I am writing a thesis about a data analysis I'm implementing and am currently describing my data preparation where I talk about what needed to be done prior to the analysis (formatting the data, removing noisy data). Now, I was wondering if I should also include steps that only helped me to work more comfortably like creating multiple different sets of data with different variables?

This has no direct impact on the analysis but I'm not sure if it still needs to be included.

2 Answers 2


In general it is good to be complete, if you can do so without being overly pedantic. But the example you give (multiple data sets) seems like an especially important one to include. This is especially true as a single data set is seldom definitive (if the analysis relies on statistical data - sampling). Multiple, similar data sets gives general confirmation.

If you say too little, you leave yourself open to questions about what you considered and didn't or whether your results are too sensitive to extraneous factors.

Give enough information so that someone else is able to independently confirm your conclusion.

  • I didn't frame that clear enough I think. The amount of rows per dataset is the same, I just include a different subset of variables per dataset. I could do it all with one dataset but it's just more bothersome workflow-wise as I'd have to specifically select the needed variables each time. Therefore, I think in this case it'd be overly pedantic.
    – beld
    Oct 3, 2019 at 13:57

The answer depends on the expectations of your target community. Therefore, your best bet is to check similar papers from the same venue (journal or conference) and see how detailed the related instructions normally are.

You must log in to answer this question.

Not the answer you're looking for? Browse other questions tagged .