Don’t try to find every last jot of data relevant to your questions; that’s impossible. But you do need data that are sufficient and representative.
I’m so, so guilty of this. In fact, I found myself doing this last night while combing over some secondary resources, though I feel to do something similar but slight different: I will take a tiny, relevant piece of a large article and use it for my purposes, even if the article itself has little to do with my subject. I’m not sure if that’s bad or not, but it seems a shame to read something and not be able to use it at all.
You’ll believe in your claim so strongly that you will see all your evidence in its favor.
Oops. This is true and speaks to what I described above. Perhaps I should have read this before doing my secondary research notes… though I think it’s important to use sources in a way that support–not contradict–your thesis. So this is a double-edged sword in a lot of ways. I’ve not yet found a source that directly argues against my thesis, but I’m going to look for one, just to make sure I’m not blinded by my own confirmation bias.
“If you come across secondhand data (data that your source reports from another source), do everything you can to locate the original source. Not only can you then be sure your report is accurate (you may be surprised to see how often secondhand sources are not), but you may find other data equally useful.”