The poll of 1,165 adults was conducted Aug. 10-14 using a sample drawn from NORC’s probability-based AmeriSpeak Panel, which is designed to be representative of the U.S. population.
That shouldn't be at the end of the article after I had to click on a separate article. Poor journalistic practices.
Poor journalistic practices
welcome to modern journalism
If your point is that it's a small sample, you're in for a surprise - that's about all you need to estimate something within about 3% for the US population. If they're doing stratification and such it might be even more efficient than that.
Sample size isn't the problem, claiming 0.0000035% applies to the other 99.9999965% is one however. All they need to do is add 'surveyed' after Americans and there would be no problem.
That shouldn't be at the end of the article after I had to click on a separate article. Poor journalistic practices.
welcome to modern journalism
If your point is that it's a small sample, you're in for a surprise - that's about all you need to estimate something within about 3% for the US population. If they're doing stratification and such it might be even more efficient than that.
Sample size isn't the problem, claiming 0.0000035% applies to the other 99.9999965% is one however. All they need to do is add 'surveyed' after Americans and there would be no problem.