In late July I wrote an entry that set a precedent in order to 'measure' the value of pre-election polls. You can see here .
After parliamentary elections in Catalonia the result was this:
and what he said the survey by the Institute Noxa was 4 months before this one:
We see that, broadly speaking, the survey was right on the change of government in Catalonia to predict a CiU comprehensive victory (touching the absolute majority of 68 seats). Finally, it has remained at 62.
The fall has been greater than forecast by the survey (32 seats), down to 28.
The survey predicted that the PP did not change the number of seats and has a nice surprise to increase by 4 more from 14 to 18.
The thud of CKD was also higher in reality than predicted by the survey, from 21 seats to 10 (instead of 13 announced by the survey.)
With ICV hit full survey.
With C's completely wrong since the survey predicted that the parliamentary spectrum disappear, and yet, has won a seat, going to 4.
The survey also erred on the incorporation of new parties in parliament because it did not include any new ones. But in reality, the new party of Joan Laporta, SI, has managed 4 seats.
In summary, a simple calculation we can see that the average error of the survey was 3 seats up or down to the parties that have achieved parliamentary representation. What I think is a very fine result considering that the survey was administered to 4 months ahead of elections.
And that raises a disturbing but at the same time, an interesting question:
Surveys undoubtedly further improve prospecting techniques "may one day replace the electoral event itself? Will he be a day when a simple partial sample of the company duly removed, may replace an entire democratic vote, saving even have to make (and support) campaign?
Not to think with the millions of tons in ballots that would save us?
One thing is certain, it would be much more environmentally friendly and respectful to the planet.
0 comments:
Post a Comment