Information Technology Reference
In-Depth Information
are a range of approaches to advising that achieve these aims, but they are all based
on the strategy of learning while doing.
Some advisors, for example, set their students problems such as verifying a proof
in a published paper and seeing whether it can be applied to variants of the theorem,
thus, in effect, getting the student to explore the limits at which the theorem no longer
applies. Another example is to attempt to confirm someone else's results, by down-
loading code or by developing a fresh implementation. The difficulties encountered
in such efforts are a fertile source of research questions. Other advisors immediately
start their students on activities that are expected to lead to a research publication. It
is in this last case that the model of advising as apprenticeship is most evident.
Typically, in the early stages the advisor specifies each small step the student
should take: running a certain experiment, identifying a suitable source of data,
searching the literature to resolve a particular question, or writing one small section
of a proposed paper. As students mature into researchers, they become more indepen-
dent, often by anticipatingwhat their advisors will ask, while advisors gradually leave
more space for their students to assert this independence. Over time, the relationship
becomes one of guidance rather than management.
The trade-offs implicit in such a relationship are complex. One is the question of
is the degree of independence. Advisors often believe that their students are either
demanding or overconfident; students, on the other hand, can feel either confined
by excessive control or at a loss due to being expected to undertake tasks without
assistance. The needs of students who are working more or less alone may be very
different to those of students who are part of an extended research group.
An area where the advisor's expertise is critical is in scoping the project. It needs
to stand sufficiently alone from other current work, yet be relevant to a group's wider
activities. It should be open enough to allow innovation and freedom, yet have a good
likelihood of success. It should be close enough to the advisor's core expertise to
allow the advisor to verify that the work is sufficiently novel, and to verify that the
appropriate literature has been thoroughly explored. The fact that an advisor finds a
topic interesting does not by itself justify asking a student to work on it. Likewise,
a student who is keen on a topic must consider whether competent supervision is
available in that area.
Advisors can be busy people. Prepare for your meetings—bring tables of results
or lists of questions, for example. Be honest; if you are trying to convince your
advisor that you have completed some particular piece of work, then the work should
have been done. Advisors are not fools. Saying that you have been reading for a
week sounds like an excuse; and, if it is true, you probably haven't spent your time
effectively.
The student-advisor relationship is not only concerned with research training,
but is a means for advisors to be involved in research on a particular topic. Thus
students and advisors often write papers together. At times, this can be a source of
conflict, when, for example, an advisor wants a student to work on a paper while the
student wants to make progress on a thesis. On the other hand, the involvement of
