Note the referendums in the UK (Brexit) and in Catalonia (independence). These show every possible example of bias and innumeracy.

Thank you Dr.Gordon for your answer but I expected any suggestion for best accuracy and methodology

Hi,

You can read this paper;

Bennett et al. "A simplified general method for cluster-sample surveys of health in developing countries."

A population study can be accurate it totally depend upon the researcher . If you make your inclusion and exclusion criteria and observe it tightly you can easily perform a correct population study.

To do any population study first you should have clear objective and research question, then your objectives and research questions suggests appropriate study design, sample size and overall research conduct. I am sure there are excess knowledge in this regards.

To observe first the target audiences (aged between x-y). Than, to pilot a inquiry or research to those population in developing countries, adapting the words to the costume of their culture&behaviour.

There are pros and cons to all methods. If you are interested in an online survey method, try RIWI:

https://www.greenbook.org/company/Riwi-Corp

It is very exceptional to talk about perfect study.

A design of a good population-based study would state clearly

The purpose

The specific objects

A brief but effective literature review

A clear methodology (study population, type of study, sample size and sampling process, pilot testing, tools, quality control and medium of analysis)

With respect to sample size calculation, three factors are involved:

a. scientific equations

2. Cost

3. How common the problem to be studied.

I advise my colleague Struga to make fairly extensive literature review to clarify some of these points