Timeline for Adding a "none of the above" option to a list of checkboxes
Current License: CC BY-SA 3.0
11 events
| when toggle format | what | by | license | comment | |
|---|---|---|---|---|---|
| Nov 14, 2014 at 14:05 | comment | added | dennislees | @Jules a) as an idea, this is far from really bad. It's actually proved to be a valuable thinking tool. b) Assuming the worst from the user, the accidental clicking of anything is a risk in any situation. This design would be bad if it increased the chances of accidental clicking, and it doesn't. 3) The problem of misinterpretation is a risk specific to the words "None of the above" and is not increased in this design. 4) It's clear from the comments on the top answer that we've moved past this suggestion. I don't think any answer here get's it right, so please feel free to add to the mix. | |
| Nov 14, 2014 at 13:26 | comment | added | Nathan Rabe | The answer from rewobs feels like it is making things more complicated for the user in order to comply with various UX guidelines, when making the user get it right is the main guideline that matters. Yes, making a checkbox perform an action is against the rules, but it will matter so rarely that users may never know it happens. Yes, users may accidentally check the wrong box, but that's always true for any data entry. | |
| Nov 14, 2014 at 7:08 | comment | added | Jules | I think this is a really bad idea. My first concern is that a user may click the "none of the above" choice by accident and not notice that it has erased their entries, causing them to submit an incorrect form accidentally. My second is that a user may misinterpret "none of the above" as "I have a condition which isn't listed above". Systems that actively prevent them from trying to check NOTA along with other conditions or which make it visually different avoid this potential problem, and the addition of a single extra radio button is hardly a huge amount of extra UI to avoid this. | |
| Nov 14, 2014 at 2:05 | comment | added | Nicholas Pappas | While users would figure it out, I don't agree that a checkbox should actually perform an action. Checkboxes allow users to choose between two opposite states, actions, or values - but should not perform said action. | |
| Nov 13, 2014 at 16:12 | history | edited | dennislees | CC BY-SA 3.0 |
added 266 characters in body
|
| Nov 13, 2014 at 15:20 | comment | added | dennislees | @Izhaki Your comment about invisible logic making things "far more complex" is barely applicable to this scenario, and is mildly ironic given your promotion of a solution that to the end user is undeniably more complex. | |
| Nov 13, 2014 at 15:16 | comment | added | dennislees | @Izhaki I'd argue that this solution is more optimal than the other one we're commenting on above, primarily because it gets the job done with much less UI - Your point about "putting users in front of wireframes" isn't helpful because we know end users aren't fully conscious actors assessing every detail in the context of their mental model, especially checking boxes on an insurance form - The only non-conventional thing about this design is that a checkbox also has a clearing function. Is that enough of an issue to cause the user's mental model to crumble? Hardly (cont.) | |
| Nov 13, 2014 at 11:55 | comment | added | Izhaki | But we shall also account for the fact people are intelligent and given the context, the option "none of the above" implies that choosing it will disable all other choices. But is this the optimal design? I'd argue it isn't. You should also consider errors and recoverability - invisible logic makes the whole story far more complex. | |
| Nov 13, 2014 at 11:53 | comment | added | Izhaki | I disagree with this view. You should consider mental models and visibility. Checkboxes, by convention are boolean controls and can be composed into a multi-selection group. That's what users expect. Any 'behind the scenes' logic is invisible to users. That is, if you put a user in front of a wireframe and ask them how this would behave (their expectations - which we all build while interpreting an interface) they could easily be confused. The controls imply one thing, the labels - another. | |
| Nov 13, 2014 at 11:39 | comment | added | André | I, too, think this is a good solution. Visually, I would make sure to set the NOTA option slightly apart from the other options though. | |
| Nov 13, 2014 at 3:17 | history | answered | dennislees | CC BY-SA 3.0 |