On Sat 07 Sep 2013 11:53:00 AM EDT, S.Kemter wrote:
Now there are some voices, that say doing a second page if you have
selected to much isnt the best way.
Jenneh and I talked about this in IRC a couple of weeks ago and I had
thought we were in agreement that the two-step process is really not the
best way. The core problem with the two-step method is that you're
setting the user up to enter into an error condition. The max number of
votes is not enforced in the selection mechanism, so you're setting a
trap for the user to select more wallpapers than you allow, and then
basically telling them on the second screen that they are wrong.
Avoiding placing the user into an error condition is one of those key
heuristics of interface design you don't ever want to violate.
Plus, having a 'vote' button on the first screen makes it seem as if
you're entering in your final vote. Some people might hit vote, then go
somewhere else, and not realize until their session has timed out that
their vote didn't actually go through. (You could probably address this
by relabeling the button, "Review Selections" or something like this.
But that's confusing too. I made my selections, why are you forcing me
to review them?
Having a selection preview in the same page as your voting eliminates
this extraneous step and also visually makes clear the user's state
(votes left, no votes left).
Having the second step is also inconsistent in that if you select
exactly 16 wallpapers, you'll never hit that screen, right? If the point
is to give people a place to further refine over their selections, that
needs to be advertised much better otherwise there are users who will
never hit it and get that benefit. (This would then re-spin the 'error
condition' into being a 'feature')
Well I am more in favor of it as I
can do in this way sorting first out all the stuff that isnt really
good and then select the ones I really think that are the best.
Thinking thats the best way with a huge amount of submissions, we will
have this time I think something around 100 submissions but with
Nuancier running for the next one, I think the submissions will
definitely increase as a lot of people have no problems anymore with
submitting to the wiki and need no account anymore.
Now, we could try to do something most of us can agree or we can use
what we have now and try what the users think about and change it
later as we concentrate right now only on the voting and all the
others things will be written later.
So what you think sticking with it like it is now or try to improve it?
I think no matter which design we ultimately choose, the state it's in
now definitely needs refinement.
At the very least, if it's to stay using the same two-step model, I think:
- The "Vote" button on the first screen needs to be relabeled. I suggest
"Finalize Selections" or "Finalize Vote." Or even "Step 2:
- The first page should say something like, "Click on as many of the
wallpapers as you like below. You'll finalize your selections in the
next step." And if the user happens to pick 16 exactly, how would you
treat step 2? I don't even know... because then they're putting in a
real vote, not finalizing their selection. Maybe the text of the button
changes depending on the state? So if you pick exactly 16 it says "Vote"
but if you pick more than 16 it says "Refine Selections" ?? But buttons
changing labels is confusing too...
It seems like the problem you're wanting to be solved is: within a
massive number of images, how can you pick within your limit when you
like more wallpapers than the limit allows and you're not sure which
should make it?
The thing is, the mockup Ryan did solves this problem - your selections
are always viewable on the screen. If they aren't when you scroll, they
should be - the pane should be fixed. It's basically like having the
second page viewable on the first page.
The other thing in Ryan's design which makes the selection task easier
is that he mocked it up with only 5 selections allowed. Limiting the
wallpaper selection to 5 means people are only going to pick those
wallpapers they are the very most excited about, and it's going to be
easier for them to pick them because they don't have as many slots to
worry about. You might worry that letting people pick 5 means we won't
have enough wallpapers voted for to make 16 for the package, but if we
have > 100 wallpapers to choose from, the probability that our users
would pick the same 5 over and over is really low. The more wallpapers,
the lower the probability. 16 selections is a pretty high cognitive load.