3 ms·
IMO relative percents should really not be used when trying to consider the impact of things because it gives you no clue about it. If the rate of something is
by somenameforme 5y ago
IMO relative percents should really not be used when trying to consider the impact of things because it gives you no clue about it. If the rate of something is 1 in 1 million then a 40% increase means your odds of having an issue go from negligible to negligible, but a 40% increase makes it sound much scarier. I think the reason for that is that a 40% increase subconsciously tickles our "40% of" intuition, even when we consciously understand that's not what it means.
That said in this case the 40% is genuinely scary, because it's a jump from 3.3% with birth defects to 5.2%, or in other words nearly 1 in 50 babies (1.9% of births) standing to be affected by this issue. Although just quickly skimming their numbers, the sample size is unclear to me. That jump was driven by only 13 birth defects among those taking the drug, based on what the article says is a sample of 1.1 million births.