92 7 Establishing Your Contribution
had worked with for several months, she quickly found reasons why it might be
unreliable. (These included measures such as shifts in sales volumes and survey
questions to consumers. You might want to think for yourself about what the short-
comings might be.) It was as if it had never occurred to her to think about whether
each measure was plausible or not, but, when faced with a direct question, she could
readily find reasons to query them. That is, she hadn’t considered whether there
were confounds that would undermine her method.
Another perspective on the issue of measurement is that, often, the researcher
has made assumptions that other researchers would not see as justified. A common
problem that I see is that measures are just too simplistic—in the food wastage case
above, obviously the total tonnage is a concern, but it is also a concern if things
with a low production cost are kept while things with a high production cost are
discarded. You need to try and expose your assumptions, and justify them. As for
your hypothesis, you need to write about the assumptions underlying your measures
in order to tighten them into a form where they can be debated with others. It may
be that the measures themselves are something you have assumed, and not made ex-
plicit; it is critical that you directly ask yourself how the outcomes will be assessed,
and what the measures are, so that these decisions are made clear to the examiner.
The third interaction between method and thesis writing is your need to be con-
fident that the evidence you are gathering will in fact test your hypothesis. It may
seem surprising, but I have often seen research projects in which this is not the case.
Quite simply, somehow it happens that people start gathering data that isn’t going
to help them. I suspect that there is no general cause for why this happens. It may
be, for example, that a researcher goes looking for some particular data (say, of
numbers of people participating in a particular sport) but finds that it is inconsistent
or unavailable, and instead decides to make use of other data (say, of numbers of
people who attend sporting events) that seems as though it will be a good proxy;
and then some other level of approximation or indirection is introduced; and all too
soon the connection to the original argument is lost. Or it may be that the project
was inspired by the appearance of some particular data source (the discovery, say,
of historical records of goods shipped to Australia on convict vessels) but that the
hypothesis that was to make use of this data had, through the course of discussion
with the supervisor, shifted away from issues that the data could resolve.
Thus it is critical that you regularly assess the relevance of your data and method
to the aims of the project. If you have been following my advice, you will have writ-
ten down the aims as part of your introduction; you need to check these aims against
your method to make sure that they continue to be consistent with each other.
Argument
The issues of measurement and relevance are aspects of the need for your conclu-
sions to be built on a robust argument. As I discuss in Chap. 8, you will use your
data and method as part of a path from raw results and numbers to new knowledge;