https://www.gibney.org/a_syntax_for_self-tracking
9625 entries and counting.
I was planning to start writing the software to analyze causations when I have 10k entries. But I recently started already. It turned out to become a new project that feels a lot like a javascript based jupyter alternative running in the browser.
My tool so far does not output p values. At the moment, it visualizes possible cause/effect relationships in kind of a dot cloud. It shifts all effects in time relative to the possible cause. So an effect would look like a dent in the graph on the right side of the origin.
Some cause/effect pairs look interesting. But I want to gather more data and implement appropriate statistical tests before I publish something.
It is an interesting question, which statistical tests to apply to this kind of open log format. If anybody here has ideas about this, I would be very interested in hearing them.
What's the density over time of the entries? I.e. How many entries per day? And are you consistently recording the same kinds of events and information about them? How many different kinds of events?