Вы находитесь на странице: 1из 10

How to use micro fit 5

1. Drag file into micro fit


2. Window will be open like this

Note: GDP ,FDI etc are my projected variables .

3.
4. Now you want to show your variables in a graph then write in commands and data
transformation box ,,,, plot gdp , plot fdi etc what variable you want to show in graph and then
run

Muhammad Saeed Aas Meo superior university Lahore Pakistan

Page 1

5.

I wrote plot gdp and run results are these

6. How to convert data in log form in micro fit


7. Lgdp=log(gdp); Lfdi=log(fdi) ; run (when you will run all variables will gone in red color next you
must write variable in this shape in data command table
8. Dgdp=lgdp-lgdp(-1) ;dfid=lfdi-lfdi(-1); etc and run (again variable gone in red color)
9. Now check log of data in plot form like this plot lgdp lfdi run
10. L stand for log while G stand for DE trending in data , we must check data in DE trend form also
so that we can understand sationarity of data

Muhammad Saeed Aas Meo superior university Lahore Pakistan

Page 2

Here we have de trend our data now when ill run


our data then a window will open that will show
stationary of data

11.
Here we can see variables are stationary
as compare to above result

12. UNIT ROTT


13. Write like this for ADF and must use leg length
Muhammad Saeed Aas Meo superior university Lahore Pakistan

Page 3

Adf lgdp(5); adf lfdi(5); etc


Adf dgdo(5); adf dfdi(5)

How to use ARDL APPROCH IN MICROFIT


1. Drag you file into micro fit select sheet on which u have your data, ok
2. When will ok then your variable will appear in red color and a message also you will receive that
you have imported data successfully now

3. Add constant then add trend


4. Go to univeriate ---ARDL to co-integression approach

Muhammad Saeed Aas Meo superior university Lahore Pakistan

Page 4

5. Window will be open like this and go to variables

6. From above arrow you must select first dependent variable then all independent variables then
you must write & you must add inpt and trend from arrow box. And run

Muhammad Saeed Aas Meo superior university Lahore Pakistan

Page 5

7. Above equation run then this window will be open by default Schwarz Bayesian SIC will be
select if you like any other criteria you can but mostly peoples use this so it is better than ok

8.
Muhammad Saeed Aas Meo superior university Lahore Pakistan

Page 6

9. You have no need to change any other option again ok means display option 1 is good
Note: this resulted window is just like a trailer, because from here you will
decide either your data is able to move ahead or not, because here you will
find autocorrelation, HSK etc.
Short run relation
Note; if trend
insignificant u can
delete it

Our f value must be


fall with in the upper
and lower bound
values

10.
This is result window it is too long so next picks is also these results part.

Which things u need


We must see serial correlation in data, hypothesis for
serial correlation
Ho: there is no S. Correlation
H1: there is a correlation
We accept Null Hypothesis because P-value is more
than 5%
Father we must check Heteroscedasticity
HO: there is a homoscedasticity (ho is accepted P
more than 5% which is good)
HI: there is a Heteroscedasticity

Note if trend is not significant you must delete it and again run Ardl model
Muhammad Saeed Aas Meo superior university Lahore Pakistan

Page 7

Step 11: close above window from close button then this window will appear,and cancel this
window also,

Close this window also, with cancel


button

Step12: When you will cancel the above window then this window will appear, and select option
2 which is selected by default, and ok

Select 2, which is select by


default, when you will press
ok button then long run
results will appear.

Muhammad Saeed Aas Meo superior university Lahore Pakistan

Page 8

These are long run results, where you can


conclude about the relationship between
projected variables on the base of porb, value
here prob value of gdp and oil is less then five
percent so you can conclude there is
relationship exist between gdp, oil and co2
emissions, while on the base of coefficient
you can conclude either positive relation exist
or negative, so gdp and oil coefficient values
are positive so we can entirely conclude that
,gdp and oil associate with co2 positively and
significantly, means increase one unit in gdp
and oil will be a cause of (.152,.002 unit,
increment in co2 Emission)

Step 13 and last step , now we will cheak short run relation , close abouve window then one an
other window will open you must close this also like

Close this window and after


this another window will
appear, select option number 3
for short run relationship.

Muhammad Saeed Aas Meo superior university Lahore Pakistan

Page 9

ERROR CORRECTION MODEL In


MICROFIT
From the above window from where we have chosen long run relation option at number three
error correction model option is also available, trough you can chose the ECM

Fdi is only one variable which


is positively significant with
the gdp in short run .

These are the results of ECM which tells us about the short run relation between project
variables,, noted ,,, ECM value must be negative ,,,, which tells about the conversion speed
toward equilibrium state.

NOTE , if f-Statistics comes higher than upper bound then we will say there is integration, if
your F-Statistics you find not higher then upper bound you can change lag order,
Another case let suppose you F-statistics values comes upper than lower bound but not higher
then upper bound ,now you must drop any one variable , or add, but if you F, statistics values
comes les the lower bound then you must change your model and last apply diagnostic test, like
check normality, autocorrelation HSK ETC.

Muhammad Saeed Aas Meo superior university Lahore Pakistan

Page 10

Вам также может понравиться