You may want “other” included for political beliefs, possibly even including a text box.
For analysis? No, not at all. Forcing people to pick from five bad options is way better than letting them say whatever they want. I seem to recall previous versions having a text box where people could write whatever else they wanted about politics, which was then ignored, and I think that’s a good solution.
I’m interested in what LWers are doing or have done to improve their lives, and how it’s working out for them, but perhaps this should be a separate post or a different survey.
Something like “how many changes to your life have you attempted in the last year?” might be interesting, but the trouble there is one person might think a change smaller than “moving to another city” isn’t worth mentioning, while another person might count that they switched brands of oatmeal. That could be somewhat informative, insofar as that number gives an idea of how fluid people see themselves as.
So non-answers are used to identify multiple-choice questions which lack adequate choice. How would you identify questions that were skipped for a reason other than “There was no right answer.”?
Hm? In the politics question, I’m pretty sure you’re forced to pick one of the five.
Actually, that reminds me of a potentially better way to measure politics: score voting. Have people give each of the five options a score from 1 to 10 based on their identification- then you can release both unnormed scores (people gave libertarianism an average rating of 3) and normed scores (when people’s scores were expanded so their score range was 1 to 10 regardless of the range they supplied, libertarianism had an average rating of 5).
This will make it possible to identify people who really believe in one of them (by giving it a 10) and people who think that all are bad, but maybe one of them is the least bad.
Can forms with radio buttons be submitted without selecting any? I hadn’t noticed that. And anyway, if you select one by accident, in all browsers I remember using there’s no easy way to unselect it without selecting another.
(Yes, by “easy” I meant ‘without having to fill out the survey again’. Which is a helluva trivial inconvenience IMO—I think I’d rather just pick the least bad answer rather than starting over.)
For analysis? No, not at all. Forcing people to pick from five bad options is way better than letting them say whatever they want. I seem to recall previous versions having a text box where people could write whatever else they wanted about politics, which was then ignored, and I think that’s a good solution.
Something like “how many changes to your life have you attempted in the last year?” might be interesting, but the trouble there is one person might think a change smaller than “moving to another city” isn’t worth mentioning, while another person might count that they switched brands of oatmeal. That could be somewhat informative, insofar as that number gives an idea of how fluid people see themselves as.
So non-answers are used to identify multiple-choice questions which lack adequate choice. How would you identify questions that were skipped for a reason other than “There was no right answer.”?
Hm? In the politics question, I’m pretty sure you’re forced to pick one of the five.
Actually, that reminds me of a potentially better way to measure politics: score voting. Have people give each of the five options a score from 1 to 10 based on their identification- then you can release both unnormed scores (people gave libertarianism an average rating of 3) and normed scores (when people’s scores were expanded so their score range was 1 to 10 regardless of the range they supplied, libertarianism had an average rating of 5).
This will make it possible to identify people who really believe in one of them (by giving it a 10) and people who think that all are bad, but maybe one of them is the least bad.
I don’t think there is a default option- one can simply not select any answer.
But the point of the open text field for responses is that it provides input regarding what should be included on the next survey.
Can forms with radio buttons be submitted without selecting any? I hadn’t noticed that. And anyway, if you select one by accident, in all browsers I remember using there’s no easy way to unselect it without selecting another.
Reload the page; ‘easy’ might vary between people and how much of the survey has already been filled out.
(Yes, by “easy” I meant ‘without having to fill out the survey again’. Which is a helluva trivial inconvenience IMO—I think I’d rather just pick the least bad answer rather than starting over.)