50 4 Making a Strong Start
The structure of a chapter should be organic, that is, should be appropriate to the
content. My student Surijit struggled with chapter organization. The body of his
thesis was designed as a series of chapters with linked experimental results, which
were tests of different mechanisms for identifying the source of contamination in
stored grains, such as rice that has been bagged for sale in retail outlets. These
mechanisms were diverse, ranging from chemical tests on the gases enclosed in the
packaging to analysis of the transportation and processing path from field to super-
market. With his supervisor as co-author Surijit had published a strong paper early
in his PhD, on a particular chemical test, and had revised this to give one of his first
‘contribution’ chapters. This established what he saw as the correct framework for
presenting research results, including a cost analysis and reflection on sources of
error. (He had those precise headings—Cost Analysis, Sources of Error.) So far so
good! But he then concluded that all of his chapters had to have the same structure,
regardless of whether it was appropriate, and even made curious decisions such as
to include a cost formula while announcing that all the variables were unknown
and could not be measured, and that the correctness of the formula could not be
determined. This shoehorning of all work into the one fixed ordering led to some
chapters that were very strange indeed. He needed to let each piece of work take a
more natural shape, and to have the confidence to explore what that shape might be.
Every chapter in a thesis should have a conclusion. The reader needs to share
with the writer a sense of what has been achieved, what is established now that
wasn’t established at the beginning of the chapter. And the conclusion should, of
course, respond to the stated aim or purpose of the chapter. There may be some
exceptions, for example, descriptive chapters outlining information on the charac-
teristics of a study area. Chapters that need strong conclusions are reviews of theory,
reviews of available research methods, reports of results, and the discussion (which
may be the conclusions to the whole thesis).
I often suggest to my students that they structure the end of a chapter as two sec-
tions, a discussion or analysis and a summary. The work reported in a chapter may
not yet allow definite conclusions, for example because the case being made in a
thesis rests on three separate strands of evidence, each reported separately, and so a
formal section of ‘conclusions’ may not be appropriate. But a typical chapter con-
tains a great deal of detail, and the reader needs the help of the author to sift through
this mass: What’s important? What overall themes can be identified? What can be
observed, or learned? A discussion is also a good place to consider shortcomings or
limitations in the work reported in the chapter. The summary then, in effect, replies
to the introduction by briefly identifying the chapter’s achievements, and sets the
scene for the next chapter.
Students often have difficulty with chapter endings. They tend to write lists of
what was in the chapter, and fail to state the significance or implications of what
was reported. (But don’t get ahead of yourself. I have often found that students offer
conclusions in background chapters that were informed by insights from their own
work. Save these for later in the thesis.)
Below is the conclusion to a chapter of a thesis describing part of a candidate’s
work.