My research is in software engineering, but in a sub-field which is very close to social science. My papers normally contain sentences like "We conducted a study with 56 participants." and "Our previous study showed that [some assumptions are true]" and "We chose to use Cramer's V as the association measure, because [explanation why we thought it is better than other association measures]".
Now that I am close to my Ph.D. thesis, I am writing more texts alone, and the thesis is legally required to be my own work. So "we" is factually wrong. But using "I" feels immodest, and it is certainly unusual. But I don't know how to change my texts to avoid it.
I can't imagine how to apply the advice from that other answer to my case. "One conducted a study with 56 participants"? "The conducted study had 56 participants"? "A study was conducted, with 56 participants"? Unlike describing a mathematical proof, these sentences sound terrible. And how to explain my decision to use Cramer's V, when it is based on personal opinion?
Any advice how to deal with the matter outside of the world of mathematical proofs?
Another example why "I" might be needed. It is not only vanity; in the not-so-exact sciences there is sometimes lots of leeway involved. Say that I code some data. This is a very subjective process, and can be error prone. It is important for the readers to know that a coding was done by a single person, as this is considered less reliable than having somebody else repeat it and discuss any differences, and also because the coder has to take responsibility for any unusual decisions or errors.
There is a more general question on the same topic. But the accepted and highly-upvoted answer is from the point of view of a mathematician, it says that the writing style is best constrained to declarative sentences such as "Since p, it follows that q.".