Slashdot Mirror
Do Zebra Stripes Actually Help?
RyoShin writes "A List Apart, an excellent resource for web development and related aesthetics, has put together an article based on original research by Jessica Enders into 'zebra striping.' From the article: 'Zebra striping [coloring alternate rows] is used when data is presented in an essentially tabular form. The user of that table will be looking for one or more data points. Their aim is to get the right points and get them as quickly as possible. Therefore, if we set a task that uses a table, and zebra striping does make things easier, then we would expect to see improvements in two things: accuracy and speed.' The conclusion of the peer reviewed paper? It's a wash. Striped tables offered only a slight increase in accuracy and speed overall. The article notes a few other benefits to using Zebra striping, so it's all up to the individual."
Although it might not provide much extra accuracy, it does make for a nicer looking GUI. That counts for something in todays widget driven environment...
Yes, much in the same way that Go Faster Stripes work...
Summation 2
Finding data in nine columns with alternating text and numbers is easy. Try upping the number of columns, using only numbers, use close spacing, and reduce the text size. Then you will see a difference. This experiment is flawed because they didn't test how the values scale with more columns and less helpful clues (like the differences between text and spacing in their sample table). This article should have been rejected for not taking into account these issues.
I don't have a study to prove it, but coloring or otherwise marking every Nth row, where n is a smallish number, say 2-5, helps.
Anyone else remember fanfold wide-format computer paper that was colored white and green in alternating blocks of 3 rows each?
Knowledge is how to play a game, intelligence is how to win, wisdom is knowing what game to play.
It depends on the program displaying the data. Some programs allow you to to click on the row and get that one row highlighted. That is a huge help. Others like tables on a web page don't allow that. In that case I say it does help.
Also the size of the table makes a difference.
See my blog http://ilovecookes.blogspot.com/ for light hearted technical information.
is irrelevant.
On narrow tables they don't make a difference. But on wide tables they're almost a necessity. Without any table cell borders, like a spreadsheet, or striping, the eye easily wanders up or down into another row when reading across. I can say anecdotally that I'm far more accurate and faster when reading a table with stripes.
Either way, they certainly can't hurt, especially if they're a pale color. So why are we even having this discussion?
Developers: We can use your help.
So they conclude it doesn't help (though their own data says that it does, even though it's slight) based on THAT table? Maybe they should try it again with a zebra striped table where the difference between the colors used is slightly more pronounced. I don't know about the rest here but I personally think I had a harder time with that because the color difference between rows was so slight than if they had left out the color. Played tricks on my eyes.
If zebra striping doesn't actually make it easier to identify which cells actually belong to a given row - maybe a rethink of what is trying to be accomplished here could help. Perhaps highlighting of the row under the cursor?
sloth jr
Study conducted of whether long-held belief has real benefits. Conclusion: Maybe a little.
News at 11.
Spelling mistakes, grammatical errors, and stupid comments are intentional.
That's why I still print out web pages on greenbar before reading them.
Cretin - a powerful and flexible CD reencoder
I can tell you why it didn't help. They formatted their table with large spaces between columns and they had only 9 rows. If they tried the same study where they also varied the number of rows I am sure they would find that as the rows increase the positive effect of zebra striping increases. It seems they had a bias built into their test in order to find something unexpected... otherwise the study would have proved pointless.
I can see the Slashdot headline now, "A practice used for over half a century still proves to be useful!" Somehow, I think such a headline falls under the category of "not news."
I see a problem with the experiment. The hard part of the questions involves scanning down a column , where horizontal striping obviously does not help.
There may be an effect, but if so, it's small enough that 281 experimental subjects and six questions are not enough to yield statistically significant results. That result alone (that the effect is small at best) makes the paper worthwhile to me. One small quibble: on a web page, I can often use scrolling and the bottom or top of the page to check alignment on a wide table. Maybe zebra stripes are more useful on paper.
But before I give up entirely on zebra stripes, I'd like to see what happens when [1] the table is made wider; [2] the table is made taller; [3] the zebra stripes are 2 or 3 rows wide instead of 1; [4] the stripes are made darker and/or a different color.
C'mon people who want publications, there are lots of other things to try here.
--- Often in error; never in doubt!
Nystagmus is a condition where your eyes oscillate at a frequency of about 1 Hz (roughly), usually horizontally. Having rows and especially columns coloured differently helps very much for someone affected by Nystagmus, to distinguish between columns.
BTW, a wider font like Verdana is also highly recommended.
"The agriculture ministry is not in charge of Gundam" - Japanese ministry official.
I'll get the wtf out of the way first
"Given that applying zebra striping in an electronic medium is a nontrivial task"
Say what? Any application that is based on columns and/or rows has trivial access to those columns/rows as separate entities. Markup for such columns/rows is easily changed. 'mod N 2 == 0? grey:white' is hardly nontrivial, it's so basic that if you can't manage to do it, you must be using the wrong software.
---
Now for the scope - it seems like the only research they have done is when data in the sheet is dense and the sheet itself is not all that wide.
Now try with a wide sheet and instead of every 'cell' or at least one of its close neighbors having data in it, imagine lots of empty cells. Now try and see if zebrastriping helps or not. I can guarantee you that without any visual cues, your lining up of something in the leftmost column to the same line on the rightmost column is going to fail far more often than you'd like.
--
Oh wait, they even admit as much:
"However, there is clearly a need for additional studies to investigate how task difficulty and the size of the table/form influence the effect of zebra striping."
No shit. I'm glad you admitted that your sample size is too low.
I thought this article was going to be about the evolutionary advantage that stripes give zebras, not spreadsheets. Who cares about spreadsheets, what about the zebras?
They also help the pack slice their dinner up fairly. For very lazy lions there are now even zebras with perforated easy-claw edges down every third black stripe.
The stripes helped my Zuba's look cool.
On printed pages, I've seen people using a ruler to help scan through tables of numbers. I thought that was where the idea for zebra striping came from. Honestly, I'm surprised that it was only a minor improvement. Maybe it's just me and my bad eyes, but I think it helps tremendously. It probably also matters how many columns there are -- the more there are, the more it helps. I'd also suspect that fatigue would kick in, so it would make less difference for, say, less than 100 exercises, and more difference after that. [Didn't RTFA. Maybe they address those points.]
I'm no GUI designer, but when I make utility web pages that use tables, I tend to use either zebra striping or a tr:hover that uses a light yellow to highlight the line under the mouse pointer. That way, if I feel I need the help to track through the table, I just run the mouse down the columns and it lines up the current row for me very nicely. IMHO, this is a nice compromise where zebra striping might not look good, but the user might want the help nonetheless.
Perhaps I'm being a bit pedantic, but am I the only one that thought some of the questions were oddly worded?
Here's what I view as correct answers:
Q What is the name of the screw that costs $35.66?
A: None. The M28 screw costs $35.66 per 50.
Q There are 664 screws of which minor diameter tolerance?
A: None. The M18 Screw has a minor diameter tolerance of 8g, and there are 664 of those, but there are 1442 screws with a tolerance of 8g.
Q: There are 292 screws of what thread pitch?
A: None. There are 292 M16 screws which have a thread pitch of 2mm, but there are 527 screws with a thread pitch of 2mm.
In addition to that excellent point, I'm skeptical about the way the table was designed. There's an image of the table here:
http://alistapart.com/d/zebrastripingdoesithelp/data-table.png
The "ordinary" rows have a background color of pure white. The "striped" rows have a background color of #F5F5F5, a very light grey. I'm all in favor of subtle design, but there is such a thing as being too subtle.
Perhaps the stripes did not help noticeably because there was insufficient contrast between the rows?
When the participants violate the very precepts of the study by creating their own striping, the study become ridiculous. It's like doing a study if walking is slower than biking, and the walkers are allowed to bring their own bikes.
But the study itself is great, I just disagree with the conclusion, it seems to show that striping is SO useful that when denied striping, people create their own.
What's up with this box everyone has to think inside of or outside of? Why does there have to be a box?
If you didn't like how they ran this test (and I agree with most of the complaints), A List Apart is running a follow-up survey over the internet. I almost missed this, as it isn't mentioned until the very end of the article.
You can find it at http://surveys.formulate.com.au/dtfu. It takes about three-five minutes. I just took it, and they appear to be using darker row colors now. It's still too narrow to see how useful it is when you have to scroll horizontally, but it's a small improvement at least.
Heh, my family keeps making fun of me because when I was teaching my daughter to safely cross the street at intersections, I told her to look at the "green guy". Apparently although the traffic light is green, the little guy is white - who knew?
Also, I can never move to Texas. The rest of the country has vertical traffic lights, but not Texas. It is bad enough to choose red and green as the stop and go colors, with the full knowledge that a large percentage of the population can't see them - it is a travesty to take that and put it on its side so that up vs down cannot be used.
And before you get to the left vs right, some of us are dyslexic...
while (sig==sig) sig=!sig;
Apart from the just plain wrong statistical reasoning, the experiment was done under uncontrolled conditions over the Internet. The sample table in the article actually had lines separating the columns and rows. Geez, with that, it's not surprising that the author finds no differences!
Zebra striping may or may not help significantly, but this paper won't tell you either way.