Hi all,
To help me pass the time and procrastinate on my thesis, and really just for the giggles of it, I took the country notification spreadsheet, modified it, and did some calculations in STATA. It was only supposed to take a good hour or so but since I'm not fluent yet in STATA I had to figure out a number of the syntaxes for panel data and whatnot. Anyway, here's what I did:
I took the notification date for each country and regressed on years, standard OLS method, and had STATA perform the calculation on every country, and save the individual constants and coefficients. Then, I put it into excel, corrected the data, and calculated the predicted date value for 2012. Basically, I took a linear regression of every country's notification dates to "predict" what time they would notify people this year, based on previous trends. I then listed the countries in order of predicted notification dates.
Obviously, the results from this are more or less bologna and you could probably tell as much just from a glance at the spreadsheet. The data is awfully limited (for some countries not even enough to do a proper regression) and obviously there are more factors than just the year involved. Still it was kind of fun to make, and the predictions are interesting if only to gawk at and ponder over while we waste more time waiting for the actual decisions. For example, it predicted that Germany and the UK notify between 3/6 and 3/7 - pretty close to what happened, right? And the real reason I did this in the first place was to see what the prediction for Turkey was, given that in the past it has been notification in March...then April...then May...
Yup. I'm supposed to hear back 6/1/12. =P.
Spreadsheet for your amusement located here: https://docs.google....0wT2JZNnc#gid=0
Good luck all!
P.S. I'm just going to reiterate that these aren't actual, real predictions so much as just regression results. Don't be mad at me if they aren't true!!!