I remember a teacher telling me to delete every occurrence of "I think," because it's redundant to tell people you think what you're writing.

That's good advice in that context and speaks to the difference in standards of evidence between normal expository or persuasive writing in school and scientific writing. In a scientific paper if you assert something it's important to flag to the reader what kind of claim this is, eg something demonstrated by other research, something that directly follows from your evidence, or something that could possibly be true given prior knowledge and your evidence.