It is often stated that condoms or 98% reliable.
How is reliability seriously computed?
If you’re talking about pregnancy, the 98% figure is in woman-years. That is, if you take 100 sexually active women and all of them use condoms correctly for one year, 2 will become pregnant, on average.
It doesn’t mean there’s a 2% chance of pregnancy from any given sexual encounter, or that there’s a 2% chance of pregnancy over the course of your lifetime.
Wow that’s really how they calculate it? Isn’t that statistic entirely dependent on how often they have sex though?
Yup. The “use correctly” bit kind of changes things too, since a lot of people don’t, which accounts for most pregnancies that occur despite condom usage.
Planned Parenthood estimates a 75-85% reliability rate based on actual usage as opposed to correct usage.
Some RA jobs are better than others.
So, that makes a lot more sense than just a crude percent as often reported but it would seem one needs to know the number of sexual encounters and the age of the women too. Wouldn’t we?
I presume that they figure that in when selecting the 3,000-ish women who participate in studies, but I don’t really know how they account for that.
If you take a large enough sample of randomly selected women who actually use condoms, then it doesn’t matter - or at least - the figure will be a correct average. It just won’t tell you much about the percentages you may actually be interested in.
I still don’t get it though. How do they know if the sample woman are using “correct usage” or “actual usage”
AFAIK, that’s based on the subjective opinions of the users.
If that’s so, it doesn’t seem fair to call it “actual usage” and “correct usage”. It should rather be “use it every single time”/“don’t”, or whatever question they asked the users to place them in those two groups.
Exit polling.