My understanding is that a Quick Format will create empty new spaces for the Root Directory and the File Allocation Table (or whatever relevant structures that NTFS uses) where sector allocation is tracked. However, it does absolutely nothing to the rest of the disk. A full format, however, tests every disk sector to decide whether it can work properly and marks any that fail so it will not be used for data. What I do not understand, though, is how that test is done. It is conceivable that it simply reads the data in the sector, calculates its checksum and compares it to the check from the sector. This presumes that the existing checksum data is correct. Other third-party utilities, I know, will do a test in which the current data is saved to RAM, a new data pattern is written to the sector and then read back, and the result is checked against what is known. Then the original data is written back to the sector. (Actually, I believe this is done on a multi-sector or full-track basis to save time, but the result is the same). Of course, the intermediate possibility is that you could write a known pattern to the sector, read it back and check it. Never mind preserving and restoring the original contents when you're Formatting a new disk, anyway. Among these three methods, I can't find a reference to say what Windows does.
EXCEPT - NOTE this from M$ Knowledge Base.
http://support.microsoft.com/kb/941961/en-us
Apparently in VISTA the new procedure in Full Format does actually write zero's to every byte of every sector. I guess that would allow them to test the sectors by reading and verifying that the data coming back out is all zero's, although the article does not say that specifically. But I suppose, too, that since this is billed as "new" behavior in VISTA, no such operation was done in previous Windows.