So my best solution to this so far has been to use dictation.

When we decided to finally get this out (thanks Kevin!), I was worried the reviewers would want us to collect more data, but luckily they didn't. They DID want us to update the analysis, so I was very pleased that I had organised the data in a sensibile way, and my old R scripts still worked!
/end

But he showed me how, and I did it, and it was fine. Just the first of many ways Saad has helped push my boundaries.

It was the start of a long scientific friendship. I remember when piloting experiment 2, Saad asked me to inject anaesthetic superficially in his skin. I was like "you know I studied psychology, right??"

And on a personal note, the data for this paper were collected in 2013, and I'm so happy we got this out of the file drawer, and into the light of day. Back then, I was finishing my PhD, and Saad was starting his post-doc.

It turns out the PCs can do all the heavy lifting and don't need help from the other mechanoreceptors when it comes to frequency discrimination. And we used a low frequency range, around 20Hz, challenging a common idea that this frequency is not optimal for PCs.

This forced our participants to rely on remotely transmitted vibrations, that are picked up by only a subset of mechanoreceptors, the super sensitive FA2 / Pacinian corpuscles (PCs). We found that under these conditions, people can still discriminate frequency just fine!

To investigate frequency discrimination, we used two different methods (anaesthetic or compression block) to block out "local" mechanoreceptors right at the spot where the vibration was applied.

Nice one Will!

Ah, sorry for slow reply! I'm not familiar with this part of R. Did you solve it? Have you tried posting on stackoverflow?

What's the code?

A black and white cat looking at a black and white world.

More snow. At Linköping central station and at Valla campus, Linköping University.

Personally as a native English speaker I don't think it adds much to my writing, but I know people who say it helps them a lot. It makes sense to me, seeing as the output it produces is nicely enough written on a sentence level.

A lot of misunderstanding about what it can do. It's a language model, not an encyclopaedia. It's most useful and defensible to use it as something like a spelling or grammar checker, improving flow and readability. Can be great for ESL, dyslexic folks, etc.

The brief version is going to be sth like: to compare effects you have to *assume* that sth (responses, variances, whatever) is commensurable and that may lead to any sort of effect size calculation, but the default of just standardizing almost seems designed to hide those assumptions.