Happy big Friday everyone,
In this issue, I wanted to talk about a new approach I’ve adopted for conducting independent research.
I’ m preparing to advance to candidacy for my Ph.D. This is an exam where I must propose at least 3 novel ideas that I will commit to publishing in research manuscripts. Currently, I’m a Ph.D student; once I advance, I’ll be a Ph.D candidate. The title change doesn’t mean much, but when I pass, I will have demonstrated to my committee that I can do the prep work to do research.
Thankfully, I already have two papers ready to be submitted to journals. This leaves a third idea I have to develop. The idea is already there, but I also have high hopes for this paper. I want to publish it in a high ranking journal in biostatistics and statistics. That’s not a requirement for my degree, but a personal aspiration.
And towards this goal, I’ve developed a new system for organizing my work.
The Central Planning Document
Most people think about the work that will go into a paper, but few think about all the work that doesn’t get in. I’m talking about the scratch code that you write to make preliminary results. Small notes to yourself about why your experiment failed. Late night breakthroughs that happen when you’re in the shower.
None of these make it into the paper, but they are essential to progressing it.
For my first two papers, I did not to keep track of all these smaller bits of information. I have a bad habit of deleting code and old writing out of a compulsion to keep my workspaces clean. With 20/20 hindsight, I cannot count how many hours have been wasted in recreating or looking for work I had already done.
To prevent history from happening again, I’m working on a system for progressing my research and organizing all the information I generate.
First, I create a document that will act as the portal to all the work that I do for this third paper. I call it my “Central Planning Document”. The Central Planning Document contains all of the questions that I need to answer to write the paper.
I use Obsidian as my main note taking app. One of the things I like about Obsidian is that it allows you to create links within documents that allow you to jump to that other documents. Crucially, Obsidian keeps track of these links and can assemble a visual network of your documents.
There are several questions I need to answer in a paper. Each of these questions receives its own page in the Central Planning Document. My advancement to candidacy only requires me to do literature review, so currently, I only have sections for the paper’s introduction and methodology:
For those curious what my idea is, I’m trying to design a personalized trial that accounts for multiple outcomes simultaneously. Most, if not all, diseases affect multiple aspects of our lives. Therefore, it would be good to create a tool that explicitly plans for this, especially since N-of-1 trials are intended to treat individuals.
On these individual question pages, I can store all the information that I gather about that particular question. Since I’m reading a lot of papers for literature reviews, I make documents for each of these too. I’ll usually write out all the salient points that the papers give to answer the question of the page that they’re on. This saves me some time on needing to come back to each reference if I forget.
I’m using this approach because hunting down references is a pain in the ass. I would rather have them all in one place when I’m writing, so I’m leveraging Obsidian’s linked documents to do this.
Intermediate Packets
In Thiago Forte’s Building a Second Brain (note: my affiliate link!), he talks about a method of information generation called “intermediate packets”. Intermediate packets are smaller pieces of rough writing that are relevant to what you’re currently working on. They can be stored for future reference and are easily shared for feedback, which is why I’ve implemented them into my system.
Writing an entire paper is a long-term push, something I expect to be finished within a year. Actually progressing on it consists of answering many smaller questions before you can synthesize all this information. For my workflow, the system of intermediate packets is a way to package each of these goals into a document that I can refer back to once I start writing the manuscript.
Every day, I’ll pick a question that I want to progress. Then, I commit to making at least one intermediate packet for that question. This intermediate packet is usually the creation of a document for a paper or resource that answers one of my questions. I’ll read the paper and summarize the main takeaways, like in the picture.
The paper is the long-term goal, while the intermediate packets are my short-term achievements. This system has worked wonders for driving my writing, but it’s an evolving object. As I get into experiments and results, my needs will change as well.
I hope this has been helpful for you. Have a great weekend, see you next week.
Christian
🧐 What Am I Enjoying Right Now?
Book — I restarted Bernoulli’s Fallacy (note: my affiliate link) recently. It’s a book about the problems of modern (frequentist) statistics and advocates for the use of more philosophically sound (Bayesian) statistics. It’s not beginner friendly, but if you’ve got a statistical background, it’s an interesting read/listen.
Show — I just finished Ted Lasso. Great show. I know from the emptiness in my heart now that I don’t have something to watch for dinner now.
📺 Most Recent Videos
The most important skill in statistics — this is a video on Monte Carlo Simulation, a technique that I think should be learned by anyone interested in statistics. Everyone from beginners to hardened Ph.Ds can benefit from being better at it.