2000 is a very small figure compared to 6 billion people in the world. I would say that this experiment would not be considered a true random sample of the population.
If it was more like a random sample of 100,000 then i would believe the numbers to be a correct representation.