Don’t just claim it, prove it or assume it!

(a.k.a. How to give a damn in opinions and start using scientific evidence and hypotheses)

This posting is a part of my “Tips on technical writing” series.

It is quite common for many students (and many many experiences researchers) to claim things as if they were facts, or to express their wishes and opinions as if they were facts. This corresponds to a person believing he or she is a celebrity or a magician. Some examples of such cases are provided below:

  • I strongly believe that my system will help users in achieving…
  • It is my personal belief that users will like to….
  • This technique will most probably reduce search time by….
  • In order to solve problem A one has to create a software B that will….

There is nothing wrong with these sentences. They just need some help in order to become a part of the scientific literature. After all, what you write in your scientific reports needs to be scientific text. Here are a couple of hints that will help:

  • If you claim something, don’t use yourself and your personal opinion as scientific evidence (you could do it if you were a celebrity) but refer to other people’s research results.
  • If you claim something, but cannot find any research results to back up your claim, then don’t treat your claim as a fact but as a hypothesis, and do your research to create evidence for that claim.
  • In theoretical fields it might be the case that you don’t need empirical evidence but can use formal or logical proof in order to convince others that you are right.

Based on these two tips, the initial list of examples can look something like this:

  • Research shows that systems similar to what I have in mind can help users achieve X and Y [ref1,ref2,ref3] (where ref1,ref2,ref3 are references to existing research backing your claim).
  • According to [ref1,ref2,ref3] users like to….
  • An assumption of our research is that this technique will help reduce search time by a factor of X. In order to prove (or disprove) this we will design an experiment where….
  • Our assumption is that the mentioned software B will help solve problem A. In order to prove this we made a prototype and tested it on users who had problem A. The users agreed with us. (of course this still does not mean the assumption is now a fact, it only shows these particular users agreed with the assumption. But this is how knowledge is built).

If you want to know more about these topics read this book.