demographics

Leah Guren on demographic differences in using Help content

Another great session from WritersUA. Leah Guren has been conducting usability tests on populations using software help systems to perform simple tasks. In this study, she found that language, experience, and age affect task performance.

Some of her conclusions:

Gender made no difference - women look for Help content, or ignore Help content, as much as men.
Experience can make a big difference. It takes 8 good successful sessions with help to make up for one bad experience where it didn't help. -- think indexes?
Age is a huge factor, and native tongue as well. These are both modified by experience, i.e., an older user who was very experienced with using other help systems were just fine, but older newbie users were not. Those working in a second language who again had more experience were just fine, but it was a barrier if the user was a newbie.
58% of users distrust Help - due to bad experiences.
Small onscreen text that appears in a bar above or below the window is not seen at all.
Side panels also disappear.

When readers look at a table of contents online that contains a ton of hierarchical structures, they don't use it. When asked why, they said that none of the headings seemed to help. Leah said that to a user, their question is first level important. We can assume that they want it up at the main heading level in a toc, or it is useless to them. So again, think indexes, we are the great levelers for people who think their search is important enough to be available on a first level basis.

I wish she had tested indexes, but oh well.

Links inside a paragraph lose users completely.
Vocabulary was a huge issue - again, there we can help teach users the domain language.