3 ms·
About 8% of all males are color-blind. If you are a scientist, and want to publish your work, there's a one in four chance that at least one of your three revie
by Derbasti 8y ago
About 8% of all males are color-blind. If you are a scientist, and want to publish your work, there's a one in four chance that at least one of your three reviewers is color-blind.
Rainbow color schemes are ridiculously bad. One study showed that medical professionals who look at rainbow-themed plots every day, immediately made fewer errors if presented with a simple grayscale plot instead, despite not having any experience with the grayscale color scheme.
Add to that that human perception is much better at seeing high-frequency lightness contrast than color contrast. Color contrasts are better-suited for categorical data.
- nbsd4life 8y ago> there's a one in four chance Closer to 1 in 7, because the reviewer might be female (who are rarely color blind).
- bradknowles 8y agoCloser to 1 in 4, because sadly there are very few reviewers who are female.