Usability Tips Based on Research Studies | Hacker News Reader

Usability Tips Based on Research Studies

sixrevisions.com

128 pointsjrwoodruff22 comments

This is the only sort of UX, interaction design and usability "top ten" list that should ever receive upvotes: one that cites references and only publishes information that has been validated in some fashion.

In several of the author's examples - like vertical attention - different studies show different results - or as new measurement tools become available, old truisms fall.

I am all for documentation / proof in UX but at the same time the problem with vetting out ideas, or common sense "this worked for me " illustrations may keep you from testing that one morsel that puts your ux / conversion rate over the top.

The article was very interesting and I was also interested in the "Removing a Button" article they linked to: http://www.uie.com/articles/three_hund_million_button/

I recently bought something from the website of a company that does that exact same thing wrong and made me hesitate a long while before purchasing. I sent the article to them and hope they change their setup, I really want them to succeed.

Re: #5.

The referenced "studies" show that users will scroll if they realize they need to access their desired content.

Okay.

That absolutely does not mean that vertical pages provide pleasing UX or good usability. It simply means the user endured the vertical layout in that case.

None of the provided citations support that conclusion of the article. Not only that - sample size, obvious confirmation bias, and extreme extrapolation of data to unsupported conclusions render the "studies" relatively useless.

In the last citation, even though they attempt to wash it away with statistical significance (SD on a sample of 15?) their data actually shows that usability, as defined by comprehension, is maximized by paging.

I think the truth is found closer to the notion that form follows function and the answer to scroll vs. paging is content and site specific.

I am finding the article, overall, unimpressive. The method involved seems to be that of looking for support of one's opinion, not extraction of good practice from objective research.

The words-read-per-page graph is presented as a curve; looks to me more like folks read up to 200 words, then start dropping out. The initial plunge is linear in other words, equal to the number of words on the page (when there are less than 200).

Can anyone provide examples or references that discuss the importance (or unimportance) of site speed for sites that aren't search engines?

http://www.zdnet.com/blog/btl/googles-marissa-mayer-speed-wi... When the Google Maps home page was put on a diet, shrunk from 100K to about 70K to 80K, traffic was up 10 percent the first week and in the following three weeks, 25 percent more, she said.

I thought the three-click rule died in the 90s. Krug debunked it in 2000 in Don't Make Me Think.

I was surprised that being above the fold didn't matter as much as I thought it did.

I assume the F-shape pattern is only effective in languages that read left-to-right?

I wonder how much the sites they used biased the results. Google's search results are very F-shaped, for example, and so is any left-to-right language text that is divided into paragraphs with headings. Does the "use an F-shaped pattern" recomendation actually boil down to "display text the way ypu normally would in your language"? If so, that's not nearly as interesting.

Well for navigation, the F-shaped or "Top-Left" pattern has been a staple of site design for at least 11 years.

In 1999, when i started making sites, the sort of nav was considered "boring" because everyone was using it, and our designers often created weird (poorly performing) nav just to buck the trend.

I had the same thought actually. I think what might be more interesting, or at least as interesting, is the flow of how the user looked at the page. What did they look at first, second, etc.

The fact that the bottom right corner is barely ever looked at directly is hardly surprising, but seeing how the user processes the page could be interesting, especially with some layouts that are more visual than than search results.

Yeah I think this point has a parallel in #6 referring to placing important content on the left of a web page...

"...for sites that are traditionally read from left to right, placing important design components at the left is a good idea; vice versa for sites whose language is read from right to left."

Number 6 had me imagining an extreme test case. All the content of a page is in a narrow vertical column over at the right of a page. Show me a heatmap of that test page with real users still spending most all of their time looking at the big blank space at the left and... well I'll eat one of these pencils on my desk I guess, as I have no hat I wish to part with.

That quote got me thinking, though, about the question they did not answer:

What about languages that are read top-to-bottom first?

I have no clue what she based it on, but my fourth grade teacher demanded that her students' booths at the school science fair be positioned so that they were to your left when you walked into the gymnasium. "People look to their left first and they'll look at your projects first."

Did she have any basis for this?

Not as far as I know. (Western language speaking) People do tend to read from left to right (and up-to-down), but there's no evidence to suggest that this applies to non-linguistic things, especially as basic as the way they look when they enter a room.

I had the pleasure of some industrial ergonomics/design courses in college, and IMHO web designers would be wise to look at the decades of research that have gone into that field. Much of it remains relevant even in software/web UI.

there's no evidence to suggest that this applies to non-linguistic things

That seemed odd to me just going off (less than scientific) bits and pieces I've read over the years, especially relating to retail design where a popular adage is that "people tend to walk to the right upon entering a store" (so that's where you put the more expensive goodies).

But a cliché isn't evidence. So I thought I'd look for stronger citations. And man, it's tough. I haven't found anything on the retail thing yet but there seems to be evidence of preferences for one direction over another in studies. For example: http://bps-research-digest.blogspot.com/2006/12/satisfaction...

Reading direction appears to have an influence on the ease with which items are detected in different hemispheres: http://pss.sagepub.com/content/16/1/15.abstract (and sort of related: http://pss.sagepub.com/content/18/6/487.abstract)

It's possible that adults with US driver's licenses, having been trained to look left first, then right, when crossing traffic, may do so in pedestrian situations as well (such as entering a crowded gymnasium).

I almost got run over by a truck once after looking the wrong way last when crossing a street in Hong Kong. People do have systematic habits shaped by their culture as to looking left first or looking right first, but maybe that is most influenced, as you suggest, by which side of the street drivers drive on.

Maybe. But the Asian languages all tend to be displayed left-to-right on the web. I've only seen Arabic (or Farsi?) script written right-to-left online.

Reply on news.ycombinator.com