dignostic tests of applied economics

19

Upload: suniya-sheikh

Post on 25-May-2015

83 views

Category:

Data & Analytics


3 download

DESCRIPTION

1.Normality test,Stationary test,3.Multicorlinearity

TRANSCRIPT

Page 1: Dignostic  Tests of Applied Economics
Page 2: Dignostic  Tests of Applied Economics

DIAGNOSTIC TESTS

Page 3: Dignostic  Tests of Applied Economics

Types of Diagnostic Test

Page 4: Dignostic  Tests of Applied Economics

Normality

Normality is a condition in which the

used variables follow the standard

normal distribution.

Page 5: Dignostic  Tests of Applied Economics

Normality Test

Histogram of residuals

Normal probability plot (NPP)

The Jarque – Bera test

Page 6: Dignostic  Tests of Applied Economics

The Jarque Bera Test

The JB test of normality is a large-sample

test. It is also based on the OLS residuals.

We make hypothesis in this test.

Page 7: Dignostic  Tests of Applied Economics

How study Result of JB

If the computed value of P > 0.05, then we

said that residuals are normally distributed.

If the computed value of P < 0.05, then we

said that residuals are not normally

distributed.

Page 8: Dignostic  Tests of Applied Economics

Example

 Variable ( LINV )

H0 : Residuals are normally distributed

H1 : Residuals are not normally distributed

Critical Region: 0.05

Results: p = 0.99

0.99 > 0.05

Conclusion : Ho Accepted

Page 9: Dignostic  Tests of Applied Economics

0

2

4

6

8

10

0.7 0.8 0.9 1.0 1.1 1.2

Series: LINVSample 1979 2010Observations 32

Mean 0.969416Median 0.968722Maximum 1.196225Minimum 0.699115Std. Dev. 0.115272Skewness 0.013718Kurtosis 3.081233

Jarque-Bera 0.009802Probability 0.995111

Page 10: Dignostic  Tests of Applied Economics

Stationarity

Time series Yt is said to be stationary,

if its mean, its variance and its

covariance remain constant over time.

Page 11: Dignostic  Tests of Applied Economics

Stationarity Tests

Graphical Analysis

Autocorrelation Function (ACF)

Differencing

The Unit Root Test

a) Dickey Fuller Test (DF)

b) Augmented Dickey Fuller Test (ADF)

c) Phillips Perron Test

Page 12: Dignostic  Tests of Applied Economics

Graphical Analysis

Page 13: Dignostic  Tests of Applied Economics

Differencing

If we wants to remove the trend

component from a (time) series

entirely to render it stationary.

we need to apply differencing, i.e.

compute absolute changes from one

period to the next.

Page 14: Dignostic  Tests of Applied Economics

Symbolically

First order Differencing ∆Yt =Yt – Yt -1 If differenced series still exhibits a trend, it needs to be

differenced again (one or more times) to render it stationary. Thus we have

Second order Differencing

Page 15: Dignostic  Tests of Applied Economics

Multicolinearity

Multicolinearity is the undesirable situation

where the correlations among the

independent variables are strong. We have

perfect muticolinearity if , the value of

correlation of two independent variables is

between +1 to -1

Page 16: Dignostic  Tests of Applied Economics

Example

Page 17: Dignostic  Tests of Applied Economics

Interpretation of Result

1.0 Perfect Multicolinearity

above 0.5 Strong Multicolinearity

below 0.5 No Multicolinearity

Page 18: Dignostic  Tests of Applied Economics

REFERENCES

Applied Econometrics

by Asteriou and Stephen

Basic Econometrics

by Damodar N. Gujarati

Normality Tests for Statistical Analysis:

A Guide for Non-Statisticians

by Ghasemi , Zahediasl

Page 19: Dignostic  Tests of Applied Economics