Oh my god a talk just bothered me so much I got up and walked out
Not ethically, they just make some really bad assumptions for their model that make me twitch (literally).
@Maenad this was the best part of statistical consulting - getting people into models that at least won't give me hives
@gzt problem isn't a lack of statistics - he's in statistical computing
Problem is his head is stuck really far up his butt
@Maenad jackiechanfrustration.gif I guess for some models that might make sense (and it becomes just a cohort variable rather than age) but I'm betting here it did not.
@gzt age specific immunity changes over time
If you ignore that, you're throwing the baby out with the bathwater
@Maenad yeah in that application this is a Very Bad Approach.
@gzt "we used yearly age classes and assumed, for simplicity, that people didn't age over our four year period"