I am having trouble with the statistical side of a research topic. I
am trying to study what effect a regional trading agreement (the EU in
this case) has on international trade. I have devised a basic
econometric model and think this equation is somewhere along the right
lines:
log Xij =
log b0 + b1 log Dij + b2 log Yi + b3 log Yj + b4 log Ni + b5 log Nj +
b7 log EUij + log eij
The dependent variable (Xij) is the dollar value of country i?s
exports to country j measured according to country j?s import data.
There are various independent variables:
Yi and Yj are the nominal GDP of countries i and j expressed as a dollar value.
Ni and Nj are the populations of the respective countries.
Tj is the tariff rate on imports.
Dij is the distance between the commercial centres of the two countries.
EUij is a dummy variable for trade between partners of the EU.
I have compiled all the data that I think is required to carry out a
statistical analysis in an excel spreadsheet. Distances will affect
trade, however, it is only necessary to test a small number of
countries I think. However, here comes the main question: I am unsure
of how to manipulate the data or what tests need to be carried out to
prove the hypothesis that the trading bloc of the EU is having an
effect on international trade.
There are a variety of tests and techniques I think I should be using,
but have completely forgotten from uni days how to, including:
regression equations, coefficients, standard errors, t-ratios,
F-ratios, R2, etc. Please can you explain in detail which of those
should be used, why they should be used, how to use them, how to carry
out a regression analysis, and anything else needed to obtain
appropriate results that will enable me to establish the effects on
trade.
I also need to be able to understand possible problems associated with
the regression analyses (e.g. autocorrelation, multicollinearity,
heteroscedasticity, and model specification problems, etc) and carry
out adequate remedy measures if required.
Please follow this link to view the data that I have compiled to carry
out the analysis.
http://s62.yousendit.com/d.aspx?id=34NOBLNIG22IQ3HU06MXV4NP8J
I realise this is a complicated question, hence, the high fee, but
would appreciate it if you could at least begin to help me carry out
the analysis.
Thanks. |
Clarification of Question by
wrangle-ga
on
24 Sep 2005 14:56 PDT
Thanks a lot for replying, btw this problem is very urgent and there
will be a considerable tip for whoever answers it.
Anyway, to address your question:
1. I am extremely confused. I have been reading a lot on regressions
and my head is about to explode. I agree now that I need the
dependent variable data (I wasn't sure before), however I haven't
sourced it because I'm not exactly sure what I need. Do I need import
data or is it export information or is it something else?
2. I understand what you mean. To gather all the distances for the
128 countries is unfeasible, that would be around 16,000 entries. So
I only want to take a small sample and still get a rough idea of how
that will influence trade (that's why I chose countries inside and
outside of the EU to be representative). So yes, it only needs to be
6 datalines for the distances regression but I still want the 128 for
the other independent variables.
|
