We’ve had some internal debates and looked at the feedback from our users over the default behavior of our Multiple Choice field (or radio inputs to you Web Developers out there). Right now, when you make a radio group (called Multiple Choice in the builder), and do not select a default option, we automatically make the first option the default for you. One of the main reasons we do this is because of the W3C:
At all times, exactly one of the radio buttons in a set is checked. If none of the
<INPUT>elements of a set of radio buttons specifies `CHECKED’, then the user agent must check the first radio button of the set initially. View
In addition to the W3C, various user interface books we have come across stress the importance of making one option selected by default. The “proper” way to make a radio group would be similar to below:
What is your favorite day?
- New Years
- 4th of July
- Thanksgiving
- None of the above
In that example, “None of the above” would be set to default. You can also use “Other”, or if the data is numerical you can do “Greater than” or “Less than.” The reasoning behind this is that it always gives the user an out in case they accidentally clicked one of the selections.
Now, even though the way Wufoo currently handles radio buttons is “right,” we’re not so sure it makes things easier for our users. For example, consider a multiple choice test you take in school. Most of the time, there is no “All of the above” answer. The student is left to pick one of four choices, and none of those choices are pre selected. In fact, a pre selected choice may even place bias on that choice.
We wanted to open up our internal discussions, and get your feedback. Should we automatically select a default radio button for you, or should the radio group load with no radio button selected?
You can follow comments on this entry by subscribing to the RSS feed.
Poll is broken:
No report was found at this URL!
But my vote is ‘No’
Posted 2 years ago by Anon.And the results pie chart display is broken in Opera 9. (I voted “no” as well and got the same error.)
Posted 2 years ago by Lynne.Ditto - “No report was found…” I voted “yes”, because doing so is compliant, and I’m a good worker ant. However, authors should be urged to choose their own default and to always provide an “out” option (“none of the above”, “other”, etc). Otherwise, switch to a set of check boxes where no selected option really means no selected option.
Posted 2 years ago by Chris Bloom.I vote “yes.”
Posted 2 years ago by Reggie.Sorry everyone. Had it linking to the wrong report url. Should be good now.
Posted 2 years ago by Kevin Hale.I would say no, but perhaps you should let turn on the feature of automatically setting a default radio button, form-wide. That way users who always want the option will be able to streamline the creation of forms with a single setting.
Posted 2 years ago by Joe Lencioni.No, leave radio buttons all unchecked.
If you create a form where some elements are required and others are not, then checking a radio button on a non-required element by default may send the wrong information.
For example, if I build a form where 1-3 are required and 4 isn’t, and 4 is a radio group (Multiple Choice), then having one of the options checked when the form is submitted, and the user hasn’t “answered” that question will make his submission incorrect.
Posted 2 years ago by Ryan M.Ryan, but what if the user accidentally clicked the radio button. He now has no way of unchecking it. The idea is that by setting default options it forces the form creator to make options that catch all cases.
It is a tricky situation though, and I see where your coming from.
Posted 2 years ago by Ryan Campbell.I think this is an occassion where the spec is wrong.
By having the behaviour of selecting a choice without user intervention (no matter how that is achieved, through mark-up, or by the user agent, or whatever) you have stepped on the toes of the user. You’ve made an assumption and spoken for the user, and that behaviour is like putting words in their mouth.
The spec writers have made the assumption that the web-master always wants populated form data, and that is a poor assumption. They effectivly make a radio set a mandatory field, which is not the job of HTML, and not the behaviour of any other form element either. Making fields mandatory is a job for your form processing script, not HTML.
Admitedly, a well chosen set of options will cater for the possibility that the user doesn’t actually want to answer the question, but to do that you need to provide an option that says ‘I dont want to choose an option’. And that’s a double negative in terms of logic, as well as a pain in the ass for the user.
Posted 2 years ago by Matt Wilcox.@ Ryan Campbell
True, but on the other hand, the only way to account for your suggestion is to include about 19 misc options (hyberbole is my specialty): “I choose not to answer”, “I forgot to answer”, “Not relevant to me”, etc…
Is it possible to create a radio group where, when submitted, it appears to the recipient as if it were not answered, even though one button is checked by default? You’d have to have a “Pick a radio button” button as the default checked, which is really lame.
Any other suggestions?
Posted 2 years ago by Ryan M.@ Matt Wilcox
Exactly.
Posted 2 years ago by Ryan M.For me, the question has always been—should we allow form designers to create forms that go go against a guideline? Do we have a responsibility to guide designers to build forms that are valid or is it their responsibility to follow the specs, our job only to provide the opportunity to do so or not?
Being a spec addict, I was inclined to think responsibility fell on us at the beginning, but as I saw the possibilities it limited and lack of alternatives available … I’m starting to side on increasing choice rather than decreasing options.
I definitely see both sides of this argument and its why we’re putting it up for our users to help us decide. One of the things I’m also starting to notice is that form design is one of those areas in web development where there isn’t a lot of detailed discussion hashing out these gray areas. Hopefully, we’ll be able to flesh out some of these ideas as it becomes easier and easier for people to create and build forms.
Posted 2 years ago by Kevin Hale.Now that Wufoo is a business charging customers for a service, the focus from a business perspective should be let the customer do what they want and specs be damned. Many cusomers will prefer the look of radio over selects, for example, even though both could be used similarly. Customers don’t care about specs. They care about getting what they want for thier money. If they want the look and feel of radio buttons and don’t want to have them be mandatory, give them what they want. Give them the option to place a “reset” button if accidental clicks are going to be troublesome.
just my half nickel…
Posted 2 years ago by Seth F..Every form with non-mandatory fields should have an “Unspecified” radio button (styled differently; in italics for instance) that allows a user to revert to an… unspecified… option
Posted 2 years ago by Kit.I’d have to agree with Matt W and Ryan M - pre-selecting an answer, even if it is “I don’t have an answer”, just feels wrong. It’s like those registration forms that pre-tick “I want to receive email newsletters” for you - let me make my own decision.
Posted 2 years ago by Matthew Pennell.In terms of asking survey type questions I think it is best to leave it unchecked. If you have something pre-filled, then I think that would tempt some participants to skip the question since validation requires no effort on their part (I know I’d skip in some circumstances). In this case I think the spec is wrong.
However, for Web Apps I think it is best to check the safe or default option initially, e.g. “Are you sure you want to delete this file?” [ ] Yes [.] No
Posted 2 years ago by Adrian D..I agree with the W3C position of having a default selected.
If having “none” selected is a meaningful option, it must be made one of the explicit choices, so that its meaning can be selected and deselected in the same way as the other choices. Especially in a situation where the user has selected an option, then changes their mind and wants to have nothing selected. With no “none” choice available they would need to reload the form.
If Wufoo doesn’t enforce this, perhaps it could strongly suggest it in some way.
Posted 2 years ago by Brook Elgie.Interesting.
Makes me think HTML radio buttons are baddly specified. The W3C should permit none to be selected and require that browser vendors provide some means to allow the user to revert to none selected after changing their mind about a selection.
Posted 2 years ago by David.I think maybe you need to re-read the spec. It states pretty clearly that UA behavior is undefined in these cases:
The bit you quoted above is from RFC 1866, and reappears in the HTML 2.0 spec, but a lot of things have changed since then ;)
Posted 2 years ago by kevin c smith.What about automatically adding a “None of the above” and making that the default if none is chosen? I hate the idea of not having any way to unchoose if one chooses by accident, and this way you avoid the ‘putting words in the users mouth’ issue.
Posted 2 years ago by Estebes.If we have group of radiobuttons in options, one radiobutton (it need be the first) need to be checked. BUT! Using radiobuttons for voting (for example) is fail. Radiobuttons have concrete behaviour and we shouldn’t change it (because in this case radiobuttons will have in voting one behaviour and in options another).
In controls like voting we need simply clickable answers no radiobuttons. To vote user need to click once, not select radiobutton and push dummy button “Vote!”.
I consider voting as example.
Posted 2 years ago by Tony.I would prefer to at least have the option as many times I get asked by my boss/client/collegue/etc to do something that goes against standards and best practices. I pick my battles with them and always recommend the best way I think it should be done. They usually listen but sometimes there isn’t enough time to explain or develop a better way (in their opinion). Time is usually money.
To have a tool further encroach on my ability to ‘break’ a standard is a deal breaker for me. Let me decide how I want to break my W3C compliance….there are far worse ways my clients could force me to on the site this form will appear on and as long as they are paying, I’ll do what they ask but will give my reasonings for doing it the ‘right way’.
I’ve seen plenty of examples with other web based apps that give a strong warning if you do something that goes against a best practice or standard…until it’s the law, it’s bend/breakable and should be an option.
Your tool should give us the choice, plain and simple or I’ll find another one which would be sad because I like this one.
Posted 2 years ago by John.Hrrm hello … I tried to unselect the selected radiobutton (by modifying the html), “just to see”, because of the content of the entry, and, to my surprise, it worked. I then tried to put some random value in the value attribute of a radio, because I am very curious >_>, and the value was put in the results.
Is it a normal behaviour ? =X
Posted 2 years ago by numerius negidus.Everyone needs a hug.
Posted 1 year ago by Victor Bachenko.I agree with Brooke that “With no “none†choice available they would need to reload the form.”
Posted 7 months ago by Tej Kohli.