I have a mostly successful go at cohort component population projections to replicate the UN's totals from their published parts, with an idea to making small changes to observations or assumptions that can build on the official projections.

I compare the GDP per capita and scores on the UN Multidimensional Vulnerability Index (MVI) of the 68 economies in the 'V20' group with other countries that aren't part of the V20.

I make some visualisations of the country scores of the UN's proposed Multidimensional Vulnerability Index

I have a go at an 'insanely hard' (actually not that hard) problem to find the radius of a circle from someone's recruitment exercise

I make an animation and a basic Shiny app to explore the United Nations' model life tables used for demographic estimates in countries where direct estimation of mortality rates by age isn't possible.

A model that is 'improved' (in terms of making standard assumptions more plausible) by using a logarithm transform of the response will not necessarily be improved for estimating population totals.

When working with complex survey data where the weights are related to a continuous variable of interest, using a weighted rather than unweighted percentile rank will lead to different results towards the middle of the distribution; but the two measures will be highly correlated with eachother. Also, R reportedly calculates weighted percentile ranks much much faster than Stata.

I demonstrate the function I use to make it simpler to draw choropleth maps based on Pacific Island countries' and territories' exclusive economic zones.

I do some simulations to reproduce a great figure by Wysocki et al; and show different data where the causal relationship between x and y is in the presence of a third variable that is either a confounder, collider or mediator.

I compare vaccination rates in the Pacific to GDP per capita and find the evidence isn't strong enough to say that there is a relationship between the two.