Random sampling is the fundamental principle upon which many statistical analyses rely. When the population is as heterogeneous as it is in the Gulf with cultural sensitivities to do with contacting females for interviews, random sampling is virtually impossible. Its concept assumes each contact has the same chance of being converted to an interview and that is basically not achievable in the Gulf.
What it means in practice is that male interviewers should not contact Muslim females with a traditional mind set. Refusal rates are higher when Asians attempt interviews with Arabs and respondents in higher income groups are difficult to reach. When interviewers do manage to get through to them, these respondents are more likely to decline participation. An attempted random sample in the Gulf, therefore unchecked by quotas, inevitably leads to a disproportionate set of respondents and the results of such a survey are not representative of the targeted population, which in turn leads to misleading results.
The norm in Arab countries is therefore to work with a sample frame and quotas, based on an estimated universe developed from secondary research. Interviewers engage respondents through the so-called snowballing technique, which is essentially a method that draws on referrals and fresh contacts. Interviewers are more successful when they approach respondents of the same gender and ethnicity, as they can strike a better rapport. For this reason, Field Managers need to operate much larger interviewer teams compared with Europe.