Hi David,
Most OCR programs try their best to produce the most accurate
scan given the capabilities of the receiving program. In Word
this means it may try to reproduce every variation that it
perceives as a difference (in fonts or other attributes).
Unfortunately this usually works out poorly unless you're using
the most pristine originals.
To get "plain text" it's sometimes helpful to scan into Notepad
rather than Word, but this often doesn't give you text that's
easy to fix. Better, usually, is to scan it into Word and
immediately select the entire output document and press CTRL+Q
(clears all paragraph formatting), then CTRL+Spacebar (clears
all character formatting).
The more complicated your original, the more work you'll face in
cleaning up the finished scan to make it presentable. Buying a
more expensive OCR program can improve this a little (sometimes
only *very* little).
