For each sentence you write, ask yourself which information is conveyed by it to the
reader.
Do not write about what made you most effort, as this is usually not interesting for the
reader.
Be clear about what (scientific) question you want to answer in your thesis, in each
chapter, and possibly in each section.
Do not be unspecific if you know the precise number. If you evaluate 58 instances
and your code fails to solve 6, then write it fails on 6 instances and not
"on a few".
In general world like "often", "seldom", "many", "few", "usually"
are not particular scientific
If you put a table or diagram into your text, describe what the reader is expected
to learn from it, e.g. which statement is support by this and why.
It is not the job of the reader to find out themselves.
If there are strange numbers or NaNs in your table or funny bends in your figures, explain
those even if they are not part of what you want to show. You should not leave the
reader wonder what the meaning of it is.
All column heads if they are not completely obvious, should be described or defined
in the text.
Think about how many digits after the decimal point make sense, i.e. convey meaning
to the reader.
Make sure your citations are complete and correct.
Make it is clear what you are citing and what is your contribution. Whenever there
is no reference, you claim this as your contribution. Moreover, even if you have thought
of it yourself, still check whether there was somebody before.
If you write about some problem, explain the motivation. Why are you trying to solve
this problem? In contrast to mountains, there is an unlimited number of problems.
So "it was there" is not a good enough motivation most of the
times.
For citations, use at least [LH32], better cite as Laurel and
Hardy (1932), or Laurel and Hardy [14]. Just having [55] is not
acceptable. (This is not an extended Abstract with a 4 page
limit).
There are two major points about any thesis: (formal)
correctness and scientific content. For a bachelor thesis it is all
about correctness. Try your best. For a master thesis correctness regarding
scientific presentation is even more important, though there is also
some expectation on the scientific content. For a PhD thesis
correctness is taken for granted and the important part is the
content.
For the notation/definition part of a bachelor/master thesis, do not try to
write this yourself. Copy, i.e., cite, it from somewhere. First, it should be
standard anyway, second, it is often surprisingly hard to get definitions
entirely correct. There is nothing to win in this part of the
thesis, only to lose.