Economist stuggling with panel data

[mattgammon]

Hi,

I am an undergraduate struggling to get to grips with panel data for my dissertation and would really appreciate any advice.

My model is as follows

Obesity = a + b1 economic inequality + b2 economic security + b3 time + b4 liberal country dummy

I have a panel data set which includes 14 countries over 11 years.

I have managed to get all of the data into eviews and have run a couple of basic regressions, however I am struggling to find out which estimation procedure is best. The procedure that seems to produce the best results is when I estimate it using the period SUR option on weights but I am clueless as to whether this is what is best to use and can not seem to get my head around the idea of whether fixed or random effects is the better option.

If anyone has any advice on what steps to take to decide on the best specifcation I would be very grateful.

Cheers,

Matt