Read on Twitter
Back in the madness that was September, @emilynz, @omaclaren, and I managed to send off a proposal to @HRCNewZealand for a project to look how we can use contagion network simulations to improve equity outcomes related to COVID in Aotearoa. 1/N
Since we got funded 
( https://twitter.com/HRCNewZealand/status/1338233361081401344) I thought I should tell people about what we plan to do.
But first to acknowledge the other investigators on the project - @AASporle, @DRR_NZ, @hobbs_PA , @Turdbull, & @JaninePaynter. 2/N
( https://twitter.com/HRCNewZealand/status/1338233361081401344) I thought I should tell people about what we plan to do. But first to acknowledge the other investigators on the project - @AASporle, @DRR_NZ, @hobbs_PA , @Turdbull, & @JaninePaynter. 2/N
As part of our work with @PunahaMatatini, we've built a large network based contagion model. This explicitly represents ~5million NZers, with each of them connected to nodes representing the contexts where they can interact: work, school, home, and community. 3/N
We've built the network (actually an ensemble of different networks*) using empirical data to look at things like: the distribution of dwelling sizes; the ages and ethnicities of people living in dwellings in different areas; who is/n't employed; who works multiple jobs etc. 4/N
*We build a collection of networks since we want to capture the structure and statistics of an Aotearoa interaction network; but we can't (and shouldn't) represent the interactions of specific identifiable individuals.
The people in our network have age, sex & ethnicity, the workplaces have industry sectors, dwellings have intergenerational structure, and all these have locations. This mean we can look at how different groups might be impacted differently by COVID spread. 5/N
Or how different vaccination strategies might play out for different parts of the populations. 6/N
For example, people working multiple jobs have more possible transmission routes. People working in tech or management jobs might be better able to reduce exposure by working from home. People of different ages develop symptoms, and seek testing, at different rates. 7/N
But we also know that the data used to build the network has issues: Census 2018 was a fiasco (technical term). More so for some groups than others. E.g. young Pasifica men are much less likely to be linked to a dwelling in Census data compared with other demographics. 8/N
So, we will also be quantifying the consequences that low quality data has on COVID contagion simulations. If people are at increased risk of COVID (e.g. in crowded housing, multiple jobs) but aren't represented in the data then simulations can't represent that risk well. 9/N
We'll also be looking at ways to better inform the construction of the interaction networks to make them more accurate, using both qualitative ( @DRR_NZ ) and quantitative ( @AASporle) methods. 10/N
Finally, we want people who _aren't_ govt ministers and Wellington policy folks to have access to COVID modelling, so we'll be talking to some community groups (exact people TBD; it's been a busy year) about what questions they want answered. 11/N
The grant kicks off in February but even getting the work this far has been a big piece of work. And we wouldn't have gotten this far without the help of @PunahaMatatini, @UoA_Physics, @Styla73, @Knhannah, @VictoriaLSm, @demivasques, @davidjxwu & others from TPM & @ScienceUoA 
