4 ms·

Read the second link on my post. A menu bar is one or two word that are easy to read. The recommendation on those posts are for long texts that need sustain att

Read the second link on my post. A menu bar is one or two word that are easy to read. The recommendation on those posts are for long texts that need sustain attention to be read. The second article shows data of readers reading speed based on font size.

medstrom 3y ago
Thank you. Good data, that people read fastest at a text size of 0.3 arc. And I'm taking away that it might be fine for paragraph text to be slightly larger than menu bar text. But a couple of factors jump out at me: 1 As they said, there's a lot of variability in how far people sit, which is a good time to remember the myth of the average[1]. You set default values to suit the "average" reader, but your efforts might serve a much smaller amount of people than you'd think -- because almost no one is "average". It's good to shoot for sane defaults, of course, but it could distract designers from strategies that would reliably serve all users. To start with, an actual safe assumption is that whoever the reader is, they've made sure that 16 CSS pixels is readable to them (even if it doesn't meet the ideal of 0.3 arc). They might have upscaled the whole OS UI or just sat closer -- either way, 16 CSS pixels still represents a size that's readable to that person. Going too far above this requires making some pretty bold assumptions which risks making users feel they're being treated like idiots or children. 2 Without sensors that measure physical distance-from-screen, I don't know how we would calculate the visual arcs anyway. We can only hope the user is average. And even if we had such sensors, there are two problems: (1) I'd prefer that iOS/Android just manage the system-wide zoom level, and then individual webpages need do nothing; and (2) it may be great and all to try for 0.3 arc as a default, but if I've scaled it up or down for any reason, I probably want that choice respected, which brings us back to point 1: you can safely assume that the user has adjusted their sitting distance to meet the arc-value they prefer! Yes, they could be idiots and wrong about what would be best for them, but they will still sit closer or zoom out, so you won't get the arc-value you were shooting for anyway! The more you upscale, the more the user downscales. So why fight? Why break expectations? 3 I don't know which font size represents 0.3 arc my screen, but I'll assume it's about 18 CSS pixels and I often experience that smaller text and tighter line-heights let me digest paragraphs and go through whole documents more quickly. The lower limit is very low, like 9px. I think studies need to split up what it means to "read fast": If you're part of the majority who read one word at a time and subvocalize every word in your head, then what's ideal is probably very different from what's ideal for a speed-reader. In my experience, my reading speed scales to some extent with the total amount of text visible on the screen! So I want a lot of text on screen. That's why I like the default HN tiny font too. Notably, line-heights above 1.6 absolutely destroy my reading speed and I have to switch to Reader mode or just close the tab. My own website has 1.3. I guess the same principles apply to reading carefully vs. "scanning"/"skimming". Every user cares about that. I think the people who designed the websites I hate probably just assumed that the user is reading everything carefully, but that makes the website a nightmare for users who are just trying to navigate and find their way around. ----- [1] Excerpt from https://www.gse.harvard.edu/ideas/ed-magazine/15/08/beyond-average https://www.gse.harvard.edu/ideas/ed-magazine/15/08/beyond-a...: > The Air Force concluded that Americans had gotten bigger over the past couple of decades and they simply needed to update their measurement of the average pilot. With the help of a young Harvard College graduate named Lieutenant Gilbert Daniels, they measured more than 4,000 pilots on 10 dimensions of size that seemed important for fitting into a cockpit — torso length and chest circumference, for example. The thinking was that once they redesigned the cockpit for the average pilot of the 1950s, controlling the plane would no longer be so troublesome. Most pilots, they assumed, would be within the average range on the majority of dimensions and that a good number would even be average on all 10 dimensions. > “Do you know how many really were?” Rose says. “Zero.” > Even when just three dimensions of size were picked, fewer than 3.5 percent of the pilots fell within the range defined by Daniels as average. Instead, what Daniels found is that every single pilot had what Rose calls a jagged profile. One pilot with long arms may also have long legs while another may not. Not everyone who was average height (5 feet 9 inches) had the same chest circumference or head size. > Finally, the Air Force had its “aha” moment: If every pilot had a jagged profile and the cockpit was designed for the average pilot, it was actually designed for no one.

medstrom 3y ago

Thank you. Good data, that people read fastest at a text size of 0.3 arc. And I'm taking away that it might be fine for paragraph text to be slightly larger than menu bar text. But a couple of factors jump out at me: 1 As they said, there's a lot of variability in how far people sit, which is a good time to remember the myth of the average[1]. You set default values to suit the "average" reader, but your efforts might serve a much smaller amount of people than you'd think -- because almost no one is "average". It's good to shoot for sane defaults, of course, but it could distract designers from strategies that would reliably serve all users. To start with, an actual safe assumption is that whoever the reader is, they've made sure that 16 CSS pixels is readable to them (even if it doesn't meet the ideal of 0.3 arc). They might have upscaled the whole OS UI or just sat closer -- either way, 16 CSS pixels still represents a size that's readable to that person. Going too far above this requires making some pretty bold assumptions which risks making users feel they're being treated like idiots or children. 2 Without sensors that measure physical distance-from-screen, I don't know how we would calculate the visual arcs anyway. We can only hope the user is average. And even if we had such sensors, there are two problems: (1) I'd prefer that iOS/Android just manage the system-wide zoom level, and then individual webpages need do nothing; and (2) it may be great and all to try for 0.3 arc as a default, but if I've scaled it up or down for any reason, I probably want that choice respected, which brings us back to point 1: you can safely assume that the user has adjusted their sitting distance to meet the arc-value they prefer! Yes, they could be idiots and wrong about what would be best for them, but they will still sit closer or zoom out, so you won't get the arc-value you were shooting for anyway! The more you upscale, the more the user downscales. So why fight? Why break expectations? 3 I don't know which font size represents 0.3 arc my screen, but I'll assume it's about 18 CSS pixels and I often experience that smaller text and tighter line-heights let me digest paragraphs and go through whole documents more quickly. The lower limit is very low, like 9px. I think studies need to split up what it means to "read fast": If you're part of the majority who read one word at a time and subvocalize every word in your head, then what's ideal is probably very different from what's ideal for a speed-reader. In my experience, my reading speed scales to some extent with the total amount of text visible on the screen! So I want a lot of text on screen. That's why I like the default HN tiny font too. Notably, line-heights above 1.6 absolutely destroy my reading speed and I have to switch to Reader mode or just close the tab. My own website has 1.3. I guess the same principles apply to reading carefully vs. "scanning"/"skimming". Every user cares about that. I think the people who designed the websites I hate probably just assumed that the user is reading everything carefully, but that makes the website a nightmare for users who are just trying to navigate and find their way around. ----- [1] Excerpt from https://www.gse.harvard.edu/ideas/ed-magazine/15/08/beyond-average https://www.gse.harvard.edu/ideas/ed-magazine/15/08/beyond-a...: > The Air Force concluded that Americans had gotten bigger over the past couple of decades and they simply needed to update their measurement of the average pilot. With the help of a young Harvard College graduate named Lieutenant Gilbert Daniels, they measured more than 4,000 pilots on 10 dimensions of size that seemed important for fitting into a cockpit — torso length and chest circumference, for example. The thinking was that once they redesigned the cockpit for the average pilot of the 1950s, controlling the plane would no longer be so troublesome. Most pilots, they assumed, would be within the average range on the majority of dimensions and that a good number would even be average on all 10 dimensions. > “Do you know how many really were?” Rose says. “Zero.” > Even when just three dimensions of size were picked, fewer than 3.5 percent of the pilots fell within the range defined by Daniels as average. Instead, what Daniels found is that every single pilot had what Rose calls a jagged profile. One pilot with long arms may also have long legs while another may not. Not everyone who was average height (5 feet 9 inches) had the same chest circumference or head size. > Finally, the Air Force had its “aha” moment: If every pilot had a jagged profile and the cockpit was designed for the average pilot, it was actually designed for no one.