Look at the flawed methodology of this Israeli poll: "About a half-dozen Israelis of Iranian origin, who speak Farsi as a mother tongue, telephoned the 530 respondents in May and early June without revealing that they were Israelis or from where they were calling." So they lied to the respondents but most importantly: this is not a representative sample of Iranians. How were the names collected? On what basis? Random, in sampling, does not mean, random.