For those of us who like looking at numbers, the full data for the S4C poll that featured on Y Byd ar Bedwar on Monday have just been published, here.
-
Update: One thing that did strike me as strange is that there is no breakdown by ability to speak Welsh. The poll was conducted at the same time as these other polls:
Welsh Westminster Voting intention
Welsh on deficit
Welsh Assembly cuts
Yet the first of these had a breakdown of Welsh speakers and non Welsh speakers, and by level of fluency too. That data is far more relevant to the S4C survey than to these others, so the omission is odd. Hopefully it is just an oversight.
-
Update 2 [13:06 12 November]: YouGov have confirmed that it was just an oversight, and have put up a revised version with a breakdown for Welsh and non Welsh speakers, here.
2 comments:
YouGov's view of using ability to speak Welsh for weighting given here: http://waleshome.org/2010/01/online-polling-is-vital-to-democracy/
I read that before, Anon, but weighting is not the point at issue here. At that time YouGov's figures showed a high percentage of Welsh speakers, and people wondered whether that had skewed the result. But what is notable from this poll (the numbers are identical, so are from the same data) is the low percentage. 222 and 984 unweighted = 18.4%, 203 and 1003 weighted = 16.8%
In the programme, Mai Davies made points about the percentage of Welsh speakers who thought this and that about S4C, so the data that led to those statements must surely be available. When we see the figures, we can make informed comment on them.
Post a Comment