Re: Screen captures

[jimmy -at- breck-mckye -dot- com]

Screen captures are only useful if they help the user find an element 
in a UI that they are unlikely to locate themselves by providing them a 
visual context, and a cue to look for when they use the application for 
real. They are not helpful when they show things the user would locate 
trivially, or don't provide enough context (my favourite example was a 
manual which illustrated the step "Click OK" with an image of a sole, 
lonely OK button, cropped such that even if the user couldn't find it on 
the UI, the image still wouldn't help them locate it).

Therefore, if you're looking for facts on when you should and shouldn't 
use screenshots, look for research on the kinds of scenarios in which 
users fail to locate UI elements. From my own experience as a UX 
designer, some common cases for this are:

- When an element is not exposed upfront, e.g. it is only provided on a 
hover state, or as part of a dialog panel that would not be reached 
naturally;
- When an element is placed below the fold, and the UI does not 
adequately indicate that content sits below the scrollpoint;
- When an element of very low visual weight is positioned above an 
element of extremely high visual weight;
- When multiple elements are densely clustered;
- When findings of usability tests indicate that it's been a problem 
before!

This is an over-simplification, but it serves to illustrate the point. 
Think about circumstances where it's actually useful to show an image, 
and think about the ways that image *actually* helps when you frame 
them.

If you wish to validate whether images are necessary, you should try to 
validate it with user testing rather than surveys. Observe users 
actually using the interface with the document at hand, and observe 
their use of the content. Simply asking users 'could images be useful' 
won't get you good data, because respondents _always_ say yes, no matter 
the reality (after all, who says 'no, this extra content could never be 
useful?').

I like the idea of thumbnails, though - effectively, you're not showing 
the images (and incurring the disruption in layout and whitespace) until 
a user actually wants a cue on how to find something. And showing the 
image in a modal dialog means you can display it at full size without 
incurring the constraints of your document layout.

If it becomes a toss-up, I prefer to avoid images. They create 
innumerable layout problems and large rivers of whitespace. They turn 
simple, one-page procedures into unwieldy, multi-pages messes. They add 
a significant overhead to maintenance and need to be duplicated for 
every OS, platform and regionalized release.


On 2013-02-10 17:13, Anonymous wrote: