How to Choose a Thesis Topic (according to David Steinberg)

From the author of ""How to complete and survive a doctoral dissertation".

* Excitement about the topic not that big of a factor in success,
since all dissertations run into similar sets of difficulties
* Essential questions:

Is it researchable?

* "One must be virtually certain that the data for the dissertation
will be available and accessible when the candidate comes around to the
collection phase of his project."

Especially timing is important, eg staff turnover rescinding access to a
population

Does it make a contribution to the field?

* "dissertation-as-too-little" or "dissertation-as-magnum-opus"
syndromes (magnus opus will come in 20 years)
* Just have to think & write about it, as well as talk to everyone

Is it original?

* What you should be looking for is daylight: after studying the
literature, and scope and ambition of other recent theses, "do I find a
hole, a gap, a missing link that my topic can contribute to plugging,
bridging or forging?"
* Again, talk to all faculty, like a travelling salesman
* Don't worry very much about being "scooped"
* Topics rarely come by chance for real: "topics almost always come
out of extensive soaking in the literature and prior research of one's
field, and are traceable consequences of that immersion"
* Expansion of a master's thesis from a dissertation not really an
easier option
* Never pick a topic with the assumption that a particular faculty
member is going to be there all the way through! It must stand on its own
* As much as possible, discuss dissertations that appeal to you with
their authors. They're almost always willing to share their time and
dissertation experience, esp within the first couple of years (often
noticeably more helpful than faculty)
* Be cautious of opportunities to use spin-off data - what if the
grant is cancelled or the faculty member leaves?