Page 9 of 9

PostPosted: Mon Dec 06, 2004 1:59 am
by Alexandros Lordos
MicAtCyp,

Your sums are indeed correct ... in my actual data, the votes for the referendum are as follows:

No - 65%, Yes - 27%, Won't say - 6%, didn't vote - 2%

As you can see, the Yes vote was over-represented and the No vote under-represented. While 27% for Yes is within the acceptable margin of error (24% + 3%) It would seem that many of the No voters chose not to reveal what they voted ... I don't know why. It must be something to do with No voters being more suspicious than Yes voters, and therefore also suspicious of the person conducting the telephone interview. Suspicious people don't easily admit what they have voted.

You are right to point out that in the light of this, the figures in 3.6 of people who voted Yes while knowing the plan extremely well have to be revised downwards ... maybe by 5%, down to 35% instead of 40%.


I would be glad to do any more correlations that you ask for, eg for party affiliation by referendum vote, and the results will be reliable so long as we keep in mind that the Yes vote for each party will be over-represented by a few percent points ...

(hmm... let me do this now ...)

OK, here it goes. Numbers are approximate, and correction for the over-representation of Yes have been made:

AKEL 25% Yes, 75% No
DISY 40% Yes, 60% No
DIKO 12% Yes, 88% No
EDEK 10% Yes, 90% No
NEO 95%+ No
EDI 95%+ Yes
OIKOLOGOI 30% Yes, 70% No

(numbers for the three small parties are not statistically reliable because the sample is too small, so read them with caution)

Now, concerning refugee and non-refugee status, the numbers are as follows: (correction for over-representation of Yes have been made)

Refugees: Yes 30%, No 70%
Non-Refugees: Yes 20%, No 80%

To me, this result is not surprising: Those refugees who happen to be in the territorial adjustment area (e.g. from Morphou, Famagusta or even Karpasia which would have had special status) would have a greater incentive to vote Yes than non-refugees, who are likely to focus more on functionality and security issues.

As for the full data for questions 1, 2 and 6 ... isn't this already published in my report? Anyhow, let me know what exactly you need and I will prepare it for you.

PostPosted: Tue Dec 07, 2004 10:22 pm
by MicAtCyp
Very interesting figures Alexandre! Thanks. So it seems AKEL is the only one which is very near to the eventual result.
On second thought, no I don't need the complete data, as it would be extremely difficult to process it myself. If I need anything, or want to have some more co-relations, I will just ask you.
Thanks again. Please be patient until I study the full report.