More info for anyone looking into this:
I have one radio question at the beginning of my experiment, which is the one above.
At the end of the experiment, I have a few radio questions in a loop. The first time I see each of them, their labels are misaligned, e.g.:
(after more testing, I’ve found that this happens even when not in fullscreen)
The alignment will adjust and become correct if I resize the window:
And apparently psychoPy remembers this, because in the second iteration of the loop, the labels are correct without having to resize the window:
I have noticed that this sometimes happens with buttons as well, and text, so maybe it’s not only a slider problem. The issue always resolves itself if I resize the window AFTER a routine has started, but not if it is resized before/during the onset of a routine.