Upload
gapusing
View
235
Download
0
Embed Size (px)
Citation preview
7/27/2019 Tutorial Single Equation Regression Model
1/32
Tutorial 1
Econometrics
Two-Variable
Regression Analysis
MEP-DD Muhammad Edhie Purnawan, MA, Ph.D
Prepared by Rafiazka Millanida H.
7/27/2019 Tutorial Single Equation Regression Model
2/32
7/27/2019 Tutorial Single Equation Regression Model
3/32
Reviewing Methodology of Econometrics
Classical econometric methodology proceeds along thefollowing lines:
1. Statement of theory or hypothesis.
2. Specification of the mathematical model of the theory
3. Specification of the statistical, or econometric, model
4. Obtaining the data
5. Estimation of the parameters of the econometric model
6. Hypothesis testing
7. Forecasting or prediction
8. Using the model for control or policy purposes.
Overview Introduction to EVIEWS
Application Exercise
7/27/2019 Tutorial Single Equation Regression Model
4/32
Differing Economic Data
Cross sectionaldata
Consists of a
sample ofvariety of unitstaken at a givenpoint in time
Ex: census of
populatioconducted bythe CensusBureau every 10
years
Time series data
Consists of
observations ona variable orseveralvariables overtime.
Ex: stock prices
Panel Data
Data are
elements ofboth time seriesand cross-section data.
Ex: Consumer
Price Index (CPI)for severalcountries for19731997
Overview Introduction to EVIEWS
Application Exercise
7/27/2019 Tutorial Single Equation Regression Model
5/32
Interpreting Regression
Regression analysis is concerned with the study
of the dependence of one variable, the
dependent variable, on one or more other
variables, the explanatory variables, with aview to estimating and/or predicting the
(population) mean or average value of the
former in terms of the known or fixed (inrepeated sampling) values of the latter.
Overview Introduction to EVIEWS
Application Exercise
7/27/2019 Tutorial Single Equation Regression Model
6/32
Contd
Case:
The rate of change of money wages in relation to theunemployment rate. The historical data is an exampleof the celebrated Phillips curve relating changes in the
money wages to the unemployment rate.
Why scattergram?- It may enable the labor economist to
predict the average change in money
wages given a certain unemploymentrate.
- It may be helpful in stating something
about the inflationary
process in an economy, for increases in
money wages are likely to be
reflected in increased prices.
Overview Introduction to EVIEWS
Application Exercise
7/27/2019 Tutorial Single Equation Regression Model
7/32
Determining Regression Statistical Versus Deterministic Relationships
- statistical not functional or deterministic
- random or stochastic (have probability distribution)
- eg: The dependence of crop yield on temperature,rainfall, sunshine, and fertilizer, for example, isstatistical in nature in the sense that the explanatoryvariables, although certainly important, will not enablethe agronomist to predict crop yield exactly because of
errors involved in measuring these variables as well asa host of other factors (variables) that collectivelyaffect the yield but may be difficult to identifyindividually.
Overview Introduction to EVIEWS
Application Exercise
7/27/2019 Tutorial Single Equation Regression Model
8/32
Contd Regression Versus Causation
-A statistical relationship can never establish causalconnection (Kendall and Stuart).
- To ascribe causality, one must appeal to a priori or
theoretical considerations.- eg: There is no statistical reason to assume thatrainfall does not depend on crop yield. The fact that wetreat crop yield as dependent on rainfall (among other
things) is due to non-statistical considerations:Common sense suggests that the relationship cannotbe reversed, for we cannot control rainfall by varyingcrop yield.
Overview Introduction to EVIEWS
Application Exercise
7/27/2019 Tutorial Single Equation Regression Model
9/32
Contd
Regression Versus Correlation
Correlation analysis aims to measure the strength
or degree of linear association between two
variables.
Regression analysis try to estimate or predict the
average value of one variable on the basis of the
fixed values of other variables.
Overview Introduction to EVIEWS
Application Exercise
7/27/2019 Tutorial Single Equation Regression Model
10/32
Selecting Methods of Estimation The three generally used methods of estimation:
1. ordinary least squares (OLS)
2. maximum likelihood (ML)
3. method of moment (MM)
By and large, it is the method of OLS that is usedextensively in regression analysis primarily
because it is intuitively appealing andmathematically much simpler than the ML andMM
Overview Introduction to EVIEWS
Application Exercise
7/27/2019 Tutorial Single Equation Regression Model
11/32
EVIEWS Window
Main Menu
Command Window
Work Area
Status Line
Title Bar
Overview Introduction to EVIEWS
Application Exercise
7/27/2019 Tutorial Single Equation Regression Model
12/32
Workfile1. Creating Workfile
File NewWorkfile
Workfile Structure Type:
- Unstructured/Undated- Dated/Regular Frequency
Annual : 1991 2011Semi-annual : 1995:01 1995:02Quarterly : 1997:03 2007:04Monthly : 2003:05 2011:12Daily : 2005:31 2011:250
- Balanced Panel
2. Saving WorkfileFile Save As ....
3. Opening WorkfileFile Open Workfile
Overview Introduction to EVIEWS
Application Exercise
7/27/2019 Tutorial Single Equation Regression Model
13/32
Data
1. Entering Data:Manual
a. By Command- Make series/variable
series series_nameshow series_name
edit +/-- Enter DataType or paste data from microsoftexcel
b. By option- Make series/variable
ObjectNew ObjectType ofObject: Series Name for Object:series_name
orQuickEmpty Group Series
- Enter dataDouble click to the series/variableedit +/-Type or paste data from microsoftexcel
2. Entering Data:Import Option
a. From Microsoft Excel
File ImportReadFile Type:
ExcelFile Name: file_nameNote: Variables in Excel start at B2.
Excel must be closed before
importing process.
b. From Text
File ImportReadFile Type:
Text-ASCII File Name: file_name
Overview Introduction to EVIEWS
Application Exercise
7/27/2019 Tutorial Single Equation Regression Model
14/32
Contd
3. Opening Data Group
a. By Command
show series1_name series2_name
b. By optionblock series right clickopenas group
or
quickshowobjects:
series1_name series2_name
4. Plotting Data
a. By Commandsingle graph:plot series1_nameseries2_name
multiple graphs:plot(m)series1_name series2_nameb. By optionsingle graph: block series
right clickopen as groupview graphmultiple
series: single graphmultiple graphs: block series right clickopen as groupview graphmultipleseries: multiple graph
Overview Introduction to EVIEWS
Application Exercise
O i I d i EVIEWS
7/27/2019 Tutorial Single Equation Regression Model
15/32
Descriptive Statistics
1. Statistical Value
a. By Command
stats series1_name series2_name
b. By optionblock series right clickopenas group
viewdescriptive stats individual sample
2. Correlation Matrix
a. By Command
cor series1_name series2_name
b. By optionblock series right clickopen
as group
viewcovariance analysis
statistics: correlation
Overview Introduction to EVIEWS
Application Exercise
O i I t d ti t EVIEWS
7/27/2019 Tutorial Single Equation Regression Model
16/32
Equationquick estimates equation specification: series1_name c
series 2_name
Overview Introduction to EVIEWS
Application Exercise
block series1_name c series 2_name right clickopen asequation proc series1_name c series 2_name
ls series1_name c series 2_nameObject name name: eq01
equation eq01.ls series1_name c series 2_name
or
or
or
O i I t d ti t EVIEWS
7/27/2019 Tutorial Single Equation Regression Model
17/32
Step of Regression
1.Hypothesis Testing
If null hypothesis isnt rejected hence independent
variable isnt significant to dependent variable
2.Interpreting Regression Coefficient3.R (R-squared)
4.Residual, Fitted, and Actual Value
5.Test for normality of Residual
Overview Introduction to EVIEWS
Application Exercise
O i I t d ti t EVIEWS
7/27/2019 Tutorial Single Equation Regression Model
18/32
1. Open EVIEWSProgram
2. Open NewWorkfile
3. Quick EmptyGroup (editseries), copy the
data from excel.
Running EVIEWS
Overview Introduction to EVIEWS
Application Exercise
Exchange Rate
Gujarati (2003), p.33
O i I t d ti t EVIEWS
7/27/2019 Tutorial Single Equation Regression Model
19/32
equation eq1.ls france c germany
Compare to value of t-stat table2
Overview Introduction to EVIEWS
Application Exercise
Compare to given =5%
O er ie Introd ction to EVIEWS
7/27/2019 Tutorial Single Equation Regression Model
20/32
Overview Introduction to EVIEWS
Application Exercise
Hypothesis Testing
The model:France = o + 1_Germany + u
Ho: 1 = 0
Ha: 1 0
1
Prob. < 0.05: Ho (Null Hypothesis) is notrejected independent variable is
significant to dependent variable.
Prob. > 0.05: Ho (Null Hypothesis) isrejected independent variable is notsignificant to dependent variable.
Prob.
0.0001 < 0.05Ho (Null Hypothesis) is not
rejected.
Independent variable is
significant to dependent
variable.
Overview Introduction to EVIEWS
7/27/2019 Tutorial Single Equation Regression Model
21/32
Overview Introduction to EVIEWS
Application Exercise
t-stat
4,997 > 2.845 or 2.086 or
1.725
4,997 > 2
Ho (Null Hypothesis) is not
rejected.
Independent variable is
significant to dependentvariable.
t-stat > t-table: Ho (Null Hypothesis) is notrejectedindependent variable is
significant to dependent variable.
t-stat < t-table: Ho (Null Hypothesis) isrejectedindependent variable is notsignificant to dependent variable.
t-table
n = 22k = 2df = n-k = 20
= 10% t-table = 1.725= 5% t-table = 2.086
= 1% t-table = 2.845 Note:Since Eviews applies level of significanceis 5% or = 5%, the rule of thumbregarding to t-stats will be for anyregressions with more than n>20 and 0.05level of significant, the critical value is 2.
t-stats > 2: Ho (Null Hypothesis) is
rejected. Independent variable is significant
to dependent variable.
t-stat < 2: Ho (Null Hypothesis) is not
rejected independent variable is not
significant to the dependent variable.
Overview Introduction to EVIEWS
7/27/2019 Tutorial Single Equation Regression Model
22/32
t-table
Degree of freedom:
n-k
One tail
Two tail
n = number of data
k = number of parameter
Or number of variable
Overview Introduction to EVIEWS
Application Exercise
Overview Introduction to EVIEWS
7/27/2019 Tutorial Single Equation Regression Model
23/32
Overview Introduction to EVIEWS
Application Exercise
Interpreting Regression Coefficient
France = 1.618887+ 2.185832Germany
2
If exchange rate of Germany increases by an
average of 1 USD, exchange rate of france
will increase by average of 1.618887 USD,
ceteris paribus.
Overview Introduction to EVIEWS
7/27/2019 Tutorial Single Equation Regression Model
24/32
Overview Introduction to EVIEWS
Application Exercise
R (R-squared) 3
The variance of independent variable (exchange
rate of Germany) can explain 55.53% variance of
the dependent variable (exchange rate of France).
Overview Introduction to EVIEWS
7/27/2019 Tutorial Single Equation Regression Model
25/32
Overview Introduction to EVIEWS
Application Exercise
Residual, Fitted, and Actual Value 4
-2.0
-1.5
-1.0
-0.5
0.0
0.5
1.0
4
5
6
7
8
9
10
78 80 82 84 86 88 90 92 94 96 98
Residual Actual Fitted
Viewactual, fitted, residual
actual, fitted, residual graph
Graph shows the goodness
of fit of the model. Actual
value is close to estimatedvalue or fitted value. The
more closer the two are,
residual is around zero.
Overview Introduction to EVIEWS
7/27/2019 Tutorial Single Equation Regression Model
26/32
Overview Introduction to EVIEWS
Application Exercise
Normality Test 5
JB Test is designed to track departures
from the normal distribution of the OLS
residuals.
viewresidual diagnostics
histogram-normality test
Ho: Disturbance terms are normallydistributed.
Ha: Ho is not true.
Overview Introduction to EVIEWS
7/27/2019 Tutorial Single Equation Regression Model
27/32
Overview Introduction to EVIEWS
Application Exercise
Contd
Prob. > 0.05: Ho (Null Hypothesis) is notrejected Disturbance terms are normallydistributed.
Prob. < 0.05: Ho (Null Hypothesis) isrejected Disturbance terms are notnormally distributed.
JB-stat < c2 table , the model is normallydistributed.
JB-stat > c2 table , the model is notnormally distributed.
JB-stat
4.363461 < 10.8508 ( =
5%; df=22-2=20)
Ho (Null Hypothesis) is not
rejected.
Disturbance terms arenormally distributed.
Prob
0.112846 > 0.05 ( = 5%)
Ho (Null Hypothesis) is notrejected.
Disturbance terms are
normally distributed.
Overview Introduction to EVIEWS
7/27/2019 Tutorial Single Equation Regression Model
28/32
c2 -tableDegree of freedom:
n-k
n = number of data
k = number of parameter
Or number of variable
Overview Introduction to EVIEWS
Application Exercise
Overview Introduction to EVIEWS
7/27/2019 Tutorial Single Equation Regression Model
29/32
Overview Introduction to EVIEWS
Application Exercise
Regress data sales.xls using eviews
as the previous.
Overview Introduction to EVIEWS
7/27/2019 Tutorial Single Equation Regression Model
30/32
Overview Introduction to EVIEWS
Application Exercise
Assignment 1
Journal draft on two-variable regression
analysis
(Abstract, Introduction, Theoretical
Framework, Data&Methodology, Model,
Result, Conclusion, Bibliography)
Due date: 3rd tutorial meeting
7/27/2019 Tutorial Single Equation Regression Model
31/32
Further Readings
Gujarati, D (2003). Basic Econometrics 4th
edition. New York: Mc Graw-Hill.
Kennedy, P (2003). A Guide to Econometrics
5th edition. Oxford: Blackwell.
Seddighi, Lawler, and Katos (2000).
Econometrics: A Practical Approach. New York:
Routledge.
7/27/2019 Tutorial Single Equation Regression Model
32/32
For any inquiry regarding to the topics please
text or email me.
Thanks.