Some criteria of a good Ph.D. dissertation have been suggested by my friend Dr. Tian Hong. These are the original criteria and translations:
* A candidate's dissertation or published work must represent a contribution to knowledge, showing evidence of originality and independent critical power.
Translation: Make sure your thesis contains something that your committee members have never heard before, ideally something they can't even understand!
* A candidate must also satisfy the Examiners, by the dissertation or published work, and in the viva voce examination, that he or she is well acquainted with the general field of knowledge to which the subject relates.
Translation: Make sure your committee members like your ideas! And try to pretend you know almost everything in your field!
* Dissertations must make a significant original contribution in the particular field of learning within which the subject falls and their literary form must be satisfactory.
Translation: You should have a beautiful hardcover Ph.D. thesis to show your parents and friends. However, if the folks don't understand what you've done, you should at least convince them you really pushed science forward in YOUR FIELD!