˆ0 ~ N ( β 0 , σ ) � Z = 0
~ N (0,1)
σ βˆ
2
βˆ0
0
ˆ −β
β
βˆ1 ~ N ( β1 , σ ) � Z = 1 1 ~ N (0,1)
σ βˆ
2
βˆ1
1
NGUYEN BA TRUNG 2016
I. THE DISTRIBUTION OF THE PARAMETERS
•
Under the assumptions from 1-6, we have the
following theorem:
Theorem 4.2: Student Distribution
βˆ j − β j
t=
~ t(n − k − 1)
se( βˆ )
j
NGUYEN BA TRUNG 2016
II. HYPOTHESES TESTING
2.1. Testing against OneSided Alternatives:
Right Hand Side
H 0 : β1 = 0
(1). Manifest the hypothesis:
H1 : β1 > 0
βˆ1 − β1
(2). Compute tstatistics: t = se( βˆ )
1
(3). Search to find t (nk1), for example, α=5%, nk1=28, t0.05
(28)=1.701 ( excel: TINV(0.1,28))
(4). Decision Rule:
Nếu t > t (nk1) thì có thể bác bo gia thiê
̉ ̉
́t H0
Nếu t < t (nk1) thì có thể chấp nhân gia thiê
̣
̉
́t H0
NGUYEN BA TRUNG 2016
II. HYPOTHESES TESTING
NGUYEN BA TRUNG 2016
Example: expenditure.wf
(1). Let’s test the following hypothesis with α = 5%:
H 0 : β1 = 0
H1 : β1 > 0
NGUYEN BA TRUNG 2016
II. HYPOTHESES TESTING
βˆ1 − β1 0.509
(2). Compute sstatistics: t = se( βˆ ) = 0.035 = 14.54
1
(3). Search to find: t0.05 (8) = 1.86
(4). Decision Rule: tstatistics = 14.54 > t0.05 (8)=1.86, we therefore
can reject the null hypothesis H0. In other word, we can accept the
alternative hypothesis H1: 1 > 0
NGUYEN BA TRUNG 2016
II. HYPOTHESES TESTING
2.2. Testing against One-Sided Alternatives: Left Hand
(1).
Side Suppose the hypothesis:
H 0 : β1 = 0
H1 : β1 < 0
(2).Compute tstatistics:
βˆ1 − β1
t=
se( βˆ1 )
(3).Search to find t (nk1)
(4). Decision Rule :
Nếu t < t (nk1) thì có thể bác bo gia thiê
̉ ̉
́t H0
NGUYEN BA TRUNG 2016
II. HYPOTHESES TESTING
2.3.
(1). SetTwo-Sided
up the hypothesis:
H :β
Alternatives H : β
0
1
=0
1
1
0
βˆ1 − 0
(2). Compute tstatistics: t = se( βˆ )
1
(3). Search to find t /2 (nk1), for example α = 5% và nk1=8 thì
t0.025 (8)=2.306 (TINV(0.05,8))
(4). Decision Rule:
if t > t /2(nk1), we can reject the null hypothesis H0
if t t /2(nk1), we can accept the null hypothesis H0
NGUYEN BA TRUNG 2016
Example: Whether income affects to expenditure?
Solution:
(1)
Manifest the hypothesis:
H 0 : β1 = 0
H1 : β1
(2)
Compute s-statistics:
βˆ − β
0.5091 − 0
t=
=
= 14.243
0.035742
se( βˆ1 )
1
(3)
0
With
1
= 5%, we have: t0.025(n-k-1) = t0.025(8) =
2.306
(4)
Because t =14.243 > t0.025(8) = 2.36, we can reject
the null hypothesis H0: β1= 0. This means that
income significantly
impacts to expenditure.
NGUYEN BA TRUNG 2016
Example: expenditure.wf
NGUYEN BA TRUNG 2016
2.4. Testing Other Hypotheses
about that
Suppose
parameter
there is a research showing that β1= 0.3,
Whether this conclusion is true or not?
Solution:
H 0 : β1 = 0.3
(1)
(2)
H1 : β1
Set up the null hypothesis:
0.3
βˆ1 − β1
t=
Compute s-statistics:se( βˆ1 )
(3)
Search to find t /2(n-k-1)
(4)
Decision:
If t > t /2(n-k-1) => Reject H0
If NGUYEN BA TRUNG 2016
t
t /2(n-k-1) => Accept H0
H 0 : β1 = 0.3
(1). Null hypothesis:
•
H1 : β1
0.3
βˆ1 − β1 0.5091 − 0.3
t-statistics: t =
=
= 5.85
ˆ
0.035742
se( β1 )
•
Search to find: t0.025(n-k-1)=t0.025(8) = 2.306
•
Because t =5.85 > t /2(8) = 2.36, we can reject H0:
β1= 0.3 with
= 5%.
NGUYEN BA TRUNG 2016
2.5. Compute P-value for t-tests
•
•
•
•
With confident level α, P-value is the probability of
mistake if we reject the null hypothesis H0
P-value: Prob(|T|>|t|)
Thus, if small P-value provides the evidence against
H0, and if large P-value provide the evidence against
H1
Normally, we often choose confident level at 1%, 5%,
and 10%.
NGUYEN BA TRUNG 2016
Ví dụ: expenditure.wf
NGUYEN BA TRUNG 2016
III. Confident Interval
Confident Interval of the parameters
βˆ0
βˆ1
1; 2
tα / 2 se( βˆ0 )
t se( βˆ )
α /2
1
Confident Interval of error term Ui
σˆ (n − 2)
σˆ (n − 2)
2
<σ < 2
2
χα /2 (n − 2)
χ1−α /2 (n − 2)
2
2
NGUYEN BA TRUNG 2016
Example: expenditure (continue)
Let’s compute the confident intervals for the parameters at confident
level 95%?
Ø
Because 1 =0.95 ➨ = 0.05 ➨ /2 = 0.025
Ø
t 0.025(8) =2.306 [TINV(0.05,8)]
ˆ
t
0
ˆ ) = 24.4545 ± 2.306 * 6.4138
se
(
/2
0
hay (9.6643 < 0 < 39.2448)
ˆ
1
t
ˆ)
se
(
/2
1
= 0.5091 ± 2.306 * 0.035742
hay (0.4268 < 1 < 0.5914)
Thí du: expenditure (continue)
̣
We can utilize confident interval in testing hypothesis,
for instance
•
H 0 : β1 = 0.3
H1 : β1
v
0.3
The confident interval for 1:
(0.4268 < 2 < 0.5914)
v
Because 1 = 0.3 is out of the confident interval, we can reject the H0
NGUYEN BA TRUNG 2016
IV. F- test
(1). Suppose that X1…X2…Xk do not effect on Y, this
means that:
H 0 : β1 = β 2 = ... = β k = 0 � R 2 = 0
(2). Compute Fstatistics as below:
ESS
(k )
2
ESS / (k )
R
/k
TSS
F (k , n − k − 1) =
=
=
RSS / (n − k − 1) RSS (n − k − 1) (1 − R 2 ) / (n − k − 1)
TSS
(3). Search to find: F(k,nk1,α)
(4). Decision Rule:
v
If F> F(k,nk1,α), we can reject H0.
v
If F ≤ F(k,nk1,α), we can not reject H0.
NGUYEN BA TRUNG 2016
Example: continuted
•
Compute F-statistics:
F=
v
0.96206(10 − 8)
= 202.86
1 − 0.96206
With α = 5%, search to find Fα (k, nk1), we have: F0.05 (1,8)= 5.32
v
Because F = 202.86 > F0.05 (1,8)= 5.32, we can reject H0
v
EVIEWs provide the pvalue of Ftest(next slide)
NGUYEN BA TRUNG 2016
Thí dụ: expenditure.wf
NGUYEN BA TRUNG 2016
V. Test the simple linear
combination
the
parameters
(1). Assume that we of
want
to
H test:
:β
=β
0
(2). Compute tstatistics as:
v
Where:
Se( βˆ1 − βˆ2 ) =
1
2
H1 : β1
β2
βˆ1 − βˆ2
t=
se( βˆ1 − βˆ2 )
( se(βˆ )
1
2
(
− 2 cov( βˆ1 , βˆ2 ) + se( βˆ2
(3). With α = 5%, search to find: tα/2(nk1)
(4). Decision: If |t| > tα/2(nk1), We can reject H0
NGUYEN BA TRUNG 2016
)
2
Example:Twoyears.wf
•
We have the following model:
log(wage) = β 0 + β1 jc + β 2univ + β 3 exp er + u
Where: Jc: trình độ cao đẳng (2 năm đại học)
v
Unive: trình độ đại học (4 năm đại học)
v
Exper: kinh nghiệm (tháng)
Question: whether wage differs between who those graduated from
colleague and university?
NGUYEN BA TRUNG 2016
v
From the estimate result, the different wage:
θ1 = βˆ1 − βˆ2 = −0.0102
v
Is this difference statistically significant?
NGUYEN BA TRUNG 2016