Formatting can be impacted by different settings. This post might help.
That said, Scrivener is not a WYSIWYG word processor. Whatever settings (font, spacing, size, indents, etc) you use when writing in Scrivener can be non-destructively overwritten when you compile.
Scrivener can only do whatever a user tells it to do, so if a user gives it instructions to output multiple files, it will output multiple files. If a user wants everything as one document (as many people do), the user just needs to instruct Scrivener to compile (not export) the project into a single Word file, RTF file, ebook, PDF, etc.
This video might help:
literatureandlatte.com/video … eLarge.mov
In essence, each Scrivener project is a database. Users can fill their databases with whatever text they want. They can tag, label, reorder, and manipulate that text while it is in the database in an infinite number of ways. When they want to, they can output all, part, or parts of their database in a variety of different formats, without impacting on the structure and content held in the database itself.
For me, I think of Scrivener as being a digital brain. And just like any brain, it can store a ton of data, reorder and refine that data in zillions of different ways, and then it can present that data in a variety of different formats. My organic brain collates ideas, works them into a cohesive whole, and then outputs them as novels, plays, emails, letters, oral stories, films, forum posts, etc. Scrivener can do much the same, collating data and then outputting the right data in the right single or multiple format that the user wants.
With a word processor, you are limited and hamstrung by the structure, content, and look of the document that you see on the screen.
With Scrivener, you are omnipotent. A project is a collection of cells and DNA. You get to construct and reconstruct those cells and that DNA in whatever way your imagination conceives.
Some additional videos here…