In a policy-dependent survey, you can use these tools every day to argue freely opposing views? Considered when creating them and their analysis is an activity rather than discretionary.
fact, I noticed how everyone tends to prevail in its case, in reporting the "revolutionary" sondaggistici data.
We enter the key issue, namely the survey DIGIS that gave me inspiration for the theme. Consists of shares of the electorate to the right and left (a total of 2200 individuals), to probe their sensitivity to issues traditionally dear to a political party in particular.
We see some of the questions.
Questions electorate right
"It's just the strict protection of civil liberties (speech, press, Association etc.).? "
" She supports the freedom of choice for women on abortion? "
Vince in both cases is , ( 89.1%, 69.2% ).
Questions electorate left
"crime and illegal immigration must be fought with tougher laws? "
" She is in favor of halving the employees the time and cost of the bureaucracy? "
Vince in both cases the you ( 67.3%, 86.1% )
I heard these data define " sensational," to demonstrate how the electorate not deviate from typical design that has ideas of right and left. But how do you define the data made sensational falsifiable; when the response is implicitly contained in the application?
Just read carefully the demands of pollster above.
do not have to be right or left to support the "freedom of choice " of women on abortion, it would have been very different if you ask for or against abortion in itself.
do not have to be right or left to be in favor of " fight crime and illegal immigration " would have been very different if you ask favorable and against racial integration.
Similarly, who would not " halve staff, time and cost of bureaucracy? .
I would like to connect to the theories of Sartori about it.
thought weak
E 'from first consider the "weak thought" that permeates the survey itself as such. Many of the views of the statistical sample certainly born at the time. About a dozen questions often the other person has a strong opinion on only 2-3 of them, while the other is easier to answer the instinct of the moment, thus creating a review labile, changing an initial investigation by the interviewee.
And this is at best . Often the other person does not even know what you're talking about, but its non-knowledge of the subject does not affect its ability to respond in the least a simple " you" or "no .
Carry a view of the Montanelli Italian Constituent Assembly while the work was still going, having to decide if the best was to bicameral parliamentary system or not. Following a survey among Italians, he wrote "I think the majority of respondents believe that the two-chamber is a double bedroom.
The importance of demand
What may be relevant to the choice of the words in question?
In American elections of 1988 between Bush and Dukakis (the Democratic candidate for president) when he was placed first in the application "Dukakis," Bush was below 12 percentage points , a difference that reduced to 4 points when the name Bush had made up his mind.
If a transposition of words alone can give such different results, because it can alter the response of the audience a question entirely under a different light, as in the survey DIGIS?
This is not to delegitimize the work behind the polls, but feel daily proclaim Berlusconi enjoys the confidence in the electorate (the shame not to mention their strong opposition to parliamentary immunity) , and each party trumpeting those percentage points collected periodically favor of one or another faction, suggests a rather uncritical reading of these.
should more often look at what lies a seemingly innocent question.
0 comments:
Post a Comment