2010, econ 77101 hypothesis testing 1: single coefficient review of hypothesis testing testing...

37
2010, ECON 7710 1 Hypothesis Testing 1: Single Coefficient • Review of hypothesis testing • Testing single coefficient • Interval estimation Objectiv es

Upload: mariah-harper

Post on 17-Jan-2016

220 views

Category:

Documents


0 download

TRANSCRIPT

Page 1: 2010, ECON 77101 Hypothesis Testing 1: Single Coefficient Review of hypothesis testing Testing single coefficient Interval estimation Objectives

2010, ECON 7710 1

Hypothesis Testing 1: Single Coefficient

• Review of hypothesis testing

• Testing single coefficient

• Interval estimation

Objectives

Page 2: 2010, ECON 77101 Hypothesis Testing 1: Single Coefficient Review of hypothesis testing Testing single coefficient Interval estimation Objectives

2010, ECON 7710 2

s.e. (9.3421) (0.8837)R2 = 0.7431, N = 20, SER = 8.5018

ii X3771.63971.103Y

Explaining weight by height (Table 1.1)

Can X really explain Y?

When X=0, what is Y?

If we suspect that the coefficient of X is 5, can we find support from the data?

Page 3: 2010, ECON 77101 Hypothesis Testing 1: Single Coefficient Review of hypothesis testing Testing single coefficient Interval estimation Objectives

2010, ECON 7710 3

1. Hypothesis testing: Revision

The principle of hypothesis testing

The value of the parameter to be tested is assumed in H0. The estimate of this parameter is compared with that assumed value.

If the estimate is far from the assumed value, then H0 is rejected. Otherwise, H0 is not rejected.

Page 4: 2010, ECON 77101 Hypothesis Testing 1: Single Coefficient Review of hypothesis testing Testing single coefficient Interval estimation Objectives

2010, ECON 7710 4

Procedures of Hypothesis Testing

1. Determine null and alternative hypotheses.

2. Specify the test statistic and its distribution as if the null hypothesis were true.

3. Select and determine the rejection region.

4. Calculate the sample value of test statistic.

5. State your conclusions.1. Revision

Page 5: 2010, ECON 77101 Hypothesis Testing 1: Single Coefficient Review of hypothesis testing Testing single coefficient Interval estimation Objectives

2010, ECON 7710 5

2. Testing a Regression Coefficient

PopulationYi = 0 + 1X1i + 2X2i + … + KXKi + i

KiKi22i11oi XˆXˆXˆˆY Sample:

3 types of tests (k = 0, 1, 2, , K):

•Ho: k = c; HA: k c

•Ho: k c; HA: k > c

•Ho: k c; HA: k < c

c is any number meaningful in your study

Page 6: 2010, ECON 77101 Hypothesis Testing 1: Single Coefficient Review of hypothesis testing Testing single coefficient Interval estimation Objectives

2010, ECON 7710 6

Probability Distribution of Least Squares Estimators

kkkˆvar,N~

)1,0(N~ˆvar

ˆZ

k

kk

2. Testing

Page 7: 2010, ECON 77101 Hypothesis Testing 1: Single Coefficient Review of hypothesis testing Testing single coefficient Interval estimation Objectives

2010, ECON 7710 7

Student's t - statistic

t has a Student-t Distribution with N – K – 1 degrees of freedom.

)1,0(N~ˆvar

ˆZ

k

kk

1KN

k

kk t~ˆse

ˆt

2. Testing

Page 8: 2010, ECON 77101 Hypothesis Testing 1: Single Coefficient Review of hypothesis testing Testing single coefficient Interval estimation Objectives

2010, ECON 7710 8

Two-Tail t-test

1. State the null & alternative hypotheses

H0: k = c

HA: k c

2. Compute the estimated t-valuecˆ

ˆSet

k

k

3. Choose a level of significance () and degrees of freedom (N – K – 1). Then find a critical t-value from the t-table (tc = tN-K-1,/2).

2. Testing

Page 9: 2010, ECON 77101 Hypothesis Testing 1: Single Coefficient Review of hypothesis testing Testing single coefficient Interval estimation Objectives

2010, ECON 7710 9

Two-Tail t-test (cont.)

4. State the decision rule.

Version I: If |t| > tc, then reject H0.

Version II: If t > tc or t < -tc, then reject H0.

5. Conclusion Acceptance region

0 tc-tc

rejection regionrejection region

2. Testing

Page 10: 2010, ECON 77101 Hypothesis Testing 1: Single Coefficient Review of hypothesis testing Testing single coefficient Interval estimation Objectives

2010, ECON 7710 10

Example 1: In the following regression results, test whether the estimated coefficient of X1 and X2 are significantly different from zero. ( = 5%)

Y = 14.32 + 0.798 X1 – 0.101 X2

se (6.1361) (0.2535) (0.08333)R2 = 0.2718, N = 30.

Hypotheses: H0: 1 = 0; HA: 1 0.

First test:

2. Testing

Page 11: 2010, ECON 77101 Hypothesis Testing 1: Single Coefficient Review of hypothesis testing Testing single coefficient Interval estimation Objectives

2010, ECON 7710 11

Computed t-value: ....

)ˆ(

ˆ1123

25350

078900

1

1

se

t

Table t-value: For = 0.05 and 30 – 2 – 1 = 27 degrees of freedom, a critical value is t27,0.025 = 2.052.

Decision rule: If |t| > 2.052, then reject H0.

Conclusion: Since |t| = 3.112 > 2.052, Ho can be rejected. The estimated coefficient of X1 is significantly different from zero.

2. Testing

Page 12: 2010, ECON 77101 Hypothesis Testing 1: Single Coefficient Review of hypothesis testing Testing single coefficient Interval estimation Objectives

2010, ECON 7710 12

One-Tail t-test

Step 1: State the null & alternative hypotheses

Right-tail test: Test whether k > c.

H0: k c; HA: k > c.

Left-tail test: Test whether k < c.

H0: k c; HA: k < c.

2. Compute the estimated t-value (same as before)

2. Testing

Page 13: 2010, ECON 77101 Hypothesis Testing 1: Single Coefficient Review of hypothesis testing Testing single coefficient Interval estimation Objectives

2010, ECON 7710 13

3. Choose a level of significance () and degrees of freedom (N – K – 1). Then find a critical t-value from the t-table (tc = tN-K-1,).

4. State the decision rule.

Right-tail test: Reject H0 if t > tc.

Left-tail test: Reject H0 if t < -tc.

One-Tail t-test (cont.)

2. Testing

Page 14: 2010, ECON 77101 Hypothesis Testing 1: Single Coefficient Review of hypothesis testing Testing single coefficient Interval estimation Objectives

2010, ECON 7710 14

0 tc < t

Right-tail

0-tct <

left-tail

2. Testing

One-Tail t-test (cont.)

5. Conclusion

Page 15: 2010, ECON 77101 Hypothesis Testing 1: Single Coefficient Review of hypothesis testing Testing single coefficient Interval estimation Objectives

2010, ECON 7710 15

Example 3: Right-tail test: Test whether 1 is greater than 0.35 at 5% level of significance.

2. Testing

Dependent Variable: YMethod: Least SquaresSample: 1 30Included observations: 30

Coefficient Std. Error t-Statistic Prob.

C 16.42782 5.936234 2.767381 0.0099X 0.7177 0.246886 2.907007 0.0071

R-squared 0.231839 Mean dependent var 32.6Adjusted R-squared 0.204405 S.D. dependent var 12.71871S.E. of regression 11.3446 Akaike info criterion 7.759701Sum squared resid 3603.597 Schwarz criterion 7.853114Log likelihood -114.3955 Hannan-Quinn criter. 7.789585F-statistic 8.450689 Durbin-Watson stat 1.338091Prob(F-statistic) 0.007061

Page 16: 2010, ECON 77101 Hypothesis Testing 1: Single Coefficient Review of hypothesis testing Testing single coefficient Interval estimation Objectives

2010, ECON 7710 16

Example 4: Left-tail test: Test whether 1 in Example 3 is smaller than 1.2. ( = 0.05)

2. Testing

Page 17: 2010, ECON 77101 Hypothesis Testing 1: Single Coefficient Review of hypothesis testing Testing single coefficient Interval estimation Objectives

2010, ECON 7710 17

A Special caseHo: k = 0

HA: k 0

Statistic

k

k

ˆse

ˆt

2. Testing

It is the lowest level of significance at which we could reject the Ho that a parameter is zero.

The p-values Reported by Regression Software

Page 18: 2010, ECON 77101 Hypothesis Testing 1: Single Coefficient Review of hypothesis testing Testing single coefficient Interval estimation Objectives

2010, ECON 7710 18

The t-statistics and P-values

Dependent Variable: YMethod: Least SquaresSample: 1 30Included observations: 30

Coefficient Std. Error t-Statistic Prob.

C 16.42782 5.936234 2.767381 0.0099X 0.7177 0.246886 2.907007 0.0071

R-squared 0.231839 Mean dependent var 32.6Adjusted R-squared 0.204405 S.D. dependent var 12.71871S.E. of regression 11.3446 Akaike info criterion 7.759701Sum squared resid 3603.597 Schwarz criterion 7.853114Log likelihood -114.3955 Hannan-Quinn criter. 7.789585F-statistic 8.450689 Durbin-Watson stat 1.338091Prob(F-statistic) 0.007061

2. Testing

Page 19: 2010, ECON 77101 Hypothesis Testing 1: Single Coefficient Review of hypothesis testing Testing single coefficient Interval estimation Objectives

2010, ECON 7710 19

The p-value of 1-hat for a two-sided test

t0

f(t)

-2.91 2.91

p/2 = 0.00355

red area = p-value = 0.0071

p/2 = 0.00355

2.048-2.048

critical values

2. Testing

Page 20: 2010, ECON 77101 Hypothesis Testing 1: Single Coefficient Review of hypothesis testing Testing single coefficient Interval estimation Objectives

2010, ECON 7710 20

3. Confidence Intervals for Regression Coefficients

Yi = 0 + 1Xi + ui (i = 1,n)

The OLS estimators for 0 and 1 are point estimators.The OLS estimates are likely to be different from the theoretical values

We have no idea of how close the OLS estimates to the theoretical values

Page 21: 2010, ECON 77101 Hypothesis Testing 1: Single Coefficient Review of hypothesis testing Testing single coefficient Interval estimation Objectives

2010, ECON 7710 21

Interval estimation:

.1 0 where, - 1 ) ˆ -ˆP(

such that 0 a findcan then we

known, is ˆ ofon distributi theIf estimator. its

be ˆ andparameter population a be Let

k

k

kk

k

k

We know the chance of including the population parameter (k )in the intervals constructed from repeated samples.

3. Confidence Interval

Page 22: 2010, ECON 77101 Hypothesis Testing 1: Single Coefficient Review of hypothesis testing Testing single coefficient Interval estimation Objectives

2010, ECON 7710 22

) ˆ , - ˆ( :estimator Interval kk

Confidence coefficient : 1 -

Level of significance :

Interval estimate : (k* - , k* + )

Population parameter: k

Estimator of k :

Estimate of k : k*

Confidence limits

3. Confidence Interval

Page 23: 2010, ECON 77101 Hypothesis Testing 1: Single Coefficient Review of hypothesis testing Testing single coefficient Interval estimation Objectives

2010, ECON 7710 23

Constructing Confidence Interval for k

Actual estimated k could be fallen intothese regions

fk

k E k k

3. Confidence Interval

Page 24: 2010, ECON 77101 Hypothesis Testing 1: Single Coefficient Review of hypothesis testing Testing single coefficient Interval estimation Objectives

2010, ECON 7710 24

fk

k E k k

a

b( ) )(

interval a interval b

Constructing Confidence Interval for k

3. Confidence Interval

Page 25: 2010, ECON 77101 Hypothesis Testing 1: Single Coefficient Review of hypothesis testing Testing single coefficient Interval estimation Objectives

2010, ECON 7710 25

( )ta )(

tb

interval ta interval tb

f(t)

ˆSe

ˆt

k

kk

0

Constructing Confidence Interval for k

3. Confidence Interval

Page 26: 2010, ECON 77101 Hypothesis Testing 1: Single Coefficient Review of hypothesis testing Testing single coefficient Interval estimation Objectives

2010, ECON 7710 26

Probability statements

P(-tc < t < tc ) = 1

P( t < -tc ) = P( t > tc ) =

1)ˆ(ˆ)ˆ(ˆ

1)ˆ(

kckkkck

c

k

kkc

setsetP

tse

tP

3. Confidence Interval

Page 27: 2010, ECON 77101 Hypothesis Testing 1: Single Coefficient Review of hypothesis testing Testing single coefficient Interval estimation Objectives

2010, ECON 7710 27

A 95% confidence interval means that, using the interval estimator and drawing samples from the population, 95% of the interval estimates would include the population value .

The probability that a particular interval estimate contains this population value is either 0 or 1.

3. Confidence Interval

The (1-)100% CI for k is

)ˆ(setˆ),ˆ(setˆk/2 ,1KNkk/2 ,1KNk

Page 28: 2010, ECON 77101 Hypothesis Testing 1: Single Coefficient Review of hypothesis testing Testing single coefficient Interval estimation Objectives

2010, ECON 7710 28

Example 5: Regressing WEIGHT on HEIGHT, 2005

3. Confidence Interval

Page 29: 2010, ECON 77101 Hypothesis Testing 1: Single Coefficient Review of hypothesis testing Testing single coefficient Interval estimation Objectives

2010, ECON 7710 29

95% confidence limits for 1:

95% confidence interval for 0:

..,.

...ˆˆ.,

7811037650

10110000325788010250681

set

3. Confidence Interval

The 95% confidence interval for 1 is (0.3765, 0.7811).

Page 30: 2010, ECON 77101 Hypothesis Testing 1: Single Coefficient Review of hypothesis testing Testing single coefficient Interval estimation Objectives

2010, ECON 7710 30

4. Applied Examples

Example 6: Restaurant location (Section 3.2)

Suppose you have been hired to determine a location for the next Woody’s. Woody’s is a moderately priced, 24-hour, family restaurant chain. Two choices are:

Location A: NN = 4.4, PP = 104, II = 20.6

Location B: NN = 2, PP = 50, II = 20

Page 31: 2010, ECON 77101 Hypothesis Testing 1: Single Coefficient Review of hypothesis testing Testing single coefficient Interval estimation Objectives

2010, ECON 7710 31

YYi = 0 + NNi + PPPi + IIIi + i. -ve +ve ?

YY: Number of customers served in thousand

N: Number of direct market competitors

PP: Population in thousand within a 3-mile radius

II: Average household income in thousand

Example 6: Restaurant location (Cont’d)

4. Examples

Page 32: 2010, ECON 77101 Hypothesis Testing 1: Single Coefficient Review of hypothesis testing Testing single coefficient Interval estimation Objectives

2010, ECON 7710 32

Woody’s: Null and Alternative Hypotheses

1. Ho: N 0; HA: N < 0

2. Ho: P 0; HA: P > 0

3. Ho: I = 0; HA: I 0

YY = 102.19*** – 9.07***N + 0.35***PP + 1.29**II se (2.0527) (0.07268) (0.5433)R2 = 0.618, R2 = 0.579, N = 33.

^

_

4. Examples

Page 33: 2010, ECON 77101 Hypothesis Testing 1: Single Coefficient Review of hypothesis testing Testing single coefficient Interval estimation Objectives

2010, ECON 7710 33

If i is normally distributed, then

• hat is normally distributed with mean k and variance var( ).

•Z = has standard normal distribution.

kk

k

kk

ˆvar

ˆ

k kˆrav •Var( ) is unobservable. is used inst

ead and is denoted by se.

•t = has t distribution with N – K – 1

degrees of freedom. k

kk

ˆse

ˆ

4. Examples

Page 34: 2010, ECON 77101 Hypothesis Testing 1: Single Coefficient Review of hypothesis testing Testing single coefficient Interval estimation Objectives

2010, ECON 7710 34

Woody’s two-sided test

Hypotheses: Ho: I = 0; HA: I 0

Statistics:

.freedom of degrees 294-33 with

ondistributi t a has ˆse

0ˆt

N

N

Decision rule: Let = 0.05. From the table the critical values are tc = t29,0.025 = 2.045. Reject Ho if |t| > 2.045.

.37.25433.0

0288.1ˆse

0ˆt

I

I

Computed t-value:Decision: Since t = 2.37 > 2.045, reject Ho. Thus I hat is significantly different from zero.

Page 35: 2010, ECON 77101 Hypothesis Testing 1: Single Coefficient Review of hypothesis testing Testing single coefficient Interval estimation Objectives

2010, ECON 7710 35

Woody’s one-sided tests

Hypotheses: Ho: N 0; HA: N < 0

Decision rule: Let = 0.05. From the table the critical value is tc = -t29,0.05 = -1.699. Reject Ho if t < -1.699.

.42.4053,2

0075,9ˆse

0ˆt

N

N

Computed t-

value:

Decision: Since t = - 4.42 < -1.699, reject Ho. Thus N hat is significantly smaller than zero.

4. Examples

Page 36: 2010, ECON 77101 Hypothesis Testing 1: Single Coefficient Review of hypothesis testing Testing single coefficient Interval estimation Objectives

2010, ECON 7710 36

Example 7: Sales of Hamburger

TRi = 0 + pPi + AAi + i. ? +

Data: Weekly observations for a hypothetical hamburger chain

TR : Weekly revenue in $1,000

P : Price in $

A : Advertising expenditure in $1,000

4. Examples

Page 37: 2010, ECON 77101 Hypothesis Testing 1: Single Coefficient Review of hypothesis testing Testing single coefficient Interval estimation Objectives

2010, ECON 7710 37

TR = 113.83*** – 10.26***P + 2.68***A se (1.6007) (0.1189)R2 = 0.8739, N = 78.

^Regression results:

a. Is the demand significantly elastic or inelastic in price?

b. Is the increase in total revenue stimulated by more advertisements significantly greater than the corresponding increased cost of advertising?

Let = 0.01. Answer the following two questions statistically.