The figure in case 3 is an error. This looks like a standard survey where they take 75% from Macedonians, 25% from Albanians + insignificant proportion for others.
From 1110, respondents, this equates to 828+/- 2 Macedonians; 280 +/- 2 Albanians; and 1 +/- 2 Other. The results are consistent in questions 1 and 2. In question 3, if the ethnic proportions are accurate, then the overall proportion for the answer "no" should be 67.57% of total answered.
From 1110, respondents, this equates to 828+/- 2 Macedonians; 280 +/- 2 Albanians; and 1 +/- 2 Other. The results are consistent in questions 1 and 2. In question 3, if the ethnic proportions are accurate, then the overall proportion for the answer "no" should be 67.57% of total answered.
Comment