Hilary Palmén

Thanks to a nudge from a colleague I've just found this wonderful research paper by Sophie Beier and Kevin Larson in 2010. It evaluates different design characteristics of commonly confused (difficult to identify) letters using multiple different identification tests. the goal being to help design more easily identifiable letters. fabulous!

https://www.researchgate.net/publication/233608849_Design_Improvements_for_Frequently_Misrecognized_Letters. It was uploaded to researchGate (Open access) by Sophie Beier. Awesome!

https://www.researchgate.net/publication/233608849_Design_Improvements_for_Frequently_Misrecognized_Letters

#LetterIdentification

Mar 21, 2023 at 2:39pmWeb
Show threadDavid JONESMar 21, 2023

@Hilary One of the few workers in this area doing reliable work in my opinion. Her book "Type Tricks" is fun too.

I was going to complain that it wasn't Open Access because on Research Gate i clicked on it and it asked me to login. But by the time i had tabbed back here, it had also spawned a page with 503 error on it, and... actually downloaded the PDF. so, lol.

Show threadHilary PalménMar 21, 2023
@drj yes, everything I've read by Sophie is normally strong from a research perspective (my lense) and from a typographical perspective, a rare double strength. Kevin Larson is also a very strong research partner, they've got my instant trust as individuals, collectively they really rock the space!
Show threadHilary PalménMar 21, 2023
@drj its so long since I signed into research gate, that I'd literally forgotten it was required - thanks for calling this out!
Show threadTiro TypeworksMar 21, 2023
@Hilary Chuckling about the insight that x-height letters benefit from not being x-height letters.
Show threadHilary PalménMar 26, 2023

@TiroTypeworks I guess that means the x height can be a larger proportion. Stumpy ascenderd and descender.

I wonder if the range of x height to descender/ascender proportions' impact on readability has been studied. With Roboto Flex, its possible to study it...

Show threadTiro TypeworksMar 27, 2023
@Hilary I don’t think that’s what Beier and Larson meant. If you compare the proportions of the letters used in the tests, it really looks like the third model of ‘a’ is taller than the other x-height letters in that group, and the ‘s’ in the same group is descending.