Countries, regions and currency • ezrsurvey

library(ezrsurvey)
#> Loading required package: dplyr
#> 
#> Attaching package: 'dplyr'
#> The following objects are masked from 'package:stats':
#> 
#>     filter, lag
#> The following objects are masked from 'package:base':
#> 
#>     intersect, setdiff, setequal, union
#> Loading required package: ggplot2
#> Loading required package: tidyr
#> Loading required package: tibble
#> Loading required package: readr
#> Loading required package: stringr
#> Loading required package: purrr

Two columns make a multi-country study harder than a single-market one. The country question is usually free text, so it comes back holding every spelling a person can produce. And if respondents answered about money in their own currency, the amounts column cannot be averaged at all.

The country column people typed

On one real open salary survey of 26,232 people, nine thousand wrote “United States”, eight thousand wrote “USA”, two and a half thousand wrote “US”, and several hundred more wrote “U.S.”, “Usa” or “usa”. Matching on the name as typed resolved about half the column.

recode_region() folds the answer down before looking it up. Case, punctuation, accents and a leading “the” all stop mattering, and then a table of the spellings people actually use is tried.

recode_region(c("United States", "USA", "U.S.", "Estados Unidos"))
#> [1] "North America" "North America" "North America" "North America"

It carries the constituent countries of the United Kingdom, which a UK respondent is as likely to write as the country itself:

recode_region(c("England", "Scotland", "Wales", "Northern Ireland"))
#> [1] "Europe" "Europe" "Europe" "Europe"

Endonyms, because a survey fielded in more than one language collects them:

recode_region(c("Deutschland", "Brasil", "Espana", "Osterreich"))
#> [1] "Europe"        "South America" "Europe"        "Europe"

And the official ISO 3166-1 names, which is what a questionnaire built on the standard puts in front of a respondent, in either word order:

recode_region(c("Viet Nam", "Syrian Arab Republic", "Republic of Moldova",
                "Cabo Verde", "Eswatini"))
#> [1] "Asia"        "Middle East" "Europe"      "Africa"      "Africa"

The fold is done letter by letter rather than left to the platform’s transliteration, so a name carrying an accent matches the same way on Windows, macOS and Linux:

recode_region(c("Côte d'Ivoire", "Réunion", "Türkiye"))
#> [1] "Africa" "Africa" "Europe"

Adding the column

add_region() does the same thing to a data frame, and subregion = TRUE adds the finer grouping alongside:

survey <- add_region(podracing_survey, demo_country, subregion = TRUE)

calc_percentage(survey, region, sort = "desc")
#> # A tibble: 5 × 3
#>   region            n   pct
#>   <fct>         <int> <dbl>
#> 1 North America   503    50
#> 2 Europe          314    31
#> 3 South America    82     8
#> 4 Asia             59     6
#> 5 Oceania          42     4

calc_percentage(survey, subregion, sort = "desc")
#> # A tibble: 6 × 3
#>   subregion                 n   pct
#>   <fct>                 <int> <dbl>
#> 1 North America           503    50
#> 2 Western Europe          286    29
#> 3 South America            82     8
#> 4 East Asia                59     6
#> 5 Australia New Zealand    42     4
#> 6 Northern Europe          28     3

Anything that did not match comes back NA with a warning naming the values, because at that point what is left is genuine typos worth seeing rather than spellings worth supporting.

Reporting against the standard

A client who consolidates research across agencies wants countries coded to ISO 3166-1, not to whatever names this analyst happened to pick. A column already coded that way needs no preparation:

recode_region(c("US", "GBR", "DE", "JPN", "AUS"))
#> [1] "North America" "Europe"        "Europe"        "Asia"         
#> [5] "Oceania"

A two-letter code that is also an ordinary English word is only read as a country when it was typed in capitals, so "NO" is Norway and "no" is somebody answering a different question:

recode_region(c("NO", "no"), quiet = TRUE)
#> [1] "Europe" NA

The bundled table carries both codes, so a country name can be replaced by its code on the way out. That is the table to join to, because country names differ between sources and codes do not:

head(country_region)
#> # A tibble: 6 × 5
#>   country      iso2  iso3  region subregion
#>   <chr>        <chr> <chr> <chr>  <chr>    
#> 1 Algeria      DZ    DZA   Africa Africa   
#> 2 Angola       AO    AGO   Africa Africa   
#> 3 Benin        BJ    BEN   Africa Africa   
#> 4 Botswana     BW    BWA   Africa Africa   
#> 5 Burkina Faso BF    BFA   Africa Africa   
#> 6 Burundi      BI    BDI   Africa Africa

survey %>%
  dplyr::left_join(dplyr::select(country_region, country, iso2, iso3),
                   by = c("demo_country" = "country")) %>%
  dplyr::count(iso3, region, sort = TRUE) %>%
  head(8)
#> # A tibble: 8 × 3
#>   iso3  region            n
#>   <chr> <chr>         <int>
#> 1 USA   North America   336
#> 2 GBR   Europe          120
#> 3 MEX   North America    93
#> 4 DEU   Europe           89
#> 5 BRA   South America    82
#> 6 FRA   Europe           77
#> 7 CAN   North America    74
#> 8 JPN   Asia             59

Namibia’s alpha-2 code is the two letters NA, not a missing value. It is worth knowing before filtering that column, and before reading a country file with anything other than na.strings = "".

The regions themselves follow the UN M49 sub-regions with two deliberate departures, because they are how consumer research buys these markets: the Middle East is a region of its own rather than part of Asia and Africa, and the Americas are split north and south.

Money in eleven currencies

A single amount converts with convert_currency():

convert_currency(50000, from = "GBP", to = "USD")
#> [1] 63291.14
convert_currency(c(50000, 60000), from = c("EUR", "CAD"), to = "USD")
#> [1] 54347.83 44117.65

A whole column converts with add_currency(), which takes the currency from a second column, so each row is converted at its own rate:

pay <- tibble::tibble(
  respondent = 1:5,
  salary = c(52000, 61000, 48000, 75000, 39000),
  currency = c("GBP", "EUR", "USD", "CAD", "AUD")
)

add_currency(pay, salary, from = currency, to = "EUR", into = "salary_eur")
#> # A tibble: 5 × 4
#>   respondent salary currency salary_eur
#>        <int>  <dbl> <chr>         <dbl>
#> 1          1  52000 GBP          60557.
#> 2          2  61000 EUR          61000 
#> 3          3  48000 USD          44160 
#> 4          4  75000 CAD          50735.
#> 5          5  39000 AUD          23605.

The rates are bundled rather than fetched, so a figure computed today is the same figure next year. That is what a published number needs, and it is also why they are a dated snapshot rather than live rates.

head(list_currencies())
#> [1] "USD" "EUR" "GBP" "JPY" "CNY" "CHF"

When accuracy matters more than reproducibility, pass your own:

own_rates <- c(USD = 1, GBP = 0.79, EUR = 0.92)
convert_currency(50000, from = "GBP", to = "EUR", rates = own_rates)
#> [1] 58227.85

A code the table does not hold is reported rather than guessed at. Real currency columns contain things like "AUD/NZD" and "Other", and those rows come back NA with a warning naming the codes.