It’s the second day of 2018, which means time to start acting on my new year’s resolution to start a data blog. In the coming year I aim to post old and new data analysis projects, and in doing so learn more about markdown, R, A/B testing, data analysis, web design, and more. In pursuit of improving my blogging skills, I will also be following more blogs, so feel free to send me your blogs!

Goodbye academia

The start of 2018 also marks the official end to my more than six year career in academia. Over the course of my entire academic career I have had the privilege of accomplishing a number of things:

  • published 3 peer-reviewed articles in top sociology journals, with 3 more and 1 book in the pipeline
  • wrote 4 conference papers
  • presented at 13 national (Dutch) and international conferences
  • collaborated on data collection for 2 large-scale surveys
  • learned Dutch, and then taught statistics in Dutch, covering basic probability to logistic regression
  • co-organized the first R-ladies chapter in the Netherlands

In all I have worked on 9 projects, the scope of which has ranged from months to years. Data cleaning was involved in all those projects, while A/B testing was only involved in one. 7 of those projects used survey data and 2 used historical register data. More details are available in the figure below.

Hello blog

As good as academia has been to (and for) me, there are three things in particular that I have yet to accomplish. It is my hope that I can achieve these with this blog and in my future career.

1. Impact

My academic research was focused on answering a very practical question: Why do some fathers spend more time with their children than others? The answer to this question can benefit children, mothers, and fathers themselves. And yet. I guarantee less than 100 people have ever read one of my articles or heard me present my findings. If only 5 people read each blog post, I will have already bested my average impact for my academic work.

2. Reproducibility

At its best, academia is (or, prides itself on being) different from business because it is about conducting the highest quality analyses in an ethical way. At its worst, behind every published, well-cited article sit 20 different analyses that were never submitted to a journal, never passed peer review, or never cited because their findings were contrary to conventional wisdom in the field. As much as possible I will post my work with code and data so that my findings can be reproduced and improved.

3. Speed

In the race between the tortoise and the hare, academia is the tree the hare falls asleep under. To be sure, more experienced researchers are faster than I am at getting published in academia. My thesis supervisor could probably take an article from idea to publication in six months, if she was using all over her connections to speed things through. But back in the real world, empires have risen and fallen in the six months it took to get that article published. I aim higher. I aim for speed. With this blog I will reverse ~6 years of training and publish one less than perfect but good enough analysis every week.