I'm not sure what you mean exactly. When you need to interact is generally pretty logical, it's for something that contains other elements. So you'll want to interact when you see a table, a tool bar, a scroll area, web content or a group. There's also a small sound effect whenever you focus something that can be interacted with. Press VO+H and go through the sounds help menu and you'll find it. Whenever you hear that sound, you'll know you might want to interact with an item.

Lastly, in the navigation section of VoiceOver utility you have an option called grouping behavior that lets you turn off interacting for a lot of controls, with some exceptions like tables. The default option is standard, which requires interaction, but you can change it to one of the other options. That being said I'd personally get used to the concept as it actually speeds up navigation in the system quite a bit. Also, you can still tab around the interface like on Windows, you won't always want to do this but for example if you're in a dialog box to open a file or in some apps like Skype, tabbing is even faster than just manually moving VoiceOver and interacting.