Panel/Pool Dummy Variable

[katiep]

Hi,

I have spent a while now searching through this forum to find an answer to my problem, I appologise now if I have missed something.

I am currently doing my dissertation and have data from 9 regions across 10 years with 1 dependent variable and 8 independent variables, basically looking at the causes of regional unemployment.

I have this data in both pool and panel, so would be happy to use which ever is easiest. My problem is I genuinely have no idea where to start.. I have carried out OLS for all regions together and now want to do it per individual region and then split data in to two sections (the north and the south), preferably using dummy's to reduce time. I have carried out one regression using one region manually but my results were all non significant which doesn't seem to make sense. ie ue_ne c edu_ne etc etc (ne=north east)

Wondered if anyone could tell me where to start.. sorry if this isnt very explanatory, I have been staring at data for about 7 hours now!

Thanks in advance

[startz]

If this is annual data, then when you only use one region you have 10 observations with 8 RHS variables. You can't really expect much to be significant.

[katiep]

Thanks, completely forgot to look at it that way. With regards to splitting my regions into two groups, how would I create these dummies? There would be around 45 observations so hopefully this step may be more successful, and I shall try to find some more data for the other steps.