This week's maxim for my #demography / #PopulationModeling students:
Do you have a question about a duration / state of being, but data on rates / a set of transitions? That's how you know you have a life table situation
If your answer is, "No, I have a question abt the duration of an experience but I don't have data on the transitions, just a cross-section of the experience," then, surprise: you have a "life table plus extra assumptions" situation
I need to work on the pithiness of my maxims
@wrigleyfield Your maxims are good, even if not pithy!
I’m trying to come up with something along the lines of “you can’t introduce time into data that doesn’t already contain it except by making assumptions,” but I feel like that’s not quite as incisive as what you’re actually saying.
@inthehands I like that one!
Something that always amazes me is how much mileage demographers get from the most limited time data
Here's an example I love. In our class "preview," where students work together to try to solve a puzzle that the next week's reading will tackle, I summarize it:
"You have a Census at 2 time points. That is all the info you have. Have mortality and fertility changed over time in this population? Was there much migration? How can you tell?"
@wrigleyfield Oh, that’s a nice question! (“Suppose between the 2 time points the •entire• population left, and an entirely new population arrived. How would you know?!”)
I get lost in the weeds of the paper’s numerous helper functions pretty quickly, but the gist of it makes perfect sense. Major Green’s theorem vibes.
Elizabeth Wrigley-Field
Paul Cantrell