In antaldaniel/retroharmonize: Ex Post Survey Data Harmonization

Arab Barometer Issues

We create another example to show the idisynchratic issues with Arab Barometer later.

to_harmonize_economy <- arabbarometer_metadata %>%
  filter( .data$var_name_orig %in% c("rowid", "country","date", "wt")|
           grepl("current economic situation", .data$label_orig)) %>%
  mutate(var_label = var_label_normalize(.data$label_orig)) %>%
  mutate(var_label = case_when(
    .data$var_name_orig == "country" ~ "Country",
    .data$var_name_orig == "rowid"   ~ "Unique ID AB English", # in pdf Unique ID AB English
    .data$var_name_orig == "date"    ~ "Date_of_interview",
    .data$var_name_orig == "wt"      ~ "Weight",
    TRUE ~ "evaluation_economic_situation")) %>%
  mutate (var_name = var_label_normalize(.data$var_label) )

set.seed(2021) # Let's see the same random example:
sample_n(to_harmonize_economy%>% 
       select ( all_of(c("id", "var_name", "var_label"))), 10)

merged_ab_economic  <- merge_waves ( waves = arab_waves, 
                           var_harmonization = to_harmonize_economy )

merged_ab_economic  <- lapply ( merged_ab_economic, 
         FUN = function(x) x  %>%
           mutate( country = normalize_country_names (country)))

merged_ab_economic   <- lapply(
  merged_ab_economic,
  FUN = function(df) df %>%
    rename_all(snakecase::to_snake_case))

harmonize_economic_situation <- function(x) {
  label_list <- list(
    from = c("very\\sgood", "1\\.\\svery good",
             "4\\.\\svery bad", "very\\sbad", 
             "2\\.\\sgood", "good", 
             "3\\.\\sbad", "bad", 
             "Don't know",  "Don't know \\(Do not read\\)", "8. i don't know", 
             "refuse", "refused", "Decline to answer (Do not read)", "not\\sresponsible",
             "9\\.\\sdeclined to answer",   "decline", 
             "missing"), 
    to = c("very_good", "very_good", 
           "very_bad", "very_bad",
           "good", "good",
           "bad", "bad",
           rep("do_not_know", 3), 
           rep("declined", 6),
           "missing"), 
    numeric_values = c(3,3,0,0,2,2,1,1, 
                       rep(99997,3),
                       rep(99998,6),
                       99999)
  )

  harmonize_values(
    x, 
    harmonize_labels = label_list, 
    na_values = c("do_not_know"=99997,
                  "declined"=99998,
                  "missing"=99999), 
    perl = TRUE
  )
}