interaction design - Does any research/study exist on what makes an interactive document better from a user experience perspective?

I am aware that there is a lot of research into page layout methodologies and the flow of design elements (text, diagrams, etc) such as the Gutenberg Diagram, Z-Pattern, F-Pattern, and so on. I'm also aware of other Stack Exchange questions (e.g. this recent one: Page layout - authoritative research into natural flow of text, tables and images).

However, this question is about interactive documents (typically interactive PDFs, but not necessarily so).

I have designed many of these over the years and after coming across the question above, got to thinking that maybe there is some respected research into what makes an interactive document much more usable from a reader's point of view.

Typically (though not always) my interactive PDFs are designed so that they're almost like an entire website within a PDF (i.e. they have a menu structure and make use of images, colour, white space etc). Users can browse and jump around seamlessly within the document as required and these seem to work very well. I've used this approach for technical documents, newsletters, interactive diagrams, help guides and so on and these seem to come across very well, but I'm always looking to continually improve my work.

For example, one area I haven't delved too much into is the use of javascript in my interactive PDFs. I have used this on two occasions with positive results, but I'm not sure if there are any dos and don'ts on using javascript.

Also, I wonder about the compatibility of interactive PDFs with screen reader software for the visually impaired?

And of course, with the explosion of mobile platforms, there's the whole issue of compatibility to ensure that an interactive document displays and functions as intended from Mac/Windows desktops to iOS/Windows Mobile/Android devices.

Obviously, there's quite a bit to this topic, hence why my question is asking for references to any research (if it exists) on best practice for creating interactive documents. My hope is that if any research does exist, some or most of these issues will be covered.

Answer

I would narrow down the problem definition - documents are supposed to convey information. And to quote from Bret Victor (http://worrydream.com/MagicInk/#p155):

Information software [...] mimics the experience of reading, not working. It is used for achieving an understanding—constructing a model within the mind. [...] For information software, all interaction is essentially navigation around a data space.

So this is a problem how to design navigation, not interactivity in general:

Other kinds of "interactive documents" are usually called applications, tools, utilities or games, so I wouldn't expect to find much useful info on the topic of "user experience of interactive documents", but a lot of research on the topics of: