iiinova — where your digital self keeps evolving

Ben Jarvis

@benjarvis.bsky.social

419 followers 680 following 47 posts

Analytical sociologist and demographer studying residential mobility, segregation, and inequality at Linköping University's Institute for Analytical Sociology.

Posts Replies Media

Ben Jarvis @benjarvis.bsky.social · 01/09/2026

New blog post. I recapitulate the known, but still underutilized, equivalence between conditional logit, log-linear, and multinomial logit models. My plan is to hammer on this until people start defaulting to conditional logit thinking. benjaminfjarvis.com/blog/clogit-...

A plot showing coefficient estimates and standard errors from saturated log-linear, multinomial, and conditional logit models of assortative mating, estimated using simulated data. Estimated values are plotted on the x-axis, the parameter names on the y-axis. For each named parameter, there are up to five estimates for different models. The plot shows that for each named parameters, estimates and standard errors are identical across models.

141

Ben Jarvis @benjarvis.bsky.social · 26/08/2026

Messing around with blogging. This post, where I generate some simulated marriage data, is the first in a series about using clogit as a go-to for modeling assortative mating and mobility. benjaminfjarvis.com/blog/actors-...

simulated spatial distribution of singles in a two-dimensional grid, with singles distinguished by sex, age, and group membership. Three groups are segregated into three distinct clusters in the map, with younger singles more centralized.

032

Ben Jarvis @benjarvis.bsky.social · 10/10/2025

Kernel density estimates of proximity to parents show clearly differences between groups, but only slight differences between stayers locations, movers' origins, and movers' destinations. Curves are multi-modal b/c of uneven distribution of residential populations across Swedish cities and towns.

011

Ben Jarvis @benjarvis.bsky.social · 10/10/2025

Discrete choice models of residential mobility that account for the uneven distribution of housing show that people in all kinds of couples are more likely to move to neighborhoods near to parents. Differences in effects are slight, but the effects appear to be weakest for exogamous immigrants.

000

Ben Jarvis @benjarvis.bsky.social · 22/11/2024

That’s weird, this very same newspaper spent months and months in the lead up to the election implying that inflation was all Biden’s fault…

210