2 ms·
No need to poll 100 million. 20,000 is pretty accurate.
by megablast 5y ago
No need to poll 100 million. 20,000 is pretty accurate.
- kstrauser 5y agoOK, sure, but that doesn't change my point. You could run all of someone's blood through a scanner and then back into them to get a complete view, but a lab only samples a few mL of it.
- dvhh 5y agoUsually your sample size would be a good indicator of the accuracy of your poll/test/experiment. Of course, like everything it could be biased by the used methodology.