Image Reorderer protocol ready for review...

[Paul Zablosky]

Hi Daphne,
    I tried to perform the tasks in the Round 1 protocol, and I must say 
I completely failed at task 2.  Well, not completely, but it took me 
many minutes to figure out how to perform it reliably.  I'm sure that no 
tester would have given me enough time.

If I have the fruit images in two rows, it is really easy to move any of 
the second row images to the centre of the first.  If there are seven 
images in Row 1, I simply select any second row image and place the 
target after Row 1, Image 3.  It doesn't matter how the avatar is 
positioned -- if the target is to the right of Image 3, my selection 
drops into the middle.  I also notice that the target is a good 
indicator of which images the one being moved will fall between.  That 
is, if the target is between the blackberry and cherry, that's where the 
one I'm moving ends up -- between the blackberry and the cherry.

So far, so good.  My success at moving images from Row 2 to Row 1 is so 
confidence-inspiring that I decide to move an image from Row 1 into Row 
2. Should sort of work the same, shouldn't it?  (Now I know that it 
won't  quite be the same, because I know that I'm really operating on a 
one dimensional list, not a grid.  So things will rearrange themselves 
to fill gaps, but I let my sense of having learned something in the 
first trial carry over.)

Now what happens? Well first of all, I find that the position of the 
target causes rather different behaviour.  If  I place the avatar over 
the image currently in the centre of Row 2, it doesn't seem to matter 
which side of it the target is on.  The current centre image moves to 
the left and the one I'm moving takes the centre position.  So, I sort 
of know how to get my image into the centre, but I'm totally confused 
about how to get my image /between/ two others.  The "between-ness" rule 
I had inferred from the previous trial doesn't work any more. 

So I experiment a bit an suddenly find that things aren't dropping where 
I expect.  I'm totally confused until I notice that the relative 
positions of the avatar and the target are important with this kind of 
move (Row 1 to Row 2).  If the centre of the avatar is a bit to the left 
of the target, the image ends up on the left side, and if it's a little 
bit to the right, the image ends up on the right side.  The rule I now 
infer is "the image my avatar is hovering over will scoot to the left, 
and my image will replace it --  the position of the target doesn't 
really matter.  This is a lot different from "a gap will open up where 
the target is now, and my image will go in between".

So I end up with two rules:

   1. When I move things /up/, the position of the *target* tells me
      where they will fall.
   2. When I move things /down /the position of the *avatar *tells me
      where they will fall.

I'd be embarrassed to tell you how long it took me to figure this out.  
I hope your test subjects are able to catch on a bit quicker than me.

One other thing I noticed which you may want to control for while 
testing. If you resize the window so that the rows have an even number 
of images, the "middle" is less well-defined than if you have an odd 
number..

Regards,
Paul





Daphne Ogle wrote:
> http://wiki.fluidproject.org/display/fluid/Image+Reorderer+User+Testing+-+Round+1 
>
>
> Any feedback would be greatly appreciated!
>
> Daphne Ogle
> Senior Interaction Designer
> University of California, Berkeley
> Educational Technology Services
> daphne at media.berkeley.edu
> cell (510)847-0308
>
>
>
> _______________________________________________________
> fluid-work mailing list - fluid-work at fluidproject.org
> To unsubscribe, change settings or access archives,
> see http://fluidproject.org/mailman/listinfo/fluid-work

-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://fluidproject.org/pipermail/fluid-work/attachments/20090325/c6cedf24/attachment.html>