do users understand the terms "ascending" and "descending"?

28 Apr 2010 - 5:12pm
3 years ago
11 replies
695 reads
Christopher Rider
2009

Hi

My team have been discussing how to word the labels presented to non-sighted users for sortable table headers.

The terms "ascending" and "descending" are popping up like gophers. My instinct is that parsing these terms requires a lot of work on the users' part, so I'd prefer to use labels like  "A-Z", "0-9" and "newest first". Naturally, I'm getting some pushback from my development team, since this requires a different label depending on the data type of the column.

Can anyone point to some concrete research on whether users understand these terms? Failing that, informed opinion and wild-eyed rants are welcome as well.

--

chris rider

Comments

28 Apr 2010 - 5:19pm
Christopher Rider
2009

I should add that the userbase for this application covers a very wide swath of the population.

29 Apr 2010 - 6:59am
laura olac
2007

Hi Chris

I cannot point you to any research on these exact two labels you are asking about. However, based on over 10 years of experience working as a user experience designer I would not recommend using the words "ascending" and "descending" especially if you are dealing with a wide variety of people using the product. The terms can be confusing for many people and they definitely require a moment of thinking. It is a lot more intuitive to use the labels you thought of (A-Z - type label) which our brain can easily recognize and understand with no extra effort. Good luck making the case to your development team! Don't give up and preferably don't spend time conducting expensive usability testing only to convince others in the company that your design expertise makes sense. If need be and if you want to be confident with your decision, put together a quick prototype and conduct some quick usability with people from your company, but outside of the engineering team.

29 Apr 2010 - 3:50am
Bowen Hendy
2009

Well as wild eyed rants go,

Perhaps some user testing might be in order?

29 Apr 2010 - 6:26am
Jakub Andrzejewski
2010

Hi,

This may or may not help since the test was carried out in Poland on a polish website. We had 12 people test a website where there was a list of products you could sort (by: name ascending, name descending, price ascending, price descending).

The task they wanted to perform was to find the cheapest product on the list.

Observation:all of the users hesitated. Most of them chose the option they didn't want.

29 Apr 2010 - 8:14am
Ania Powers
2010

Hi,

Why not to use terms users are already familiar with (because of real estate websites and online stores), and try something self-explanatory like "highest first" vs "lowest first" ? The simpler the better (not to mention words "descending" and "ascending" may be not understood by users with first language other than English).

29 Apr 2010 - 8:18am
Ania Powers
2010

edit: for some reason, it was sent twice, sorry.

29 Apr 2010 - 11:44am
Ian Roberts
2004

I have to agree with the others, not from research but only from personal experience. I'm a pretty educated guy, and I still don't know which is which when I have to sort a column. My usual approach is then to click it once and see if it's right, then click again if it's wrong Smile

29 Apr 2010 - 8:30pm
Alan James Salmoni
2008

I encountered this recently: even I got confused about which was which after a while, particularly if non alphanumerics are in the fields. You are right - something like "lowest first" or "A-Z" would work better (given my experience).

 

Like Ian said, it's possible to get around it with trial and error but page refreshes are so boring - and a destructive sort can be damaging to users. Is it possible for you to print out an example table on paper, go onto the street, grab passers-by and ask them what they would expect to see if they did an ascending sort? Should take you an hour to get convincing information - just a count of errors should be enough but try to record what people say when they're shown the correct solution.

 

30 Apr 2010 - 9:50am
bkillam
2010

I agree. Some users do and some don't understand ascending. And even those who do often show signs of hesitation to translate that term into its effect. More design direct references like "Lowest to highest" for price seems to work better for everyone.

Bill

Bill Killam, MA CHFP President, User-Centered Design, Inc. 20548 Deerwatch Place Ashburn, VA 20147 703-729-0998 (Office) 703-626-6318 (Mobile) http://www.user-centereddesign.com

On Apr 29, 2010, at 9:48 PM, Alan James Salmoni wrote:

> I encountered this recently: even I got confused about which was which after a while, particularly if non alphanumerics are in the fields. You are right - something like "lowest first" or "A-Z" would work better (given my experience). > > Like Ian said, it's possible to get around it with trial and error but page refreshes are so boring - and a destructive sort can be damaging to users. Is it possible for you to print out an example table on paper, go onto the street, grab passers-by and ask them what they would expect to see if they did an ascending sort? Should take you an hour to get convincing information - just a count of errors should be enough but try to record what people say when they're shown the correct solution. > >

5 May 2010 - 12:01am
cfmdesigns
2004

My biggest annoyance with this is the directional triangle which often accompanies the sort to indicate the direction of the sort.

When sort order is "ascending", so that more recent dates are at the bottom, the arrowhead is usually made to point up. While an up arrow might translate to "going up" (ascending), it's pointing in the opposite direction of the sorted list. One of those cases where we adhere to the "standard" despite its flaws or confusion points.

-- Jim

On Apr 29, 2010, at 7:23 PM, Alan James Salmoni wrote:

> I encountered this recently: even I got confused about which was which after a while, particularly if non alphanumerics are in the fields. You are right - something like "lowest first" or "A-Z" would work better (given my experience). > >

29 Apr 2010 - 10:24pm
bjminihan
2010

One of the biggest problems with this behavior is not what the label says, but what it does.  Without carefully reading the table contents, it's hard to know whether a button that says "A-Z" will sort that column from A-Z or represents that the column is already sorted from A-Z.  If you have two buttons on every column, side by side, with A-Z and Z-A, then you've corrected for that problem (but probably have too many buttons =]).

I've had to solve this problem at least three times.  Inevitably, your version works best with users, even if it takes a little more development time (sorry developers, I've had to program this, too, in Ruby, .NET C# and Java).  The last time, I found (through observation and feedback) that you only need 2 types:  A-Z/Z-A and 0-9/9-0 - the first for all strings, and the second can cover both numbers and dates, since people generally had no problem seeing that 0-9 would sort a date column by earliest first (strange, but true).

Also keep in mind you're talking about a very narrow use case:  non-sighted users who want to sort every column in a table.  Unless sorting is CRITICAL to your interface (probably shouldn't be), you're probably talking about a fraction of your audience.  That might explain the push-back you're getting from the dev team.

Good luck =]

Bryan

Syndicate content Get the feed