This is not such a good idea assuming your text is reasonably readable for your audience (e.g., it’s not their second language, they’re not kids just starting to read, and you use decent font style, size, and color contrast). There’s evidence that simultaneously presenting the same text verbally and visually increases cognitive load and degrades learning. See:
Kalyuga S, Chandler P, & Sweller J (2004) When redundant on-screen
text in multimedia technical instruction can interfere with learning (pdf).
Human Factors 46, 568-581.
The problem seems to be that the two text presentations interfere with one another, forcing the user to mentally coordinate them, which detracts resources from understanding the material.
Simultaneous visual and voice work best when they compliment each other, rather than repeat each other. For example, you could graphically represent what the voice is saying, such as showing Jack and Jill sitting next to each other to show they’re friends, and highlighting the links in question when the voice gets to “Click to find out what they do together.”). Or you could provide non-verbal acoustic accompaniment to written text, like maybe playing “friendly” music when introducing Jack and Jill, and sound an inviting chime when showing “Click below to find out what they do together” in order to signal it’s time for the user to interact. I’m guessing the former would work better in your case.
Of course, if you’re really concerned about holding user interest, you probably should work more on the content, not the format. A flashing format for boring content just comes across as ridiculous and I believe can itself distract from what the user really needs to learn. Make the content personally relevant to the user. Make it clear why it’s important to learn this stuff (what’s in it for them?). Make it come alive as a story. Add dramatic tension and conflict. Make them care and be curious about the content and characters.
Jack and Jill insist they are just close friends, and not lovers. But
they sure do a lot together. They’re always:
And you probably already know that you need to be careful with sound since it can be annoying in certain contexts. You may want to give the user the option of replacing voice with written text.