Urgent Help "Panel Data Issues"

[annie]

Dear Users,

I am having several issues with my panel data, the first problem I get is of "insufficient number of observation".

I have tried removing my independent variable but then when I try to regress my equation, it gives me the message of "near singular matrix".

I am fairly new to this but I have done all my reading, I urgently need the results so any advice would be useful.

I am attaching both my excel and eviews file.

This is my equation (vardi c bsize bindep rcom rcidep instiowner bm size cap ratio news)

Thanks

[startz]

The problem may be that you've read in some of your variables as alphas rather than numbers.

[annie]

Hiya,

How can I overcome this?
Thankyou

Note: Are you able to open my eviews file and edit it, would be much appreciated. (this is taking over my life)

Ok I got it thanks.

So I think I should be able to run the equation.

[startz]

I would think you can just drop your excel file on EViews.

[annie]

I am able to run my equation, although it is not giving me the results I want.

When I try to do fixed or random effects, it says "the equation does not contain fixed or random effects".

Do u know why this may be.

Thankyou

[startz]

You need to use the estimate tab to tell EViews if you want fixed or random effects.

[annie]

I am getting negative results where it should be positive, my first five independent variables should all be positive. Is this to do with my data perhaps. Is there any way I can overcome this problem.

[startz]

It could be your data. It could be your specification.

Some days research doesn't go as expected.

[annie]

It could be your data. It could be your specification.

Some days research doesn't go as expected.

Thankyou, could it be the problem of my decimal places, should all the numbers be in decimal in my excel sheet, I am just confused because it should be all positive.

[startz]

Decimal places are unlikely to matter.

How do you know what the coefficients "should" be?