Org parsing, shorthand generation, and paired comparisons
I get annoyed by a breakage in Org which turns out to be reasonable. We look at a swoosh generated by JavaScript. And we learn how paired comparisons get their power.
Hello! I hope your week is going well.
New articles
org-element Reimplements Emphasis Parsing
A small annoyance in the Org export made it hard for me to use smallcaps in parts of words. I thought it was a bug in Org, but I'm no longer convinced. Either way, the mailing list had a workaround.
Full article (2–5 minute read): org-element Reimplements Emphasis Parsing
Generating Shorthand SVG in JavaScript
Learning shorthand consists of 2 % memorising shapes, 8 % writing shorthand on one's own, and 90 % reading well-written shorthand for practice. But where does one find well-written shorthand these days? Not happening. I've always liked the idea of turning a web page into a well-written shorthand document, but thought it would be too difficult.
I no longer think it is too difficult. However, I do realise for this sort of application, the main effort is going to lie in polish and making it practically usable. This is time I'm not willing to spend, so this article is just a preview of the idea.
Full article (1–3 minute read): Generating Shorthand SVG in JavaScript
Flashcard of the week
Since I'm looking forward to re-reading Fisher's Statistical Methods once I'm through the backlog of all the other things I want to do "soon", here's a flashcard from the first time I read it:
When testing the difference between two treatments as the mean of the pairwise differences rather than the difference between the means of the two samples, we sometimes get more power. Where does this power come from?
We can probably answer this with intuition: assuming that a paired comparison was appropriate, that means the pairs were of the "all else equal" kind, and that the only difference between the measurements would be the effect of the treatment. Thus, by excluding other effects, we get more power.
But that's not strictly true! If we wanted to find out the difference in height between male and female humans, we cannot construct "all else equal" pairs (unless we happen to sit on a pile of identical twins that are fed and raised exactly the same). It might seem we're stuck to comparing the means of all males and all females – but we're not! We can measure the height differences between siblings of different sex. They're not "all else equal" pairs, but they're more similar than a comparison across the entire populations. Thus, the power of a paired comparison comes from
Positive correlation between pairs.
... somehow. I'm still not exactly sure on the details here – one of many question marks I hope will be elucidated by re-reading.
Premium newsletter
The new premium newsletter went out last week!
It had a preview of a future article, a review of Code That Fits In Your Head, links to two of my favourite text adventures you can play for free, and a few more web links. To get access to it, a premium subscription is currently only $2/month, cancelable at any time, no questions asked.
That money also gets you the past issues, which total five longer book reviews/summaries, 22 brief book recommendations, forecasting/world modeling tips, and around 12 great web links.
To upgrade, click the subscription link at the top of this newsletter and fill in your email again.
Your opinions
I cannot improve without feedback. Reply to this email to share your thoughts on any of the topics above, or anything else!Hello! I hope your week is going well.