Is a 4,000 Person Sample Big Enough?

How can a 4,000-person sample capture a country of 300 million?

Dr. Henk Pretorius, co-founder of Timelaps, explains that sampling theory is why you don't need to speak to millions of people, even when the population is that large. Timelaps uses margin of error, the universally recognized standard for this kind of sampling. On a sample of 1,000, the margin of error is roughly 3% at a 95% confidence interval. That means results can be accurate without surveying a full census. Timelaps runs 1,000 respondents every quarter, rolling up to 4,000 across the year, so each wave has a sufficiently large sample to report on. The same principles underpin public opinion polling and academic research.

Henk spent 20+ years in market research and built and sold the largest digital research agency in Africa. Learn more: https://timelaps.io