Would it be helpful to suggest that this redundancy can also be a selling point? Podcasting is starting to gain popularity.

Maybe explain the narration with screen text is only useful in certain situations: job aids, hearing impaired, low literacy, etc.

Regardless of the efficacy of text over-load, it seems to me that one's aural learners would benefit greatly from narration. It's lonely enough to sit through asynchronous e-learning modules - hearing a voice read the text could be beneficial (assuming one has the option to silence said voice if one chooses to do so).

In a perfect world we would not need to narrate on-screen text, but it can be helpful for a variety of reasons due to the technical limitations reading text on a computer screen.

On-screen text has a low DPI (96) compared to about 300 dpi in a newspaper. This makes reading more difficult. The refresh rates of computers can also give people who read onscreen text for too long a headache.

Although referring to the cognitive theory is a good place to start when considering the use of narration, it's not the only theory worth considering. Technical limitations should also help guide choices when keeping the learner in mind. Additionally, user habits of skipping large blocks of text may be reduced if they have the option to listen rather than read. (At 3:30 PM, after lunch reading can feel a heck of a lot harder than listening.)

My preference is user controlled audio.

quote:

In a perfect world we would not need to narrate on-screen text

In a perfect world, there would be hardly any to narrate. :-)

quote:

My preference is user controlled audio.

...and the user needs to be able to control the playback speed. I find that 200% of normal (100% increase) is much easier to listen to than normal speed. I have friends who prefer to listen at up to 400% of normal -- I'm working up to that.

If it seems strange to you, try it: you won't want to go back to normal speed.