On Jun 5, 2008, at 4:21 PM, Wendy Goodfriend wrote:
Can anyone direct me to research discussing the advantages and disadvantages of using one vs two email fields? I am also looking for inline validation code that addresses the one vs two field issue as well.
I haven't seen anything formally published. However, here's what we've found in our research at UIE:
The problem you're trying to solve is mistyping email addresses. Depending on the audience, context, and design, you can see typos in anywhere from 0.75% to 5% of email addresses entered. (Even here at UIE, we have, on average, 2 out of every 100 email addresses are entered incorrectly. These are designers and developers with a lot of internet experience, so it's not just a matter of sophistication.)
Several sites try solve the problem by asking for the email twice. The thinking is that, if the user enters it the same twice, then it must be correct. As people have discussed, that doesn't always happen because more sophisticated users will use cut & paste, which will only propagate a typo in the second field, making a false positive.
In my opinion, one of the best examples of the 2-field verification is at Fire Eagle. (It's currently invite only and, for reasons I can't explain, I received an invite, so I put a copy of the page here: http://tinyurl.com/5cjaqs ) In this case, it's the reason in the copy that helps people understand why they are being asked twice.
One problem with the 2-field verification is that the typo isn't always restricted to the first entry. If the user types it correctly in the first field, but incorrectly in the second field, the validation fails. However, the user *had* typed it accurately the first time and will often feel frustrated that it was the verification that failed. This doesn't improve the user's experience.
So, the alternative is a one-field entry. Here, we've found the best way to reduce problems is to make sure the input field and verification display is large. Both the length of the field and the size of the font should be large enough so people can discern characters. For example, does an "rn" combination look similar to an "m" character? Does a "1" look distinct from an "l"? With the right font, size, and field length, the user should spot typos easier. How much does "[EMAIL PROTECTED]" look like "[EMAIL PROTECTED]"?
Most typos are doubled characters or "neighbor" keys (a "j" when the user meant a "k", for example). Again, a large, clear verification display will help a lot. Designers, when showing users what they've typed, often don't ask themselves, "Is this display going to help a user spot a subtle typo they could otherwise miss?" Looking at the verification display for that quality can help reduce errors tremendously.
For algorithms, there's limitations to what you can do. You can check for spaces -- they're illegal. You can look at the domain portion and have an validation check that it contains an MX record. You can ensure there's only one @. Beyond that, there's not much else.
Hope that helps, Jared Jared M. Spool User Interface Engineering 510 Turnpike St., Suite 102, North Andover, MA 01845 e: [EMAIL PROTECTED] p: +1 978 327 5561 http://uie.com Blog: http://uie.com/brainsparks ________________________________________________________________ Welcome to the Interaction Design Association (IxDA)! To post to this list ....... [EMAIL PROTECTED] Unsubscribe ................ http://www.ixda.org/unsubscribe List Guidelines ............ http://www.ixda.org/guidelines List Help .................. http://www.ixda.org/help
