I like this idea in general (I have an Android app that does the same thing as your page, except it uses draggable sliders).
Here are some things I believe you could improve about the implementation:
- There's no explanation behind the math of weighing that's done. "these scores were calculated based on your ratings for each choice and the weight you gave to each factor" isn't an explanation. "(e.g., "super important" has a weight of 4 and "not that important" has a weight of 1)" doesn't tell me why my choice of one "super important" and one "not that important" gives me the score of 14. Not knowing the math you're doing makes me not trust the result. I mean, I suspect you're doing "sum of factors multiplied by weights", but a non-math-savvy person may take a while to figure where the numbers come from.
- Too many pages. I know you're aiming for dead simple (technically) UI, but I think it would be better if I could enter choices and factors on the same page.
- I'd consider changing the factor score UI from:
- ( ) 4 (best)
- ( ) 3
- (x) 2
- ( ) 1 (worst)
into something inline, like:
4 - best: ( ) ( ) (x) ( ) :1 - worst
People are used into this into (that's how you rate stuff on most pages on the Internet), and it will cut your use of vertical space by at least a factor of 4, and also make it more readable - with this, it would be easier to compare between factors and across options.
--
Quite honestly, a perfect incarnation of this idea to me would be a spreadsheet. This problem domain naturally fits into a shape like this:
A spreadsheet interface would make this tool much more convenient, and offers plenty of opportunities for futzing with UI (color coding, sliders for input, whatnot).
Something I think you should consider for version 2.0 or as an alternative to test out. Hell, I think I'll make myself one in Excel right now.
I'm pretty good at UX and I don't have any issues with the current experience. It actually really helped me for my use case.
I wouldn't use a spreadsheet. While I clicked 'How was this calculated?', I realised I didn't care.
Your mistake is thinking the purpose of the app is to give the user the perfect objective recommendation. For most decisions, you just want some kind of psychological justification - for some people that can even be a coin flip.
The "objective recommendation" and "just psychological justification" groups seem to me a vastly different target audiences. For the latter, this tool is already overcomplicated; it could be done as a list of input fields and a button that highlights one of them at random.
I'm glad the tool helped in your case. I can imagine it helping in mine once, but me getting immediately frustrated. In my own philosophy of UX, I generally don't care much about tools you're only going to use once in your life; I focus on ones that you're going to use repeatedly. Repeated use has different priorities - in particular, efficiency over hand holding.
Here are some things I believe you could improve about the implementation:
- There's no explanation behind the math of weighing that's done. "these scores were calculated based on your ratings for each choice and the weight you gave to each factor" isn't an explanation. "(e.g., "super important" has a weight of 4 and "not that important" has a weight of 1)" doesn't tell me why my choice of one "super important" and one "not that important" gives me the score of 14. Not knowing the math you're doing makes me not trust the result. I mean, I suspect you're doing "sum of factors multiplied by weights", but a non-math-savvy person may take a while to figure where the numbers come from.
- Too many pages. I know you're aiming for dead simple (technically) UI, but I think it would be better if I could enter choices and factors on the same page.
- I'd consider changing the factor score UI from:
into something inline, like: People are used into this into (that's how you rate stuff on most pages on the Internet), and it will cut your use of vertical space by at least a factor of 4, and also make it more readable - with this, it would be easier to compare between factors and across options.--
Quite honestly, a perfect incarnation of this idea to me would be a spreadsheet. This problem domain naturally fits into a shape like this:
A spreadsheet interface would make this tool much more convenient, and offers plenty of opportunities for futzing with UI (color coding, sliders for input, whatnot).Something I think you should consider for version 2.0 or as an alternative to test out. Hell, I think I'll make myself one in Excel right now.