Page 1 of 1

Regression with dummy variable

Posted: Mon Jul 15, 2013 11:01 am
by Sha
Hello everyone,
I'm trying to run my regression model which is as follows:
Y = a + β1 X1 + β2 X2 + β3 X3 + β4 X4 + β5 X5 + e
Where;
·         Y is Internet Financial Reporting index;
·         a is constant
·         X1 is company size (Log of the book value of total assets)
·         X2 is leverage / Debt ratio (total liabilities divided by total assets)
·         X3 is profitability/ return on equity (net profit divided by equity)
·         X4 is liquidity (Cash divided by total assets)
·         X5 is a dummy variable for audit firm size (International audit firm = 1 and small audit firm
·         e is error term.

Question:
1) Is my ratios/data set to be used on Eviews appropriate? (kindly find attached data set)

2) Please confirm when i click on - quick - estimate equation , my equation should be as follows:
ifr c cosize lever profit liqui audit

OR for dummy i should write it in another way?

3) I run my data set with the above equation and window appeared indication ''Near singular matrix error. Regressors may be perfectly collinear). Please help.

Many thanks in advance.

Re: Regression with dummy variable

Posted: Mon Jul 15, 2013 11:02 am
by Sha
Data set for below regression.

Re: Regression with dummy variable

Posted: Mon Jul 15, 2013 1:11 pm
by EViews Gareth
1) Yes.
2) Yes.
3) Your dummy always takes the value of "1". There are no Zeros. Thus it is perfectly colinear with the constant.