I get it!
You've re-weighted the internals and Bush's lead is 49.14 to kerry's 46.61.
Thanks!
I get it!
You've re-weighted the internals and Bush's lead is 49.14 to kerry's 46.61.
Thanks!
Well, what you see is just how I calculated the numbers by distributing the 1,461 Registered voters according to the published Gallup internals. Since Gallup (as do others) rounds their answers to no decimal places this explains the differences from my numbers. The only re-weighting done in the table is the very speculative right hand column which attempts to rigidly fix the Gallup internals to the year 2000 VNS exit poll data as do some pollsters (Zogby, ARG).
This is also part of the methodology I use to calculate the percentages of Republicans, Democrats and Independents in the sample. If the pollsters would publish their results to a single decimal point of precision, then none of what I do would be necessary...
Now compare the computed table above to the Gallup published table below. If my data table is rounded to NO decimal places, then it matches the results which Gallup published. This is where the 'extra' precision came from in the above data tables. It's the rounding...
Hope this helps...
dvwjr