I hate that when I look for statistical methods or formulas, they are often written in such a technical way that I can’t understand a single thing without looking up each individual concept. Academic papers are usually this technical, or so it seems. I don’t have a degree in statistics, and so I don’t have enough prior knowledge to make sense of things without spending inordinate amounts of time trying to find the precise term I’m missing. I’m also willing to guess that many people who study statistics don’t necessarily remember everything they learnt.
Statistics can also be written in such a basic way, that the more advanced concepts I’m looking for are completely absent. Looking for a key bit a knowledge to complete an analysis can take much, much longer than expected simply because the information is incredibly difficult to find.
Now I have spent some time wading through statistics for epidemiology and evidence synthesis, I want to help. In this section of this site I will write about methods of evidence synthesis for medical research (combining information, for example several medical studies) in Stata, and sometimes R (statistical packages, more later). I will describe the methods in as much detail as possible so anyone, even someone with no statistical background, will be able to understand what is going on. If you have any trouble with anything, let me know and I will do what I can to make things clearer.
All the statistical code I use will be displayed and explained. I know how to program in Stata already, but am slowly learning R, so although everything should work (no guarantees) my code may be inefficient, inelegant, or just plain ugly. I think most people who write any form of code feel similar.
My hope is that whatever I write, whether it’s about cleaning or transforming data, statistical method underlying evidence synthesis, meta-analysis or anything else I can think of, anyone could dive in and understand how something is done and why it works. I also hope this blog will be a good resource for learning Stata (and R) in general, and specifically for evidence synthesis.
- 29/10/2017 Albatross plots: Part 1
- 26/10/2017 Series on Evaluation of Scientific Publications
- 22/10/2017 Introduction: Questions about Stata, R and other things