3 ms·
And yet, this Pew Research poll from two days ago: > "Overall, 56 percent of Americans consider the NSA’s accessing of telephone call records of millions of Am
by gmisra 13y ago
And yet, this Pew Research poll from two days ago:
> "Overall, 56 percent of Americans consider the NSA’s accessing of telephone call records of millions of Americans through secret court orders “acceptable,” while 41 percent call the practice “unacceptable.”" [1]
The Pew question only specifies phone data monitoring, while the Gallup question mentions both phone and internet data. I'm sure the poll-question-phrasing-savvy can provide some more insight on the differences.
How much of the difference can be attributed to people's minds changing in response to further revelations, the quality of the government's response, the opinions of their peers, etc?
[1] http://www.washingtonpost.com/politics/most-americans-support-nsa-tracking-phone-records-prioritize-investigations-over-privacy/2013/06/10/51e721d6-d204-11e2-9f1a-1a7cdee20287_story.html http://www.washingtonpost.com/politics/most-americans-suppor...
- mpyne 13y agoWell we've always known that both the wording and sequence of questions can affect poll results. I don't know by how much exactly but the effect has been seen to be significant enough to mislead pundits on seemingly every side of a given issue until right before an election.
- jessaustin 13y agoThis summary has been somewhat debunked: http://erratasec.blogspot.com/2013/06/nsa-poll-you-are-reading-numbers-wrong.html http://erratasec.blogspot.com/2013/06/nsa-poll-you-are-readi...