Updated SurveyData function to handle haven-labelled data

[Dewi-Amaliah]

Updated SurveyData function to automatically read haven-labelled data to survey question and survey responses

[Dewi-Amaliah]

Hi all,

Here is the example code of using the haven functionality

survey_dat <- SurveyData$new(feline_survey_haven)
survey_dat$questions()
survey_dat$responses()

[lauken13]

This is great! Some thoughts:

• the code is failing on the checks because the new haven labelled data hasn't been documented
• if there is a variable that is missing labels for the levels BUT is a haven labelled variable, it won't be included e.g:

`survey_dat <- SurveyData$new(feline_survey_haven)
survey_dat$questions()
survey_dat$responses()

generate_dictionary(feline_survey_haven) %>% View()

feline_survey_haven2 <- feline_survey_haven %>%
mutate(pet_own = remove_val_labels(pet_own))

survey_dat <- SurveyData$new(feline_survey_haven2)
survey_dat$questions()
survey_dat$responses()`

• Do we want to be able to do a mix of labelled and unlabelled?

[lauken13]

From the meeting:

• Decided to have a preference list:
  For each variable:
  -> if it's explicitly named in the function input then use those labels
  -> else if its a haven labelled variable then use those variables
  -> else if its a factor labelled variable then use the factor level names
  -> else if its a binary variable then code 0 or 1 as the label