At some point we made a decision to collapse the "I don't know" responses with NAs. This happens, for example, in the Q8 questions.
The most raw survey data in the repo right now is raw_hegs_rpr.rds, but the "I don't know" responses in that file are already recorded as NAs. Based on this:
It seems like an "excluding" option in Qualtrics modified the behavior of the API to pull NAs for any type of response that was excluded.
We do have code in the cleaning scripts to deal with the "I don't know" style responses
Will be better to reset Qualtrics to include these types of responses in the API pull and then re-save raw_hegs_rpr.rds inclusive of those responses.
Crosstab Q8_6 with subfields & methods prior to switching 'i don't know' to NA
I just did some searching and I think this setting would have to be changed at the level of every question & excluded response. I'm not sure that it's worth the effort to do that at this point.
At some point we made a decision to collapse the "I don't know" responses with NAs. This happens, for example, in the Q8 questions.
The most raw survey data in the repo right now is
raw_hegs_rpr.rds
, but the "I don't know" responses in that file are already recorded as NAs. Based on this:raw_hegs_rpr.rds
inclusive of those responses.