Data-Driven Article Replication and Analysis

Replication of a data-driven article and analysis of the UK’s healthcare systems among other countries; by British editor Tom Calver

Replication of a data-driven article/analysis of the impact (if any) of President Trump and his administration on international tourism into the United States; by Josh Holder et. al

Here I explore some ways to draw cool looking objects, here snowflakes namely using fractals and their properties - inspired by Micmaths.

This personal website serves as a repository for my replications of interesting studies/articles, analysis, and explorations of topics that spark my curiosity. As a data enthusiast, I enjoy diving into datasets, uncovering insights, and sharing my methods/findings - in as many coding languages as I possibly can.

On Replications: There has been a clear increase in the use of data ubiquitously and over recent years, news articles are therefore no exception. Embedded in news articles these days you can, and without much effort, find anything from simple tables to fairly involved/advanced visualizations. But of course, you seldom find the exact methodology the author(s) used in presenting said visualizations to their readers. The purpose here then becomes to reverse-engineer the final article output to generate a clear path to each ‘element’ that is required to make said output - almost like reconstructing a puzzle.

“Where there is a will, there is a way”; similarly, where there is HTML, there is its source page- less poetic but equally true. This will be the leitmotif for most data extraction from articles I work on replicating; unless they are behind a paywall (where an account has to be created first -with or without a paid subscription); or JS is intentionally set to not redirect; not allowing for the actual data to be revealed under the hood.

Along the way, I hope I do a good job explaining why I went with one method over another based on high certainty; meaning that this is what the author did as well; or purely making an educated guess, and then seeing how well the guessed method’s output matches the article’s. I find the exercise interesting for multiple reasons, one of which being thinking the same way the author thought so that both outputs align; that can be a useful learning experience for me and hopefully for you as well !

In terms of coding choices; I will mostly be using R in tandem with Quarto along with python and potentially explore additional languages/frameworks along the way.

On Analysis: Once things get streamlined and time to publish additional replication projects shrinks, I will most likely begin devoting time to dive into articles or more generally online sources that I will either support or critique. This will be relying more on the analytical side of things in making my point of whether or not I think the article/piece drives home their point with rigor; and if not, suggest concrete improvements; be it on the choice of methods and/or how the message is conveyed visually.

On Opinion: This section will really focus more on (potentially pompous) philosophical and metaphysical takes (I warned you) to take a step back and some hindsight into deciphering the speed with which things are evolving, tackling many topics..more to come on that section !