Why do I leave codebook documentation until after the survey is launched? If I had created the codebook while I was still scripting the survey I would have noticed that I coded all the research support responses of “not offered” as 3s but I should have coded them as 0s. Now I’ll have to recode those responses before I start analysis so I don’t confuse myself when I’m interpreting the results. Sigh, such is the messy life of an academic.
