I have experienced and seen and heard from MMED candidates that
starting & completing the research component of the MMED program is
always a struggle. I have done some searching and came up with some
helpful tips that I think will be beneficial to others.

The tips were compiled from a seminar done in the US. Here they are:

*Originality is a principal criterion of a good
topic. You can be original in diverse ways. You
may examine material that has never been studied
before; or you can examine well-known material,
but provide new interpretation.

* Another way to view these different concepts of
originality is to recognize that some topics are
central to the field and that there is always new
work being done; other topics are on the
periphery and have been neglected.

* It is important to choose a topic that is
congenial to you, that you think is worthwhile
not only within the framework of the discipline,
but on a personal level. It is not all irrelevant
to consider how much you like interviewing,
computers, dealing with insects-or whatever it is
that a topic demands.

* The specific topic that you study may have a
personal and idiosyncratic origin. It is no
accident that research on certain groups is
likely to be pioneered by people of that group:
women have often led the way in women's history,
Blacks in Black history, immigrants in the
history of immigration.

* You should have a doable thesis that has
boundaries; you have to be able at least to
imagine where and when it would end. It if hard
to start a thesis, it can be even harder to end
one.

* This means that you should be ambitious
intellectually, but not too ambitious, think of
it as a task that will enable you to get on with
your career. Students sometimes ask if their
dissertation should include A, B, C, and D after
the dissertation is finished.

* One speaker put this idea in a different way.
He suggested that instead of writing a
dissertation prospectus it is best simply to
write a dissertation chapter. He explained that
what he really meant was that it is best to do a
little piece of research think small. If it is
interesting it will lead to a bigger problem. The
best proposal is a pilot project; once you have
picked a path you can add on different forks as
you go along. He observed that everyone knows the
BIG IDEAS, it is harder to do the little ones.

* Modesty is also helpful in choosing a
manageable topic. Some students set out to write
a dissertation that will change the world; others
just want to write a dissertation. In terms of
results, there seems to be no correlation between
the quality of the dissertation and the ambitious
nature of the topic.

* They noted that it is useful to make the
dissertation separable into parts with short-term
goals. Work on the dissertation often competes
poorly with other tasks that offer more immediate
gratification. Confronting the dissertation as a
whole can lead to endless postponements.

* There was also a warning that dissertations
seldom turn out as planned; it is important to
hedge your bets and be prepared in case you do
not find data that speaks to the issue.

* A good dissertation topic should also allow you
to say something that is convincing to other
people. Each field has its own rules as to what
is compelling evidence. There is always a topic
of explanation and there must be interpretable
results.

* One speaker suggested that topics that involve
comparisons provide a more structured framework
than studies of individual subjects. He also
recommended building on the work of others. This
does not mean replication, but rather looking for
gaps or for ways to extend other investigations.
He stressed that very few things start de novo.
Having a framework, testing things that others
have done is very helpful.

* To find out what it is you would like to do, it
is helpful to be attentive to your reactions in
your scholarly reading. If you find yourself
saying "I wish I had written that," you can use
that as a key to finding something similar.

* Preparing a research design also requires
conversation. Research is often a solitary
activity, but designing research is an activity
that should be carried out collaboratively.
Decisions made at the stage of research design
are so crucial to the value of subsequent labor
that issues must be talked out thoroughly at the
outset. Even highly experienced researchers often
collaborate with colleagues, teach courses on
methodology with them, or pop into each other's
office with a query twice a day. Rule numbers one
for graduate students beginning their first large
research projects is: engage in an extended
conversation with your advisors. Even Jove, with
his legendary powers, could not generate a good
research design full-blown from his head.

* Looking to the future, the speakers addressed
the relationship between the dissertation topic
and job prospects. Both agreed that job
considerations should be subordinate to
intellectual interests. In any case, predicting
the market is like "guessing in the dark." A
topic that is in the mainstream of the discipline
might appear to be safer, but it may be in an
overcrowded field. That problem is not completely
solved by choosing a more peripheral topic, since
there may be less demand. In general, you should
avoid choosing a topic because you think it is
fashionable. They also added that the
dissertation topic does not necessarily identify
your field that precisely-hiring departments tend
to work by broad fields.

Good luck

ritaki