library(tidyverse) library(careless) library(formattable) library(ggmap) library(ggrepel) library(ggthemes) library(magrittr) library(mapproj) library(maps) library(naniar) library(RColorBrewer) library(rio) library(scales) library(trend) load("wci_df_reprex.Rdata") #IMPORT# ##import wci data index_import <- import("wci_data.csv") #replace missing values index_import <- index_import %>% mutate(across(where(is.character), ~na_if(., "--"))) index_import <- index_import %>% mutate(across(where(is.character), ~na_if(., "N/A"))) #name columns colnames(index_import) <- c("respID", "nationality", "residence", "technical1", "technical2", "technical3", "technical4", "technical5", "technical1_impact", "technical1_professional", "technical1_techskill", "technical2_impact", "technical2_professional", "technical2_techskill", "technical3_impact", "technical3_professional", "technical3_techskill", "technical4_impact", "technical4_professional", "technical4_techskill", "technical5_impact", "technical5_professional", "technical5_techskill", "attack1", "attack2", "attack3", "attack4", "attack5", "attack1_impact", "attack1_professional", "attack1_techskill", "attack2_impact", "attack2_professional", "attack2_techskill", "attack3_impact", "attack3_professional", "attack3_techskill", "attack4_impact", "attack4_professional", "attack4_techskill", "attack5_impact", "attack5_professional", "attack5_techskill", "data1", "data2", "data3", "data4", "data5", "data1_impact", "data1_professional", "data1_techskill", "data2_impact", "data2_professional", "data2_techskill", "data3_impact", "data3_professional", "data3_techskill", "data4_impact", "data4_professional", "data4_techskill", "data5_impact", "data5_professional", "data5_techskill", "scams1", "scams2", "scams3", "scams4", "scams5", "scams1_impact", "scams1_professional", "scams1_techskill", "scams2_impact", "scams2_professional", "scams2_techskill", "scams3_impact", "scams3_professional", "scams3_techskill", "scams4_impact", "scams4_professional", "scams4_techskill", "scams5_impact", "scams5_professional", "scams5_techskill", "cash1", "cash2", "cash3", "cash4", "cash5", "cash1_impact", "cash1_professional", "cash1_techskill", "cash2_impact", "cash2_professional", "cash2_techskill", "cash3_impact", "cash3_professional", "cash3_techskill", "cash4_impact", "cash4_professional", "cash4_techskill", "cash5_impact", "cash5_professional", "cash5_techskill", "expert_crimetype", "expert_crimetype_other", "expert_region", "expert_region_other", "comments") #CLEANING# #reshaping each variable to long technical1_long <- index_import |> select(respID:technical1, technical1_impact:technical1_techskill) |> pivot_longer(cols = technical1_impact:technical1_techskill, names_to = "measures", values_to = "rating") technical1_long$position <- 1 technical1_long$crimetype <- "technical" names(technical1_long) <- c("respID", "nationality", "residence", "country", "measures", "rating", "position", "crimetype") technical2_long <- index_import |> select(respID:residence, technical2, technical2_impact:technical2_techskill) |> pivot_longer(cols = technical2_impact:technical2_techskill, names_to = "measures", values_to = "rating") technical2_long$position <- 2 technical2_long$crimetype <- "technical" names(technical2_long) <- c("respID", "nationality", "residence", "country", "measures", "rating", "position", "crimetype") technical3_long <- index_import |> select(respID:residence, technical3, technical3_impact:technical3_techskill) |> pivot_longer(cols = technical3_impact:technical3_techskill, names_to = "measures", values_to = "rating") technical3_long$position <- 3 technical3_long$crimetype <- "technical" names(technical3_long) <- c("respID", "nationality", "residence", "country", "measures", "rating", "position", "crimetype") technical4_long <- index_import |> select(respID:residence, technical4, technical4_impact:technical4_techskill) |> pivot_longer(cols = technical4_impact:technical4_techskill, names_to = "measures", values_to = "rating") technical4_long$position <- 4 technical4_long$crimetype <- "technical" names(technical4_long) <- c("respID", "nationality", "residence", "country", "measures", "rating", "position", "crimetype") technical4_long$country <- str_replace(technical4_long$country, "Iran ", "Iran") technical5_long <- index_import |> select(respID:residence, technical5, technical5_impact:technical5_techskill) |> pivot_longer(cols = technical5_impact:technical5_techskill, names_to = "measures", values_to = "rating") technical5_long$position <- 5 technical5_long$crimetype <- "technical" names(technical5_long) <- c("respID", "nationality", "residence", "country", "measures", "rating", "position", "crimetype") attack1_long <- index_import |> select(respID:residence, attack1, attack1_impact:attack1_techskill) |> pivot_longer(cols = attack1_impact:attack1_techskill, names_to = "measures", values_to = "rating") attack1_long$position <- 1 attack1_long$crimetype <- "attack" names(attack1_long) <- c("respID", "nationality", "residence", "country", "measures", "rating", "position", "crimetype") attack2_long <- index_import |> select(respID:residence, attack2, attack2_impact:attack2_techskill) |> pivot_longer(cols = attack2_impact:attack2_techskill, names_to = "measures", values_to = "rating") attack2_long$position <- 2 attack2_long$crimetype <- "attack" names(attack2_long) <- c("respID", "nationality", "residence", "country", "measures", "rating", "position", "crimetype") attack3_long <- index_import |> select(respID:residence, attack3, attack3_impact:attack3_techskill) |> pivot_longer(cols = attack3_impact:attack3_techskill, names_to = "measures", values_to = "rating") attack3_long$position <- 3 attack3_long$crimetype <- "attack" names(attack3_long) <- c("respID", "nationality", "residence", "country", "measures", "rating", "position", "crimetype") attack4_long <- index_import |> select(respID:residence, attack4, attack4_impact:attack4_techskill) |> pivot_longer(cols = attack4_impact:attack4_techskill, names_to = "measures", values_to = "rating") attack4_long$position <- 4 attack4_long$crimetype <- "attack" names(attack4_long) <- c("respID", "nationality", "residence", "country", "measures", "rating", "position", "crimetype") attack5_long <- index_import |> select(respID:residence, attack5, attack5_impact:attack5_techskill) |> pivot_longer(cols = attack5_impact:attack5_techskill, names_to = "measures", values_to = "rating") attack5_long$position <- 5 attack5_long$crimetype <- "attack" names(attack5_long) <- c("respID", "nationality", "residence", "country", "measures", "rating", "position", "crimetype") data1_long <- index_import |> select(respID:residence, data1, data1_impact:data1_techskill) |> pivot_longer(cols = data1_impact:data1_techskill, names_to = "measures", values_to = "rating") data1_long$position <- 1 data1_long$crimetype <- "data" names(data1_long) <- c("respID", "nationality", "residence", "country", "measures", "rating", "position", "crimetype") data2_long <- index_import |> select(respID:residence, data2, data2_impact:data2_techskill) |> pivot_longer(cols = data2_impact:data2_techskill, names_to = "measures", values_to = "rating") data2_long$position <- 2 data2_long$crimetype <- "data" names(data2_long) <- c("respID", "nationality", "residence", "country", "measures", "rating", "position", "crimetype") data3_long <- index_import |> select(respID:residence, data3, data3_impact:data3_techskill) |> pivot_longer(cols = data3_impact:data3_techskill, names_to = "measures", values_to = "rating") data3_long$position <- 3 data3_long$crimetype <- "data" names(data3_long) <- c("respID", "nationality", "residence", "country", "measures", "rating", "position", "crimetype") data4_long <- index_import |> select(respID:residence, data4, data4_impact:data4_techskill) |> pivot_longer(cols = data4_impact:data4_techskill, names_to = "measures", values_to = "rating") data4_long$position <- 4 data4_long$crimetype <- "data" names(data4_long) <- c("respID", "nationality", "residence", "country", "measures", "rating", "position", "crimetype") data5_long <- index_import |> select(respID:residence, data5, data5_impact:data5_techskill) |> pivot_longer(cols = data5_impact:data5_techskill, names_to = "measures", values_to = "rating") data5_long$position <- 5 data5_long$crimetype <- "data" names(data5_long) <- c("respID", "nationality", "residence", "country", "measures", "rating", "position", "crimetype") scams1_long <- index_import |> select(respID:residence, scams1, scams1_impact:scams1_techskill) |> pivot_longer(cols = scams1_impact:scams1_techskill, names_to = "measures", values_to = "rating") scams1_long$position <- 1 scams1_long$crimetype <- "scams" names(scams1_long) <- c("respID", "nationality", "residence", "country", "measures", "rating", "position", "crimetype") scams2_long <- index_import |> select(respID:residence, scams2, scams2_impact:scams2_techskill) |> pivot_longer(cols = scams2_impact:scams2_techskill, names_to = "measures", values_to = "rating") scams2_long$position <- 2 scams2_long$crimetype <- "scams" names(scams2_long) <- c("respID", "nationality", "residence", "country", "measures", "rating", "position", "crimetype") scams3_long <- index_import |> select(respID:residence, scams3, scams3_impact:scams3_techskill) |> pivot_longer(cols = scams3_impact:scams3_techskill, names_to = "measures", values_to = "rating") scams3_long$position <- 3 scams3_long$crimetype <- "scams" names(scams3_long) <- c("respID", "nationality", "residence", "country", "measures", "rating", "position", "crimetype") scams4_long <- index_import |> select(respID:residence, scams4, scams4_impact:scams4_techskill) |> pivot_longer(cols = scams4_impact:scams4_techskill, names_to = "measures", values_to = "rating") scams4_long$position <- 4 scams4_long$crimetype <- "scams" names(scams4_long) <- c("respID", "nationality", "residence", "country", "measures", "rating", "position", "crimetype") scams5_long <- index_import |> select(respID:residence, scams5, scams5_impact:scams5_techskill) |> pivot_longer(cols = scams5_impact:scams5_techskill, names_to = "measures", values_to = "rating") scams5_long$position <- 5 scams5_long$crimetype <- "scams" names(scams5_long) <- c("respID", "nationality", "residence", "country", "measures", "rating", "position", "crimetype") cash1_long <- index_import |> select(respID:residence, cash1, cash1_impact:cash1_techskill) |> pivot_longer(cols = cash1_impact:cash1_techskill, names_to = "measures", values_to = "rating") cash1_long$position <- 1 cash1_long$crimetype <- "cash" names(cash1_long) <- c("respID", "nationality", "residence", "country", "measures", "rating", "position", "crimetype") cash2_long <- index_import |> select(respID:residence, cash2, cash2_impact:cash2_techskill) |> pivot_longer(cols = cash2_impact:cash2_techskill, names_to = "measures", values_to = "rating") cash2_long$position <- 2 cash2_long$crimetype <- "cash" names(cash2_long) <- c("respID", "nationality", "residence", "country", "measures", "rating", "position", "crimetype") cash3_long <- index_import |> select(respID:residence, cash3, cash3_impact:cash3_techskill) |> pivot_longer(cols = cash3_impact:cash3_techskill, names_to = "measures", values_to = "rating") cash3_long$position <- 3 cash3_long$crimetype <- "cash" names(cash3_long) <- c("respID", "nationality", "residence", "country", "measures", "rating", "position", "crimetype") cash4_long <- index_import |> select(respID:residence, cash4, cash4_impact:cash4_techskill) |> pivot_longer(cols = cash4_impact:cash4_techskill, names_to = "measures", values_to = "rating") cash4_long$position <- 4 cash4_long$crimetype <- "cash" names(cash4_long) <- c("respID", "nationality", "residence", "country", "measures", "rating", "position", "crimetype") cash5_long <- index_import |> select(respID:residence, cash5, cash5_impact:cash5_techskill) |> pivot_longer(cols = cash5_impact:cash5_techskill, names_to = "measures", values_to = "rating") cash5_long$position <- 5 cash5_long$crimetype <- "cash" names(cash5_long) <- c("respID", "nationality", "residence", "country", "measures", "rating", "position", "crimetype") expertise_crimetype_long <- index_import |> select(respID:residence, expert_crimetype:expert_crimetype_other) |> pivot_longer(cols=expert_crimetype:expert_crimetype_other, names_to ="ex", values_to = "expertise_crimetype") names(expertise_crimetype_long) <- c("respID", "nationality", "residence", "ex", "expertise_crimetype") expertise_crimetype_long$ex <- NULL expertise_crimetype_long <- expertise_crimetype_long |> group_by(respID, nationality, residence) |> summarise(expertise_crimetype = paste0(expertise_crimetype, collapse=",")) expertise_region_long <- index_import |> select(respID:residence, expert_region_other) |> pivot_longer(cols=expert_region_other, names_to ="ex", values_to = "expertise_region_other" ) expertise_region_long$ex <- NULL names(expertise_region_long) <- c("respID", "nationality", "residence", "expertise_region") expertise_region_long <- expertise_region_long |> group_by(respID, nationality, residence) |> summarise(expertise_region = paste0(expertise_region, collapse=",")) comments_long <- index_import |> select(respID:residence, comments) |> pivot_longer(cols=comments, names_to ="ex", values_to = "comments" ) names(comments_long) <- c("respID", "nationality", "residence", "ex", "comments") comments_long$ex=NULL nationality_long <- index_import |> select(respID:nationality) names(nationality_long) <- c("respID", "nationality") residence_long <- index_import |> select(respID, residence) names(residence_long) <- c("respID", "residence") ##join long variables together into new df, "index_long" index_long <- bind_rows(technical1_long, technical2_long, technical3_long, technical4_long, technical5_long, attack1_long, attack2_long, attack3_long, attack4_long, attack5_long, data1_long, data2_long, data3_long, data4_long, data5_long, scams1_long, scams2_long, scams3_long, scams4_long, scams5_long, cash1_long, cash2_long, cash3_long, cash4_long, cash5_long) index_long <- index_long |> left_join(expertise_crimetype_long) index_long <- index_long |> left_join(expertise_region_long) index_long <- index_long |> left_join(comments_long) na_strings <- c("N/A", "NA", "NA,", "Prefer not to say", "", " ") index_long <- index_long |> replace_with_na_all(condition = ~.x %in% na_strings) index_long$country <- str_replace(index_long$country, "Korea, North", "North Korea") index_long$country <- str_replace(index_long$country, "Korea, South", "South Korea") index_long$country <- str_replace(index_long$country, "Swaziland", "Eswatini") index_long$country <- str_replace(index_long$country, "Korea North", "North Korea") ##Frequency & region frequency variables - part of cleaning the dataset to ensure each country is counted the correct number of times ##country frequency (need to divide by 3 because each country has 3 measures/observations associated with it) index_long <- index_long |> group_by(country) |> mutate(country_freq=n()) |> mutate(country_freq=country_freq / 3) #country region frequency variable index_long<- index_long |> select(respID:country_freq) |> mutate(country_region = ifelse( grepl("Argentina|Bolivia|Brazil|Chile|Colombia|Ecuador|Guyana|Paraguay|Peru|Suriname|Uruguay|Venezuela", country),"South America", ifelse( grepl("Antigua and Barbuda|Bahamas|Barbados|Belize|Canada|Costa Rica|Cuba|Dominica|Dominican Republic|El Salvador|Grenada|Guatemala|Haiti|Honduras|Jamaica|Mexico|Nicaragua|Panama|Saint Kitts and Nevis|Saint Lucia|Saint Vincent and the Grenadines|Trinidad and Tobago|United States",country),"North America", ifelse( grepl("Albania|Andorra|Armenia|Austria|Azerbaijan|Belarus|Belgium|Bosnia and Herzegovina|Bulgaria|Croatia|Cyprus|Czech Republic|Denmark|Estonia|Finland|France|Georgia|Germany|Greece|Hungary|Iceland|Ireland|Italy|Kazakhstan|Latvia|Liechtenstein|Lithuania|Luxembourg|Malta|Moldova|Monaco|Montenegro|Netherlands|North Macedonia|Norway|Poland|Portugal|Romania|Russia|San Marino|Serbia|SLovakia|Slovenia|Spain|Sweden|Switzerland|Turkey|Ukraine|United Kingdom|Vatican City",country),"Europe", ifelse( grepl("Afghanistan|Australia|Bahrain|Bangladesh|Bhutan|Brunei|Cambodia|China|Cook Islands|East Timor|Egypt|Fiji|India|Indonesia|Iran|Iran |Iraq|Israel|Japan|Jordan|Kazakhstan|Kiribati|Kuwait|Kyrgyzstan|Laos|Lebanon|Libya|Malaysia|Maldives|Marshall Islands|Micronesia|Mongolia|Myanmar|Nauru|Nepal|New Zealand|Niue|North Korea|Oman|Pakistan|Palau|Palestine|Papua New Guinea|Philippines|Qatar|Samoa|Saudi Arabia|Singapore|Solomon Islands|South Korea|Sri Lanka|Syria|Tajikistan|Thailand|Tonga|Turkmenistan|Tuvalu|Uzbekistan|United Arab Emirates|Vanuatu|Vietnam|Yemen",country), "Asia Pacific", ifelse(grepl("Algeria|Angola|Benin|Botswana|Burkina Faso|Burundi|Cameroon|Cape Verde|Central African Republic|Chad|Comoros|Congo, Democratic Republic|Congo, Republic|Djibouti|Egypt|Equatorial Guinea|Eritrea|Eswatini|Ethiopia|Gabon|Gambia|Ghana|Guinea|Guinea-Bissau|Ivory Coast|Kenya|Lesotho|Liberia|Libya|Madagascar|Malawi|Mali|Mauritania|Mauritius|Morocco|Mozambique|Namibia|Niger|Nigeria|Rwanda|Sao Tome and Principe|Senegal|Seychelles|Sierra Leone|Somalia|South Africa|South Sudan|Sudan|Tanzania|Togo|Tunisia|Uganda|Zambia|Zimbabwe", country),"Africa", ifelse(grepl("Prefer not to say", nationality),"Prefer not to say", ifelse(grepl(NA, country),"NA", "Other")))))))) #nationality region frequency variable index_long<- index_long |> select(respID:country_region) |> mutate(nationality_region = ifelse( grepl("Argentina|Bolivia|Brazil|Chile|Colombia|Ecuador|Guyana|Paraguay|Peru|Suriname|Uruguay|Venezuela", nationality),"South America", ifelse( grepl("Antigua and Barbuda|Bahamas|Barbados|Belize|Canada|Costa Rica|Cuba|Dominica|Dominican Republic|El Salvador|Grenada|Guatemala|Haiti|Honduras|Jamaica|Mexico|Nicaragua|Panama|Saint Kitts and Nevis|Saint Lucia|Saint Vincent and the Grenadines|Trinidad and Tobago|United States",nationality),"North America", ifelse( grepl("Albania|Andorra|Armenia|Austria|Azerbaijan|Belarus|Belgium|Bosnia and Herzegovina|Bulgaria|Croatia|Cyprus|Czech Republic|Denmark|Estonia|Finland|France|Georgia|Germany|Greece|Hungary|Iceland|Ireland|Italy|Kazakhstan|Latvia|Liechtenstein|Lithuania|Luxembourg|Malta|Moldova|Monaco|Montenegro|Netherlands|North Macedonia|Norway|Poland|Portugal|Romania|Russia|San Marino|Serbia|SLovakia|Slovenia|Spain|Sweden|Switzerland|Turkey|Ukraine|United Kingdom|Vatican City",nationality),"Europe", ifelse( grepl("Afghanistan|Australia|Bahrain|Bangladesh|Bhutan|Brunei|Cambodia|China|Cook Islands|East Timor|Egypt|Fiji|India|Indonesia|Iran|Iran |Iraq|Israel|Japan|Jordan|Kazakhstan|Kiribati|Kuwait|Kyrgyzstan|Laos|Lebanon|Libya|Malaysia|Maldives|Marshall Islands|Micronesia|Mongolia|Myanmar|Nauru|Nepal|New Zealand|Niue|North Korea|Oman|Pakistan|Palau|Palestine|Papua New Guinea|Philippines|Qatar|Samoa|Saudi Arabia|Singapore|Solomon Islands|South Korea|Sri Lanka|Syria|Tajikistan|Thailand|Tonga|Turkmenistan|Tuvalu|Uzbekistan|United Arab Emirates|Vanuatu|Vietnam|Yemen",nationality), "Asia Pacific", ifelse(grepl("Algeria|Angola|Benin|Botswana|Burkina Faso|Burundi|Cameroon|Cape Verde|Central African Republic|Chad|Comoros|Congo, Democratic Republic|Congo, Republic|Djibouti|Egypt|Equatorial Guinea|Eritrea|Eswatini|Ethiopia|Gabon|Gambia|Ghana|Guinea|Guinea-Bissau|Ivory Coast|Kenya|Lesotho|Liberia|Libya|Madagascar|Malawi|Mali|Mauritania|Mauritius|Morocco|Mozambique|Namibia|Niger|Nigeria|Rwanda|Sao Tome and Principe|Senegal|Seychelles|Sierra Leone|Somalia|South Africa|South Sudan|Sudan|Tanzania|Togo|Tunisia|Uganda|Zambia|Zimbabwe", nationality),"Africa", ifelse(grepl("Prefer not to say", nationality),"Prefer not to say", ifelse(grepl(NA, nationality),"NA", "Other")))))))) #residence region frequency variable index_long<- index_long |> select(respID:nationality_region) |> mutate(residence_region = ifelse( grepl("Argentina|Bolivia|Brazil|Chile|Colombia|Ecuador|Guyana|Paraguay|Peru|Suriname|Uruguay|Venezuela", residence),"South America", ifelse( grepl("Antigua and Barbuda|Bahamas|Barbados|Belize|Canada|Costa Rica|Cuba|Dominica|Dominican Republic|El Salvador|Grenada|Guatemala|Haiti|Honduras|Jamaica|Mexico|Nicaragua|Panama|Saint Kitts and Nevis|Saint Lucia|Saint Vincent and the Grenadines|Trinidad and Tobago|United States",residence),"North America", ifelse( grepl("Albania|Andorra|Armenia|Austria|Azerbaijan|Belarus|Belgium|Bosnia and Herzegovina|Bulgaria|Croatia|Cyprus|Czech Republic|Denmark|Estonia|Finland|France|Georgia|Germany|Greece|Hungary|Iceland|Ireland|Italy|Kazakhstan|Latvia|Liechtenstein|Lithuania|Luxembourg|Malta|Moldova|Monaco|Montenegro|Netherlands|North Macedonia|Norway|Poland|Portugal|Romania|Russia|San Marino|Serbia|SLovakia|Slovenia|Spain|Sweden|Switzerland|Turkey|Ukraine|United Kingdom|Vatican City",residence),"Europe", ifelse( grepl("Afghanistan|Australia|Bahrain|Bangladesh|Bhutan|Brunei|Cambodia|China|Cook Islands|East Timor|Egypt|Fiji|India|Indonesia|Iran|Iran |Iraq|Israel|Japan|Jordan|Kazakhstan|Kiribati|Kuwait|Kyrgyzstan|Laos|Lebanon|Libya|Malaysia|Maldives|Marshall Islands|Micronesia|Mongolia|Myanmar|Nauru|Nepal|New Zealand|Niue|North Korea|Oman|Pakistan|Palau|Palestine|Papua New Guinea|Philippines|Qatar|Samoa|Saudi Arabia|Singapore|Solomon Islands|South Korea|Sri Lanka|Syria|Tajikistan|Thailand|Tonga|Turkmenistan|Tuvalu|Uzbekistan|United Arab Emirates|Vanuatu|Vietnam|Yemen",residence), "Asia Pacific", ifelse(grepl("Algeria|Angola|Benin|Botswana|Burkina Faso|Burundi|Cameroon|Cape Verde|Central African Republic|Chad|Comoros|Congo, Democratic Republic|Congo, Republic|Djibouti|Egypt|Equatorial Guinea|Eritrea|Eswatini|Ethiopia|Gabon|Gambia|Ghana|Guinea|Guinea-Bissau|Ivory Coast|Kenya|Lesotho|Liberia|Libya|Madagascar|Malawi|Mali|Mauritania|Mauritius|Morocco|Mozambique|Namibia|Niger|Nigeria|Rwanda|Sao Tome and Principe|Senegal|Seychelles|Sierra Leone|Somalia|South Africa|South Sudan|Sudan|Tanzania|Togo|Tunisia|Uganda|Zambia|Zimbabwe", residence),"Africa", ifelse(grepl("Prefer not to say", nationality),"Prefer not to say", ifelse(grepl(NA, residence),"NA", "Other")))))))) #more frequency variables (need to divide by multiples of 3 b/c each country has 3 observations associated with it) index_long <- index_long |> group_by(country_region) |> mutate(country_region_freq=n()) |> mutate(country_region_freq=country_region_freq / 3) index_long <- index_long |> group_by(nationality) |> mutate(nationality_freq=n()) |> mutate(nationality_freq=nationality_freq / 75) index_long <- index_long |> group_by(nationality_region) |> mutate(nationality_region_freq=n()) |> mutate(nationality_region_freq=nationality_region_freq / 75) index_long <- index_long |> group_by(residence) |> mutate(residence_freq=n()) |> mutate(residence_freq=residence_freq / 75) index_long <- index_long |> group_by(residence_region) |> mutate(residence_region_freq=n()) |> mutate(residence_region_freq=residence_region_freq / 75) #local & not local variable for nationality and residence #this will be helpful for later analysis/regressions index_long <- index_long |> select(everything()) |> mutate( nat_local = case_when( country == nationality ~ "local", country != nationality ~ "not_local", .default = NA ) ) index_long <- index_long |> select(everything()) |> mutate( res_local = case_when( country == residence ~ "local", country != residence ~ "not_local", .default = NA ) ) #re-order variables index_long <- index_long |> relocate(nationality_freq, .after=nationality) index_long <- index_long |> relocate(residence_freq, .after=residence) index_long <- index_long |> relocate(country_freq, .after=country) index_long <- index_long |> relocate(country_region, .after=country_freq) index_long <- index_long |> relocate(country_region_freq, .after=country_region) index_long <- index_long |> relocate(nationality_freq, .after=nationality) index_long <- index_long |> relocate(nationality_region, .after=nationality_freq) index_long <- index_long |> relocate(nationality_region_freq, .after=nationality_region) index_long <- index_long |> relocate(residence_region, .after=residence_freq) index_long <- index_long |> relocate(residence_region_freq, .after=residence_region) index_long <- index_long |> relocate(country:country_region_freq, .after=residence_region_freq) index_long <- index_long |> relocate(res_local, .after=residence_freq) index_long <- index_long |> relocate(nat_local, .after=nationality_freq) index_long <- index_long |> arrange(respID) #add "survey position" to do regressions that look at how rating behaviour changed as the survey progressed #we won't be able to do this with the next survey, which randomises the order of the cybercrime categories index_long <- index_long |> add_column(c(rep(1:5, each=15, length.out=6900))) |> rename(survey_position="c(rep(1:5, each = 15, length.out = 6900))") summary(lm(rating ~ survey_position, data = index_long)) plot(lm(rating ~ survey_position, data = index_long)) ggplot(index_long, aes(survey_position, rating))+ geom_point()+ geom_smooth(method="lm")+ labs(title="Ratings as survey progressed", x="Survey progression", y="Rating") #WCI INDICES# #first count the number of times a country was nominated within a cybercrime category. This will be helpful when composing the full index later tech_count <- index_long |> filter(str_detect(crimetype, "technical")) |> distinct() |> group_by(country) |> mutate(noms_tech=n()) |> mutate(noms_tech=noms_tech/3) |> select(country, noms_tech) |> distinct() |> arrange(desc(noms_tech)) attack_count <- index_long |> filter(str_detect(crimetype, "attack")) |> distinct() |> group_by(country) |> mutate(noms_attack=n()) |> mutate(noms_attack=noms_attack/3) |> select(country, noms_attack) |> distinct() |> arrange(desc(noms_attack)) data_count <- index_long |> filter(str_detect(crimetype, "data")) |> distinct() |> group_by(country) |> mutate(noms_data=n()) |> mutate(noms_data=noms_data/3) |> select(country, noms_data) |> distinct() |> arrange(desc(noms_data)) scams_count <- index_long |> filter(str_detect(crimetype, "scams")) |> distinct() |> group_by(country) |> mutate(noms_scams=n()) |> mutate(noms_scams=noms_scams/3) |> select(country, noms_scams) |> distinct() |> arrange(desc(noms_scams)) cash_count <- index_long |> filter(str_detect(crimetype, "cash")) |> distinct() |> group_by(country) |> mutate(noms_cash=n()) |> mutate(noms_cash=noms_cash/3) |> select(country, noms_cash) |> distinct() |> arrange(desc(noms_cash)) #Index by cybercrime type# ##First I filter index_long for each measure (impact, professionalism, technical skill) and summarise() to get the mean. I make a separate df for each measure. ##Second I combine these dfs into a single index and clean up. ##Finally I calculate the overall and WCI scores from the mean scores for each measure. ##Tech Index## tech_index_impact <- index_long |> filter(if_all(everything(), ~str_detect(crimetype, "technical"))) |> filter(if_all(everything(), ~str_detect(measures, "impact"))) |> select(everything()) |> group_by(country, country_freq) |> summarise(impact=mean(rating)) |> arrange(desc(country_freq)) tech_index_prof <- index_long |> filter(if_all(everything(), ~str_detect(crimetype, "technical"))) |> filter(if_all(everything(), ~str_detect(measures, "professional"))) |> select(everything()) |> group_by(country, country_freq) |> summarise(professionalism=mean(rating)) |> arrange(desc(country_freq)) tech_index_ts <- index_long |> filter(if_all(everything(), ~str_detect(crimetype, "technical"))) |> filter(if_all(everything(), ~str_detect(measures, "techskill"))) |> select(everything()) |> group_by(country, country_freq) |> summarise(technical_skill=mean(rating)) |> arrange(desc(country_freq)) tech_index <- left_join(tech_index_impact, tech_index_prof) tech_index <- left_join(tech_index, tech_index_ts) tech_index <- left_join(tech_index, tech_count) tech_index <- rename(tech_index, noms=noms_tech) tech_index <- drop_na(tech_index) tech_index$country_freq=NULL ##calculating "overall" (mean of impact+professionalism+technical skill) ##and WCI score (overall*(noms/92)*10) tech_index <- tech_index |> select(everything()) |> group_by(country) |> mutate(overall=mean(c(impact, professionalism, technical_skill))) tech_index <- tech_index |> mutate(wci = (overall*(noms/92))*10) |> arrange(desc(wci)) |> ungroup() |> select(everything()) |> mutate(rank = min_rank(desc(wci))) |> relocate(noms, .before = country) |> relocate(rank, .before = noms) colnames(tech_index) <- c("Rank", "Noms", "Country", "Impact", "Professionalism", "Technical_skill", "Overall", "WCI_Score") ##Attack Index## attack_index_impact <- index_long |> filter(if_all(everything(), ~str_detect(crimetype, "attack"))) |> filter(if_all(everything(), ~str_detect(measures, "impact"))) |> select(everything()) |> group_by(country, country_freq) |> summarise(impact=mean(rating)) |> arrange(desc(country_freq)) attack_index_prof <- index_long |> filter(if_all(everything(), ~str_detect(crimetype, "attack"))) |> filter(if_all(everything(), ~str_detect(measures, "professional"))) |> select(everything()) |> group_by(country, country_freq) |> summarise(professionalism=mean(rating)) |> arrange(desc(country_freq)) attack_index_ts <- index_long |> filter(if_all(everything(), ~str_detect(crimetype, "attack"))) |> filter(if_all(everything(), ~str_detect(measures, "techskill"))) |> select(everything()) |> group_by(country, country_freq) |> summarise(technical_skill=mean(rating)) |> arrange(desc(country_freq)) attack_index <- left_join(attack_index_impact, attack_index_prof) attack_index <- left_join(attack_index, attack_index_ts) attack_index <- left_join(attack_index, attack_count) attack_index <- rename(attack_index, noms=noms_attack) attack_index <- drop_na(attack_index) attack_index$country_freq=NULL #calculating "overall" & WCI score attack_index <- attack_index |> select(everything()) |> group_by(country) |> mutate(overall=mean(c(impact, professionalism, technical_skill))) attack_index <- attack_index |> mutate(wci = (overall*(noms/92))*10) |> arrange(desc(wci)) |> ungroup() |> select(everything()) |> mutate(rank = min_rank(desc(wci))) |> relocate(noms, .before = country) |> relocate(rank, .before = noms) colnames(attack_index) <- c("Rank", "Noms", "Country", "Impact", "Professionalism", "Technical_skill", "Overall", "WCI_Score") ##Data Index## data_index_impact <- index_long |> filter(if_all(everything(), ~str_detect(crimetype, "data"))) |> filter(if_all(everything(), ~str_detect(measures, "impact"))) |> select(everything()) |> group_by(country, country_freq) |> summarise(impact=mean(rating)) |> arrange(desc(country_freq)) data_index_prof <- index_long |> filter(if_all(everything(), ~str_detect(crimetype, "data"))) |> filter(if_all(everything(), ~str_detect(measures, "professional"))) |> select(everything()) |> group_by(country, country_freq) |> summarise(professionalism=mean(rating)) |> arrange(desc(country_freq)) data_index_ts <- index_long |> filter(if_all(everything(), ~str_detect(crimetype, "data"))) |> filter(if_all(everything(), ~str_detect(measures, "techskill"))) |> select(everything()) |> group_by(country, country_freq) |> summarise(technical_skill=mean(rating)) |> arrange(desc(country_freq)) data_index <- left_join(data_index_impact, data_index_prof) data_index <- left_join(data_index, data_index_ts) data_index <- left_join(data_index, data_count) data_index <- rename(data_index, noms=noms_data) data_index <- drop_na(data_index) data_index$country_freq=NULL data_index <- drop_na(data_index) #calculating "overall" & WCI score data_index <- data_index |> select(everything()) |> group_by(country) |> mutate(overall=mean(c(impact, professionalism, technical_skill))) data_index <- data_index |> mutate(wci = (overall*(noms/92))*10) |> arrange(desc(wci)) |> ungroup() |> select(everything()) |> mutate(rank = min_rank(desc(wci))) |> relocate(noms, .before = country) |> relocate(rank, .before = noms) colnames(data_index) <- c("Rank", "Noms", "Country", "Impact", "Professionalism", "Technical_skill", "Overall", "WCI_Score") ##Scams Index## scams_index_impact <- index_long |> filter(if_all(everything(), ~str_detect(crimetype, "scams"))) |> filter(if_all(everything(), ~str_detect(measures, "impact"))) |> select(everything()) |> group_by(country, country_freq) |> summarise(impact=mean(rating)) |> arrange(desc(country_freq)) scams_index_prof <- index_long |> filter(if_all(everything(), ~str_detect(crimetype, "scams"))) |> filter(if_all(everything(), ~str_detect(measures, "professional"))) |> select(everything()) |> group_by(country, country_freq) |> summarise(professionalism=mean(rating)) |> arrange(desc(country_freq)) scams_index_ts <- index_long |> filter(if_all(everything(), ~str_detect(crimetype, "scams"))) |> filter(if_all(everything(), ~str_detect(measures, "techskill"))) |> select(everything()) |> group_by(country, country_freq) |> summarise(technical_skill=mean(rating)) |> arrange(desc(country_freq)) scams_index <- left_join(scams_index_impact, scams_index_prof) scams_index <- left_join(scams_index, scams_index_ts) scams_index <- left_join(scams_index, scams_count) scams_index <- rename(scams_index, noms=noms_scams) scams_index <- drop_na(scams_index) scams_index$country_freq=NULL scams_index <- drop_na(scams_index) #calculating "overall" & WCI score scams_index <- scams_index |> select(everything()) |> group_by(country) |> mutate(overall=mean(c(impact, professionalism, technical_skill))) scams_index <- scams_index |> mutate(wci = (overall*(noms/92))*10) |> arrange(desc(wci)) |> ungroup() |> select(everything()) |> mutate(rank = min_rank(desc(wci))) |> relocate(noms, .before = country) |> relocate(rank, .before = noms) colnames(scams_index) <- c("Rank", "Noms", "Country", "Impact", "Professionalism", "Technical_skill", "Overall", "WCI_Score") ##Cash Index## cash_index_impact <- index_long |> filter(if_all(everything(), ~str_detect(crimetype, "cash"))) |> filter(if_all(everything(), ~str_detect(measures, "impact"))) |> select(everything()) |> group_by(country, country_freq) |> summarise(impact=mean(rating)) |> arrange(desc(country_freq)) cash_index_prof <- index_long |> filter(if_all(everything(), ~str_detect(crimetype, "cash"))) |> filter(if_all(everything(), ~str_detect(measures, "professional"))) |> select(everything()) |> group_by(country, country_freq) |> summarise(professionalism=mean(rating)) |> arrange(desc(country_freq)) cash_index_ts <- index_long |> filter(if_all(everything(), ~str_detect(crimetype, "cash"))) |> filter(if_all(everything(), ~str_detect(measures, "techskill"))) |> select(everything()) |> group_by(country, country_freq) |> summarise(technical_skill=mean(rating)) |> arrange(desc(country_freq)) cash_index <- left_join(cash_index_impact, cash_index_prof) cash_index <- left_join(cash_index, cash_index_ts) cash_index <- left_join(cash_index, cash_count) cash_index[sapply(cash_index, is.nan)]<-NA cash_index <- rename(cash_index, noms=noms_cash) cash_index$country_freq=NULL cash_index <- drop_na(cash_index) #calculating "overall" & WCI score cash_index <- cash_index |> select(everything()) |> group_by(country) |> mutate(overall=mean(c(impact, professionalism, technical_skill))) cash_index <- cash_index |> mutate(wci = (overall*(noms/92))*10) |> arrange(desc(wci)) |> ungroup() |> select(everything()) |> mutate(rank = min_rank(desc(wci))) |> relocate(noms, .before = country) |> relocate(rank, .before = noms) colnames(cash_index) <- c("Rank", "Noms", "Country", "Impact", "Professionalism", "Technical_skill", "Overall", "WCI_Score") ##WCI Overall Index## overall_index_impact <- index_long |> filter(if_all(everything(), ~str_detect(measures, "impact"))) |> select(everything()) |> group_by(country, country_freq) |> summarise(impact=mean(rating)) |> arrange(desc(country_freq)) |> print() overall_index_prof <- index_long |> filter(if_all(everything(), ~str_detect(measures, "professional"))) |> select(everything()) |> group_by(country, country_freq) |> summarise(professionalism=mean(rating)) |> arrange(desc(country_freq)) overall_index_ts <- index_long |> filter(if_all(everything(), ~str_detect(measures, "techskill"))) |> select(everything()) |> group_by(country, country_freq) |> summarise(technical_skill=mean(rating)) |> arrange(desc(country_freq)) overall_index <- left_join(overall_index_impact, overall_index_prof) overall_index <- left_join(overall_index, overall_index_ts) overall_index <- drop_na(overall_index) overall_index <- rename(overall_index, noms = country_freq) #calculating "overall" & WCI score overall_index <- overall_index |> select(everything()) |> group_by(country) |> mutate(overall=mean(c(impact, professionalism, technical_skill))) view(overall_index) overall_index <- overall_index |> mutate(wci = (overall*(noms/460))*10) |> arrange(desc(wci)) |> ungroup() |> select(everything()) |> mutate(rank = min_rank(desc(wci))) |> relocate(noms, .before = country) |> relocate(rank, .before = noms) colnames(overall_index) <- c("Rank", "Noms", "Country", "Impact", "Professionalism", "Technical_skill", "Overall", "WCI_Score") #complete WCI index includes all 197 UN-recognised countries #code below creates dfs for each index with all countries country.list <- c("Afghanistan", "Albania", "Algeria", "Andorra", "Angola", "Antigua and Barbuda", "Argentina", "Armenia", "Australia", "Austria", "Azerbaijan", "The Bahamas", "Bahrain", "Bangladesh", "Barbados", "Belarus", "Belgium", "Belize", "Benin", "Bhutan", "Bolivia", "Bosnia and Herzegovina", "Botswana", "Brazil", "Brunei", "Bulgaria", "Burkina Faso", "Burundi", "Cambodia", "Cameroon", "Canada", "Cape Verde", "Central African Republic", "Chad", "Chile", "China", "Colombia", "Comoros", "Congo, Democratic Republic of the", "Congo, Republic of the", "Cook Islands", "Costa Rica", "Croatia", "Cuba", "Cyprus", "Czech Republic", "Denmark", "Djibouti", "Dominica", "Dominican Republic", "East Timor", "Ecuador", "Egypt", "El Salvador", "Equatorial Guinea", "Eritrea", "Estonia", "Eswatini", "Ethiopia", "Fiji", "Finland", "France", "Gabon", "Gambia", "Georgia", "Germany", "Ghana", "Greece", "Grenada", "Guatemala", "Guinea", "Guinea-Bissau", "Guyana", "Haiti", "Honduras", "Hungary", "Iceland", "India", "Indonesia", "Iran", "Iraq", "Ireland", "Israel", "Italy", "Ivory Coast", "Jamaica", "Japan", "Jordan", "Kazakhstan", "Kenya", "Kiribati", "Kuwait", "Kyrgyzstan", "Laos", "Latvia", "Lebanon", "Lesotho", "Liberia", "Libya", "Liechtenstein", "Lithuania", "Luxembourg", "Madagascar", "Malawi", "Malaysia", "Maldives", "Mali", "Malta", "Marshall Islands", "Mauritania", "Mauritius", "Mexico", "Micronesia", "Moldova", "Monaco", "Mongolia", "Montenegro", "Morocco", "Mozambique", "Myanmar", "Namibia", "Nauru", "Nepal", "Netherlands", "New Zealand", "Nicaragua", "Niger", "Nigeria", "Niue", "North Korea", "North Macedonia", "Norway", "Oman", "Pakistan", "Palau", "Palestine", "Panama", "Papua New Guinea", "Paraguay", "Peru", "Philippines", "Poland", "Portugal", "Qatar", "Romania", "Russia", "Rwanda", "Saint Kitts and Nevis", "Saint Lucia", "Saint Vincent and the Grenadines", "Samoa", "San Marino", "Sao Tome and Principe", "Saudi Arabia", "Senegal", "Serbia", "Seychelles", "Sierra Leone", "Singapore", "Slovakia", "Slovenia", "Solomon Islands", "Somalia", "South Africa", "South Korea", "South Sudan", "Spain", "Sri Lanka", "Sudan", "Suriname", "Sweden", "Switzerland", "Syria", "Tajikistan", "Tanzania", "Thailand", "Togo", "Tonga", "Trinidad and Tobago", "Tunisia", "Turkey", "Turkmenistan", "Tuvalu", "Uganda", "Ukraine", "United Arab Emirates", "United Kingdom", "United States", "Uruguay", "Uzbekistan", "Vanuatu", "Vatican City", "Venezuela", "Vietnam", "Yemen", "Zambia", "Zimbabwe") country.list.index <- as.data.frame(country.list) country.list.index <- cbind(Rank = NA, Noms = NA, country.list, Impact = NA, Professionalism = NA, Technical_skill = NA, Overall = NA, WCI_Score = NA) colnames(country.list.index) <- c("Rank", "Noms", "Country", "Impact", "Professionalism", "Technical_skill", "Overall", "WCI_Score") country.list.index <- as.data.frame(country.list.index) country.list.index$Rank <- as.numeric(country.list.index$Rank) country.list.index$Noms <- as.numeric(country.list.index$Noms) country.list.index$Impact <- as.numeric(country.list.index$Impact) country.list.index$Professionalism <- as.numeric(country.list.index$Professionalism) country.list.index$Technical_skill <- as.numeric(country.list.index$Technical_skill) country.list.index$Overall <- as.numeric(country.list.index$Overall) country.list.index$WCI_Score <- as.numeric(country.list.index$WCI_Score) #Tech index with all countries WCI_tech <- bind_rows(tech_index, country.list.index) WCI_tech <- WCI_tech |> distinct(Country, .keep_all = T) view(WCI_tech) WCI_tech[45:197,1] <-rep(45) #Attack index with all countries WCI_attack <- bind_rows(attack_index, country.list.index) WCI_attack <- WCI_attack |> distinct(Country, .keep_all = T) view(WCI_attack) WCI_attack[46:197,1] <-rep(46) #Data index with all countries WCI_data <- bind_rows(data_index, country.list.index) WCI_data <- WCI_data |> distinct(Country, .keep_all = T) view(WCI_data) WCI_data[55:197,1] <-rep(55) #Scams index with all countries WCI_scams <- bind_rows(scams_index, country.list.index) WCI_scams <- WCI_scams |> distinct(Country, .keep_all = T) view(WCI_scams) WCI_scams[53:197,1] <-rep(53) #Cash index with all countries WCI_cash <- bind_rows(cash_index, country.list.index) WCI_cash <- WCI_cash |> distinct(Country, .keep_all = T) view(WCI_cash) WCI_cash[67:197,1] <-rep(67) #Overall index with all countries WCI_overall <- bind_rows(overall_index, country.list.index) WCI_overall <- WCI_overall |> distinct(Country, .keep_all = T) view(WCI_overall) WCI_overall[98:197,1] <-rep(98) #exporting index_wide to test the WCI score calculations #index_wide <- index_long |> #group_by(measures) |> #mutate(row = row_number()) |> #pivot_wider(names_from=measures, values_from = rating) |> #select(-row) #index_wide_temp <- index_wide |> #select(country, country_freq, technical1_impact:cash5_techskill) |> #arrange(country) #index_wide_temp <- index_wide_temp[1:18,] #selecting Afghanistan, Algeria, Angola, Argenti, Armenia, Australia #write.csv(index_wide_temp,"wci_overall_test.csv") #second test #index_wide_temp <- index_wide |> #select(country, country_freq, technical1_impact:cash5_techskill) |> #arrange(country) #index_wide_temp <- index_wide_temp[19:116,] #write.csv(index_wide_temp,"wci_overall_test2.csv") #rounding decimals for the WCI indices WCI_attack$Impact <- round(WCI_attack$Impact, digits=2) WCI_attack$Professionalism <- round(WCI_attack$Professionalism, digits=2) WCI_attack$Technical_skill <- round(WCI_attack$Technical_skill, digits=2) WCI_attack$Overall <- round(WCI_attack$Overall, digits=2) WCI_attack$WCI_Score <- round(WCI_attack$WCI_Score, digits=2) WCI_tech$Impact <- round(WCI_tech$Impact, digits=2) WCI_tech$Professionalism <- round(WCI_tech$Professionalism, digits=2) WCI_tech$Technical_skill <- round(WCI_tech$Technical_skill, digits=2) WCI_tech$Overall <- round(WCI_tech$Overall, digits=2) WCI_tech$WCI_Score <- round(WCI_tech$WCI_Score, digits=2) WCI_data$Impact <- round(WCI_data$Impact, digits=2) WCI_data$Professionalism <- round(WCI_data$Professionalism, digits=2) WCI_data$Technical_skill <- round(WCI_data$Technical_skill, digits=2) WCI_data$Overall <- round(WCI_data$Overall, digits=2) WCI_data$WCI_Score <- round(WCI_data$WCI_Score, digits=2) WCI_scams$Impact <- round(WCI_scams$Impact, digits=2) WCI_scams$Professionalism <- round(WCI_scams$Professionalism, digits=2) WCI_scams$Technical_skill <- round(WCI_scams$Technical_skill, digits=2) WCI_scams$Overall <- round(WCI_scams$Overall, digits=2) WCI_scams$WCI_Score <- round(WCI_scams$WCI_Score, digits=2) WCI_cash$Impact <- round(WCI_cash$Impact, digits=2) WCI_cash$Professionalism <- round(WCI_cash$Professionalism, digits=2) WCI_cash$Technical_skill <- round(WCI_cash$Technical_skill, digits=2) WCI_cash$Overall <- round(WCI_cash$Overall, digits=2) WCI_cash$WCI_Score <- round(WCI_cash$WCI_Score, digits=2) view(WCI_attack) WCI_overall$Impact <- round(WCI_overall$Impact, digits=2) WCI_overall$Professionalism <- round(WCI_overall$Professionalism, digits=2) WCI_overall$Technical_skill <- round(WCI_overall$Technical_skill, digits=2) WCI_overall$Overall <- round(WCI_overall$Overall, digits=2) WCI_overall$WCI_Score <- round(WCI_overall$WCI_Score, digits=2) view(WCI_overall) #DATA VISUALISATION# #WCIoverall barplot #first add region to overall_index overall_index<- overall_index |> mutate(region = ifelse( grepl("Argentina|Bolivia|Brazil|Chile|Colombia|Ecuador|Guyana|Paraguay|Peru|Suriname|Uruguay|Venezuela", Country),"South America", ifelse( grepl("Antigua and Barbuda|Bahamas|Barbados|Belize|Canada|Costa Rica|Cuba|Dominica|Dominican Republic|El Salvador|Grenada|Guatemala|Haiti|Honduras|Jamaica|Mexico|Nicaragua|Panama|Saint Kitts and Nevis|Saint Lucia|Saint Vincent and the Grenadines|Trinidad and Tobago|United States",Country),"North America", ifelse( grepl("Albania|Andorra|Armenia|Austria|Azerbaijan|Belarus|Belgium|Bosnia and Herzegovina|Bulgaria|Croatia|Cyprus|Czech Republic|Denmark|Estonia|Finland|France|Georgia|Germany|Greece|Hungary|Iceland|Ireland|Italy|Kazakhstan|Latvia|Liechtenstein|Lithuania|Luxembourg|Malta|Moldova|Monaco|Montenegro|Netherlands|North Macedonia|Norway|Poland|Portugal|Romania|Russia|San Marino|Serbia|SLovakia|Slovenia|Spain|Sweden|Switzerland|Turkey|Ukraine|United Kingdom|Vatican City",Country),"Europe", ifelse( grepl("Afghanistan|Australia|Bahrain|Bangladesh|Bhutan|Brunei|Cambodia|China|Cook Islands|East Timor|Egypt|Fiji|India|Indonesia|Iran|Iran |Iraq|Israel|Japan|Jordan|Kazakhstan|Kiribati|Kuwait|Kyrgyzstan|Laos|Lebanon|Libya|Malaysia|Maldives|Marshall Islands|Micronesia|Mongolia|Myanmar|Nauru|Nepal|New Zealand|Niue|North Korea|Oman|Pakistan|Palau|Palestine|Papua New Guinea|Philippines|Qatar|Samoa|Saudi Arabia|Singapore|Solomon Islands|South Korea|Sri Lanka|Syria|Tajikistan|Thailand|Tonga|Turkmenistan|Tuvalu|Uzbekistan|United Arab Emirates|Vanuatu|Vietnam|Yemen",Country), "Asia Pacific", ifelse(grepl("Algeria|Angola|Benin|Botswana|Burkina Faso|Burundi|Cameroon|Cape Verde|Central African Republic|Chad|Comoros|Congo, Democratic Republic|Congo, Republic|Djibouti|Egypt|Equatorial Guinea|Eritrea|Eswatini|Ethiopia|Gabon|Gambia|Ghana|Guinea|Guinea-Bissau|Ivory Coast|Kenya|Lesotho|Liberia|Libya|Madagascar|Malawi|Mali|Mauritania|Mauritius|Morocco|Mozambique|Namibia|Niger|Nigeria|Rwanda|Sao Tome and Principe|Senegal|Seychelles|Sierra Leone|Somalia|South Africa|South Sudan|Sudan|Tanzania|Togo|Tunisia|Uganda|Zambia|Zimbabwe", Country),"Africa", ifelse(grepl(NA, Country),"NA", "Other"))))))) overall_barplot <- ggplot( overall_index[1:50,], aes( x = reorder(Country, WCI_Score), y = WCI_Score, fill= region, position = "dodge" ) ) overall_barplot+ geom_col(width=0.9)+ coord_flip()+ scale_y_continuous(limits = c(0,60), expand = expansion(mult = c(0, 0)))+ scale_x_discrete(guide = guide_axis(n.dodge=1.75))+ labs(x="Country", y=expression(WCI[overall]*" Score"), fill="Region")+ geom_text(aes(y=WCI_Score, label=sprintf("%0.2f", round(WCI_Score, digits=2)), group=WCI_Score), position = position_dodge(width=1), hjust = -0.05, size=2.5, inherit.aes = TRUE)+ scale_fill_manual(values=c("#984EA3", "#4DAF4A", "#377EB8", "#FF7F00", "#E41A1C"))+ theme(axis.text.x=element_text(size=8), axis.text.y=element_text(size=8), axis.title.y=element_text(size=13), axis.title.x=element_text(size=13)) ggsave("overall_barplot_PLOS.png", scale=1, dpi = 300) #title code: (title=expression("Top 50 countries by "*WCI[overall]*" Score"), x=...) ###ANALYSIS### #Local/not local regressions #regressions below are testing to see if "locals" #rated their country of residence/nationality #differently to non-locals #(I created the local/not-local variable earlier in the script, line ~343) #nationality local/not local local_nat_lm <- index_long local_nat_lm <- local_nat_lm |> select(respID, country, rating, nat_local) |> arrange(respID) local_nat_lm$nat_local <-as.factor(local_nat_lm$nat_local) model1 <- lm(rating~nat_local, data=local_nat_lm) summary(model1) #residence local/not local local_res_lm <- index_long local_res_lm <- local_res_lm |> select(respID, country, rating, res_local) |> arrange(respID) local_res_lm$res_local <-as.factor(local_res_lm$res_local) model2 <- lm(rating~res_local, data=local_res_lm) summary(model2) #Crime type experts #regressions below are testing to see if experts in [crime type] #rate countries in [crime type] differently to non-experts #Technical crime experts (experts n=45) tech_lm <- index_long tech_lm <- tech_lm |> select(respID, country, rating, expertise_crimetype, crimetype) |> group_by(expertise_crimetype) |> mutate( expertise = case_when( (grepl("Technical products / services,", expertise_crimetype)) ~ "tech_expert", (!grepl("Technical products / services,", expertise_crimetype)) ~ "not_tech_expert", .default = NA ) ) tech_lm <- tech_lm |> select(everything()) |> filter(str_detect(crimetype, "technical")) tech_lm$expertise <-as.factor(tech_lm$expertise) model3 <- lm(rating~expertise, data=tech_lm) summary(model3) #Attacks & extortions experts (experts n=55) attack_lm <- index_long attack_lm <- attack_lm |> select(respID, country, rating, expertise_crimetype, crimetype) |> group_by(expertise_crimetype) |> mutate( expertise = case_when( (grepl("Attacks and extortions,", expertise_crimetype)) ~ "attack_expert", (!grepl("Attacks and extortions,", expertise_crimetype)) ~ "not_attack_expert", .default = NA ) ) attack_lm <- attack_lm |> select(everything()) |> filter(str_detect(crimetype, "attack")) attack_lm$expertise <-as.factor(attack_lm$expertise) model4 <- lm(rating~expertise, data=attack_lm) summary(model4) #Cashing out experts (experts n=45) cash_lm <- index_long cash_lm <- cash_lm |> select(respID, country, rating, expertise_crimetype, crimetype) |> group_by(expertise_crimetype) |> mutate( expertise = case_when( (grepl("Cashing out/money laundering,", expertise_crimetype)) ~ "cash_expert", (!grepl("Cashing out/money laundering,", expertise_crimetype)) ~ "not_cash_expert", .default = NA ) ) cash_lm <- cash_lm |> select(everything()) |> filter(str_detect(crimetype, "cash")) cash_lm$expertise <-as.factor(cash_lm$expertise) model5 <- lm(rating~expertise, data=cash_lm) summary(model5) #Data/identity theft experts (experts n=53) data_lm <- index_long data_lm <- data_lm |> select(respID, country, rating, expertise_crimetype, crimetype) |> group_by(expertise_crimetype) |> mutate( expertise = case_when( (grepl("Data/identity theft,", expertise_crimetype)) ~ "data_expert", (!grepl("Data/identity theft,", expertise_crimetype)) ~ "not_data_expert", .default = NA ) ) data_lm <- data_lm |> select(everything()) |> filter(str_detect(crimetype, "data")) data_lm$expertise <-as.factor(data_lm$expertise) model6 <- lm(rating~expertise, data=data_lm) summary(model6) #Scams experts (experts n=45) scams_lm <- index_long scams_lm <- scams_lm |> select(respID, country, rating, expertise_crimetype, crimetype) |> group_by(expertise_crimetype) |> mutate( expertise = case_when( (grepl("Scams,", expertise_crimetype)) ~ "scams_expert", (!grepl("Scams,", expertise_crimetype)) ~ "not_scams_expert", .default = NA ) ) scams_lm <- scams_lm |> select(everything()) |> filter(str_detect(crimetype, "scams")) scams_lm$expertise <-as.factor(scams_lm$expertise) model7 <- lm(rating~expertise, data=scams_lm) summary(model7) #Region experts #regressions below are testing to see if experts in [region] #rate countries from that region differently to non-[region]-experts #Africa experts (experts n=12) africa_lm <- index_long africa_lm <- africa_lm |> select(respID, country, rating, expertise_region) |> group_by(expertise_region) |> mutate( expertise = case_when( (grepl("Africa|EMEA", expertise_region)) ~ "africa_expert", (!grepl("Africa|EMEA", expertise_region)) ~ "not_africa_expert", .default = NA ) ) africa_lm <- africa_lm |> select(everything()) |> filter(str_detect(country, "Algeria|Angola|Benin|Botswana|Burkina Faso|Burundi|Cameroon|Cape Verde|Central African Republic|Chad|Comoros|Congo, Democratic Republic|Congo, Republic|Djibouti|Egypt|Equatorial Guinea|Eritrea|Ethiopia|Gabon|Gambia|Ghana|Guinea|Guinea-Bissau|Ivory Coast|Kenya|Lesotho|Liberia|Libya|Madagascar|Malawi|Mali|Mauritania|Mauritius|Morocco|Mozambique|Namibia|Niger|Nigeria|Rwanda|Sao Tome and Principe|Senegal|Seychelles|Sierra Leone|Somalia|South Africa|South Sudan|Sudan|Swaziland|Tanzania|Togo|Tunisia|Uganda|Zambia|Zimbabwe")) africa_lm$expertise <-as.factor(africa_lm$expertise) model8 <- lm(rating~expertise, data=africa_lm) summary(model8) #Asia Pacific experts (experts n=13) asia_lm <- index_long asia_lm <- asia_lm |> select(respID, country, rating, expertise_region) |> group_by(expertise_region) |> mutate( expertise = case_when( (grepl("Asia Pacific|APAC|EMEA|Eurasian", expertise_region)) ~ "asia_expert", (!grepl("Asia Pacific|APAC|EMEA|Eurasian", expertise_region)) ~ "not_asia_expert", .default = NA ) ) asia_lm <- asia_lm |> select(everything()) |> filter(str_detect(country, "Afghanistan|Australia|Bahrain|Bangladesh|Bhutan|Brunei|Cambodia|China|Cook Islands|East Timor|Egypt|Fiji|India|Indonesia|Iran|Iraq|Israel|Japan|Jordan|Kazakhstan|Kiribati|Kuwait|Kyrgyzstan|Laos|Lebanon|Libya|Malaysia|Maldives|Marshall Islands|Micronesia|Mongolia|Myanmar|Nauru|Nepal|New Zealand|Niue|North Korea|Korea, North|Korea, South|Oman|Pakistan|Palau|Palestine|Papua New Guinea|Philippines|Qatar|Samoa|Saudi Arabia|Singapore|Solomon Islands|South Korea|Sri Lanka|Syria|Tajikistan|Thailand|Tonga|Turkmenistan|Tuvalu|Uzbekistan|United Arab Emirates|Vanuatu|Vietnam|Yemen")) asia_lm$expertise <-as.factor(asia_lm$expertise) model9 <- lm(rating~expertise, data=asia_lm) summary(model9) #Europe experts (experts n=42) europe_lm <- index_long europe_lm <- europe_lm |> select(respID, country, rating, expertise_region) |> group_by(expertise_region) |> mutate( expertise = case_when( (grepl("Europe|BiH|EMEA|Eurasian|ex-USSR|RuNet|UK|EU|Nordics", expertise_region)) ~ "europe_expert", (!grepl("Europe|BiH|EMEA|Eurasian|ex-USSR|RuNet|UK|EU|Nordics", expertise_region)) ~ "not_europe_expert", .default = NA ) ) europe_lm <- europe_lm |> select(everything()) |> filter(str_detect(country, "Albania|Andorra|Armenia|Austria|Azerbaijan|Belarus|Belgium|Bosnia and Herzegovina|Bulgaria|Croatia|Cyprus|Czech Republic|Denmark|Estonia|Finland|France|Georgia|Germany|Greece|Hungary|Iceland|Ireland|Italy|Kazakhstan|Latvia|Liechtenstein|Lithuania|Luxembourg|Malta|Moldova|Monaco|Montenegro|Netherlands|North Macedonia|Norway|Poland|Portugal|Romania|Russia|San Marino|Serbia|SLovakia|Slovenia|Spain|Sweden|Switzerland|Turkey|Ukraine|United Kingdom|Vatican City")) asia_lm$expertise <-as.factor(asia_lm$expertise) model10 <- lm(rating~expertise, data=europe_lm) summary(model10) #north america experts (experts n=3) north_lm <- index_long north_lm <- north_lm |> select(respID, country, rating, expertise_region) |> group_by(expertise_region) |> mutate( expertise = case_when( (grepl("North America", expertise_region)) ~ "north_expert", (!grepl("North America", expertise_region)) ~ "not_north_expert", .default = NA ) ) north_lm <- north_lm |> select(everything()) |> filter(str_detect(country, "Antigua and Barbuda|Bahamas|Barbados|Belize|Canada|Costa Rica|Cuba|Dominica|Dominican Republic|El Salvador|Grenada|Guatemala|Haiti|Honduras|Jamaica|Mexico|Nicaragua|Panama|Saint Kitts and Nevis|Saint Lucia|Saint Vincent and the Grenadines|Trinidad and Tobago|United States")) north_lm$expertise <-as.factor(north_lm$expertise) model11 <- lm(rating~expertise, data=north_lm) summary(model11) #south america experts (experts n=5) south_lm <- index_long south_lm <- south_lm |> select(respID, country, rating, expertise_region) |> group_by(expertise_region) |> mutate( expertise = case_when( (grepl("South America|Latin America", expertise_region)) ~ "south_expert", (!grepl("South America|Latin America", expertise_region)) ~ "not_south_expert", .default = NA ) ) south_lm <- south_lm |> select(everything()) |> filter(str_detect(country, "Argentina|Bolivia|Brazil|Chile|Colombia|Ecuador|Guyana|Paraguay|Peru|Suriname|Uruguay|Venezuela")) south_lm$expertise <-as.factor(south_lm$expertise) model12 <- lm(rating~expertise, data=south_lm) summary(model12) #T-score Tscore <- tech_index[c(1:11, 13:14, 24:25), c(3,8)] Tscore <- full_join(Tscore, attack_index[c(1:11, 13, 23, 28, 36), c(3,8)], by="Country") colnames(Tscore) <- c("Country", "wci_tech", "wci_attack") Tscore <- full_join(Tscore, data_index[c(1:12, 15, 21, 52), c(3,8)], by="Country") colnames(Tscore) <- c("Country", "wci_tech", "wci_attack", "wci_data") Tscore <- full_join(Tscore, WCI_scams[c(1:11, 19, 30, 33, 67), c(3,8)], by="Country") colnames(Tscore) <- c("Country", "wci_tech", "wci_attack", "wci_data", "wci_scams") Tscore <- full_join(Tscore, WCI_cash[c(1:7, 9, 12, 15:16, 18, 25, 48, 76), c(3,8)], by="Country") colnames(Tscore) <- c("country", "wci_tech", "wci_attack", "wci_data", "wci_scams", "wci_cash") Tscore$wci_scams <- Tscore$wci_scams |> replace_na(0) Tscore$wci_cash <- Tscore$wci_cash |> replace_na(0) Tscore$wci_tech <- Tscore$wci_tech*2 Tscore$wci_attack <- Tscore$wci_attack*1 Tscore$wci_data <- Tscore$wci_data*0 Tscore$wci_scams <- Tscore$wci_scams*-1 Tscore$wci_cash <- Tscore$wci_cash*-2 Tscore <- Tscore |> mutate(t_score=c(wci_tech+wci_attack+wci_data+wci_scams+wci_cash)) colnames(Tscore) <- c("Country", "wci_tech", "wci_attack", "wci_data", "wci_scams", "wci_cash", "t_score") Tscore <- full_join(Tscore, overall_index[c(1:15),c(3,8)], by="Country") colnames(Tscore) <- c("country", "wci_tech", "wci_attack", "wci_data", "wci_scams", "wci_cash", "t_score", "wci_score") Tscore$t_score <- round(Tscore$t_score, digits=2) flatplot <- c(0,0,0,0,0,0,0,0,0,0,0,0,0,0,0) #Tscore plot, no color ggplot(data=Tscore, aes(x=t_score, y=flatplot, label=country)) + geom_line()+ geom_point(size=3)+ geom_vline(xintercept=c(0,0), linetype="dotted")+ geom_label_repel(point.size=5, max.overlaps = Inf, min.segment.length = 0)+ labs(title=expression("T-score of the Top 15 "*WCI[overall]*" countries"), y=NULL, x="T-score")+ theme(axis.text.y=element_blank(), axis.ticks.y=element_blank() )+ expand_limits(y = c(-1, 1), x=c(-125, 125)) #Tscore plot, color ggplot(data=Tscore, aes(x=t_score, y=flatplot, label=country)) + geom_line()+ geom_point(size=3, color = ifelse(Tscore$t_score < 0, "red", "black"))+ geom_vline(xintercept=c(0,0), linetype="dotted")+ geom_label_repel(point.size=5, max.overlaps = Inf, min.segment.length = 0)+ labs(y=NULL, x="T-score", color="T-score")+ theme(axis.text.y=element_blank(), axis.ticks.y=element_blank() )+ expand_limits(y = c(-0.2, 0.2), x=c(-140, 100)) ggsave("Tscore_pnas.png", dpi=200) #NOTE TO SELF deleted "heatmap" code bc didnt generate in R in the end #Participant nationality, nationality region, residence, and residence region nationality.region.tab <- index_long$nationality_region |> replace_na("Prefer not to say") nationality.region.tab <- table(nationality.region.tab) nationality.region.tab <- as.data.frame(nationality.region.tab) colnames(nationality.region.tab) <- c("Nationality (region)", "Frequency") nationality.region.tab <- nationality.region.tab |> mutate(Frequency=Frequency/75) |> arrange(desc(Frequency)) nationality.region.tab <- nationality.region.tab |> mutate(Percentage=formattable::percent(Frequency/92)) colnames(nationality.region.tab) <- c("Nationality (region)", "Frequency", "Percentage") nationality.tab <- index_long$nationality |> replace_na("Prefer not to say") nationality.tab <- table(nationality.tab) nationality.tab <- as.data.frame(nationality.tab) colnames(nationality.tab) <- c("Nationality (country)", "Frequency") nationality.tab <- nationality.tab |> mutate(Frequency=Frequency/75) |> arrange(desc(Frequency)) nationality.tab <- nationality.tab |> mutate(Percentage=formattable::percent(Frequency/92)) colnames(nationality.tab) <- c("Nationality (country)", "Frequency", "Percentage") residence.tab <- index_long$residence |> replace_na("Prefer not to say") residence.tab <- table(residence.tab) residence.tab <- as.data.frame(residence.tab) colnames(residence.tab) <- c("Residence (country)", "Frequency") residence.tab <- residence.tab |> mutate(Frequency=Frequency/75) |> arrange(desc(Frequency)) residence.tab <- residence.tab |> mutate(Percentage=formattable::percent(Frequency/92)) colnames(residence.tab) <- c("Residence (country)", "Frequency", "Percentage") residence.region.tab <- index_long$residence_region |> replace_na("Prefer not to say") residence.region.tab <- table(residence.region.tab) residence.region.tab <- as.data.frame(residence.region.tab) colnames(residence.region.tab) <- c("Residence (region)", "Frequency") residence.region.tab <- residence.region.tab |> mutate(Frequency=Frequency/75) |> arrange(desc(Frequency)) residence.region.tab <- residence.region.tab |> mutate(Percentage=formattable::percent(Frequency/92)) colnames(residence.region.tab) <- c("Residence (region)", "Frequency", "Percentage") #SUPPLEMENTARY INFO# ##Expert crimetypes and regions## #30.6.23 expertise_bind <- left_join(expertise_crimetype_long, expertise_region_long) expertise_bind[expertise_bind==""] <- NA expertise_tech <- expertise_bind |> select(respID:`expertise_crimetype`) |> filter(str_detect(`expertise_crimetype`, regex("tech", ignore_case = T))) |> mutate(no.tech.expert = n_distinct(expertise_tech$respID))|> mutate(percentage.tech=formattable::percent(no.tech.expert/92)) |> mutate(tech="tech") expertise_attack <- expertise_bind |> select(respID:`expertise_crimetype`) |> filter(str_detect(`expertise_crimetype`, regex("attack", ignore_case = T)))|> mutate(no.attack.expert = n_distinct(expertise_attack$respID))|> mutate(percentage.attack=formattable::percent(no.attack.expert/92)) |> mutate(attack="attack") expertise_cash <- expertise_bind |> select(respID:`expertise_crimetype`) |> filter(str_detect(`expertise_crimetype`, regex("cash", ignore_case = T)))|> mutate(no.cash.expert = n_distinct(expertise_cash$respID))|> mutate(percentage.cash=formattable::percent(no.cash.expert/92)) |> mutate(cash="cash") expertise_data <- expertise_bind |> select(respID:`expertise_crimetype`) |> filter(str_detect(`expertise_crimetype`, regex("data", ignore_case = T)))|> mutate(no.data.expert = n_distinct(expertise_data$respID))|> mutate(percentage.data=formattable::percent(no.data.expert/92)) |> mutate(data="data") expertise_scams <- expertise_bind |> select(respID:`expertise_crimetype`) |> filter(str_detect(`expertise_crimetype`, regex("scam", ignore_case = T))) |> mutate(no.scam.expert = n_distinct(expertise_scams$respID))|> mutate(percentage.scam=formattable::percent(no.scam.expert/92)) |> mutate(scams="scams") expertise_other <- expertise_bind |> select(respID:`expertise_crimetype`) |> filter(str_detect(`expertise_crimetype`, regex("other", ignore_case = T))) |> mutate(no.other.expert = n_distinct(expertise_other$respID))|> mutate(percentage.other=formattable::percent(no.other.expert/92)) |> mutate(other="other") expertise_none<- expertise_bind |> select(respID:`expertise_crimetype`) |> filter(str_detect(`expertise_crimetype`, regex("NA,", ignore_case = T))) |> mutate(no.none.expert = n_distinct(expertise_none$respID))|> mutate(percentage.none=formattable::percent(no.none.expert/92)) |> mutate(none="none") expertise_crimetype_bind <- left_join(expertise_bind, expertise_attack) expertise_crimetype_bind <- left_join(expertise_crimetype_bind, expertise_cash) expertise_crimetype_bind <- left_join(expertise_crimetype_bind, expertise_data) expertise_crimetype_bind <- left_join(expertise_crimetype_bind, expertise_scams) expertise_crimetype_bind <- left_join(expertise_crimetype_bind, expertise_tech) expertise_crimetype_bind <- left_join(expertise_crimetype_bind, expertise_other) expertise_crimetype_bind <- left_join(expertise_crimetype_bind, expertise_none) view(expertise_crimetype_bind) expertise.crimetype.freqpercent <-expertise_crimetype_bind |> select(no.attack.expert:percentage.none) |> distinct(.keep_all=T) expertise.crimetype.freqpercent <- pivot_longer(expertise.crimetype.freqpercent, cols = c(no.data.expert, no.attack.expert, no.cash.expert, no.scam.expert, no.tech.expert, no.other.expert, no.none.expert), names_to="expertise", values_to="number experts") expertise.crimetype.freqpercent <- pivot_longer(expertise.crimetype.freqpercent, cols = c(percentage.cash, percentage.data, percentage.scam, percentage.tech, percentage.attack, percentage.other, percentage.none), names_to="exp percentage", values_to="percentage total") view(expertise.crimetype.freqpercent) expert.attack <- expertise_crimetype_bind |> distinct(respID, .keep_all = T) |> summarise("no.attack.experts"=sum(!is.na(expertise_crimetype_bind$no.attack.expert))) expert.data <- expertise_crimetype_bind |> distinct(respID, .keep_all = T) |> summarise("no.data.experts"=sum(!is.na(expertise_crimetype_bind$no.data.expert))) expert.cash <- expertise_crimetype_bind |> distinct(respID, .keep_all = T) |> summarise("no.cash.experts"=sum(!is.na(expertise_crimetype_bind$no.cash.expert))) expert.scams <- expertise_crimetype_bind |> distinct(respID, .keep_all = T) |> summarise("no.scam.experts"=sum(!is.na(expertise_crimetype_bind$no.scam.expert))) expert.tech <- expertise_crimetype_bind |> distinct(respID, .keep_all = T) |> summarise("no.tech.experts"=sum(!is.na(expertise_crimetype_bind$no.tech.expert))) expert.other<- expertise_crimetype_bind |> distinct(respID, .keep_all = T) |> summarise("no.other.experts"=sum(!is.na(expertise_crimetype_bind$no.other.expert))) expert.none<- expertise_crimetype_bind |> distinct(respID, .keep_all = T) |> summarise("no.none.experts"=sum(!is.na(expertise_crimetype_bind$no.none.expert))) expert.crimebind <- bind_rows(expert.attack, expert.data, expert.cash, expert.scams, expert.tech, expert.other, expert.none) expert.crimebind <- pivot_longer(expert.crimebind, cols = c(no.attack.experts:no.none.experts), names_to ="expertise", values_to = "number.experts" ) expertise.crimetype.freqpercent <- expert.crimebind |> drop_na(number.experts) |> mutate(Percentage.of.respondents=formattable::percent(number.experts/92)) |> mutate(Percentage.of.experts=formattable::percent(number.experts/85)) |> arrange(expertise) expertise.crimetype.freqpercent[expertise.crimetype.freqpercent == "no.attack.experts"] <- "Attacks and extortion" expertise.crimetype.freqpercent[expertise.crimetype.freqpercent == "no.data.experts"] <- "Data and identity theft" expertise.crimetype.freqpercent[expertise.crimetype.freqpercent == "no.cash.experts"] <- "Cashing out and money laundering" expertise.crimetype.freqpercent[expertise.crimetype.freqpercent == "no.scam.experts"] <- "Scams" expertise.crimetype.freqpercent[expertise.crimetype.freqpercent == "no.tech.experts"] <- "Technical products and services" expertise.crimetype.freqpercent[expertise.crimetype.freqpercent == "no.none.experts"] <- "None" expertise.crimetype.freqpercent[expertise.crimetype.freqpercent == "no.other.experts"] <- "Other" colnames(expertise.crimetype.freqpercent) <- c("Expertise", "No.", "% respondent pool", "% expert pool") expertise.crimetype.freqpercent <- expertise.crimetype.freqpercent[c("Expertise", "No.", "% expert pool", "% respondent pool")] expertise.crimetype.freqpercent <- expertise.crimetype.freqpercent |> arrange(desc(No.)) |> distinct(Expertise, .keep_all = T) ##Expert regions dfs #expertise region df expert_asiapacific_df<- group_by(index_long, expertise_region) |> filter(str_detect(expertise_region, "Afghanistan|Australia|Bahrain|Bangladesh|Bhutan|Brunei|Cambodia|China|Cook Islands|East Timor|Egypt|Fiji|India|Indonesia|Iran|Iran |Iraq|Israel|Japan|Jordan|Kazakhstan|Kiribati|Kuwait|Kyrgyzstan|Laos|Lebanon|Libya|Malaysia|Maldives|Marshall Islands|Micronesia|Mongolia|Myanmar|Nauru|Nepal|New Zealand|Niue|North Korea|Oman|Pakistan|Palau|Palestine|Papua New Guinea|Philippines|Qatar|Samoa|Saudi Arabia|Singapore|Solomon Islands|South Korea|Sri Lanka|Syria|Tajikistan|Thailand|Tonga|Turkmenistan|Tuvalu|Uzbekistan|United Arab Emirates|Vanuatu|Vietnam|Yemen|APAC|EMEA|Eurasian|HK|Asia Pacific|Asia")) expert_africa_df <- group_by(index_long, expertise_region) |> filter(str_detect(expertise_region, "Algeria|Angola|Benin|Botswana|Burkina Faso|Burundi|Cameroon|Cape Verde|Central African Republic|Chad|Comoros|Congo, Democratic Republic|Congo, Republic|Djibouti|Egypt|Equatorial Guinea|Eritrea|Eswatini|Ethiopia|Gabon|Gambia|Ghana|Guinea|Guinea-Bissau|Ivory Coast|Kenya|Lesotho|Liberia|Libya|Madagascar|Malawi|Mali|Mauritania|Mauritius|Morocco|Mozambique|Namibia|Niger|Nigeria|Rwanda|Sao Tome and Principe|Senegal|Seychelles|Sierra Leone|Somalia|South Africa|South Sudan|Sudan|Swaziland|Tanzania|Togo|Tunisia|Uganda|Zambia|Zimbabwe|EMEA|Africa|Congo")) expert_europe_df<- group_by(index_long, expertise_region) |> filter(str_detect(expertise_region,"Albania|Andorra|Armenia|Austria|Azerbaijan|Belarus|Belgium|Bosnia and Herzegovina|Bulgaria|Croatia|Cyprus|Czech Republic|Denmark|Estonia|Finland|France|Georgia|Germany|Greece|Hungary|Iceland|Ireland|Italy|Kazakhstan|Latvia|Liechtenstein|Lithuania|Luxembourg|Malta|Moldova|Monaco|Montenegro|Netherlands|North Macedonia|Norway|Poland|Portugal|Romania|Russia|San Marino|Serbia|SLovakia|Slovenia|Spain|Sweden|Switzerland|Turkey|Ukraine|United Kingdom|Vatican City|BiH|EMEA|Eurasian|ex-USSR|RuNet|UK|EU|Nordics|Eastern Europe|Europe")) expert_namer_df <- group_by(index_long, expertise_region) |> filter(str_detect(expertise_region, "Antigua and Barbuda|Bahamas|Barbados|Belize|Canada|Costa Rica|Cuba|Dominica|Dominican Republic|El Salvador|Grenada|Guatemala|Haiti|Honduras|Jamaica|Mexico|Nicaragua|Panama|Saint Kitts and Nevis|Saint Lucia|Saint Vincent and the Grenadines|Trinidad and Tobago|United States|North America")) expert_samer_df <- group_by(index_long, expertise_region) |> filter(str_detect(expertise_region, "South America|Argentina|Bolivia|Brazil|Chile|Colombia|Ecuador|Guyana|Paraguay|Peru|Suriname|Uruguay|Venezuela|Latin America|LATAM")) expert_other_region_df <- group_by(index_long, expertise_region) |> filter(!str_detect(expertise_region, "Afghanistan|Australia|Bahrain|Bangladesh|Bhutan|Brunei|Cambodia|China|Cook Islands|East Timor|Egypt|Fiji|India|Indonesia|Iran|Iran |Iraq|Israel|Japan|Jordan|Kazakhstan|Kiribati|Kuwait|Kyrgyzstan|Laos|Lebanon|Libya|Malaysia|Maldives|Marshall Islands|Micronesia|Mongolia|Myanmar|Nauru|Nepal|New Zealand|Niue|North Korea|Oman|Pakistan|Palau|Palestine|Papua New Guinea|Philippines|Qatar|Samoa|Saudi Arabia|Singapore|Solomon Islands|South Korea|Sri Lanka|Syria|Tajikistan|Thailand|Tonga|Turkmenistan|Tuvalu|Uzbekistan|United Arab Emirates|Vanuatu|Vietnam|Yemen|APAC|EMEA|Eurasian|HK|Asia Pacific|Asia|Algeria|Angola|Benin|Botswana|Burkina Faso|Burundi|Cameroon|Cape Verde|Central African Republic|Chad|Comoros|Congo, Democratic Republic|Congo, Republic|Djibouti|Egypt|Equatorial Guinea|Eritrea|Eswatini|Ethiopia|Gabon|Gambia|Ghana|Guinea|Guinea-Bissau|Ivory Coast|Kenya|Lesotho|Liberia|Libya|Madagascar|Malawi|Mali|Mauritania|Mauritius|Morocco|Mozambique|Namibia|Niger|Nigeria|Rwanda|Sao Tome and Principe|Senegal|Seychelles|Sierra Leone|Somalia|South Africa|South Sudan|Sudan|Swaziland|Tanzania|Togo|Tunisia|Uganda|Zambia|Zimbabwe|EMEA|Africa|Congo|Albania|Andorra|Armenia|Austria|Azerbaijan|Belarus|Belgium|Bosnia and Herzegovina|Bulgaria|Croatia|Cyprus|Czech Republic|Denmark|Estonia|Finland|France|Georgia|Germany|Greece|Hungary|Iceland|Ireland|Italy|Kazakhstan|Latvia|Liechtenstein|Lithuania|Luxembourg|Malta|Moldova|Monaco|Montenegro|Netherlands|North Macedonia|Norway|Poland|Portugal|Romania|Russia|San Marino|Serbia|SLovakia|Slovenia|Spain|Sweden|Switzerland|Turkey|Ukraine|United Kingdom|Vatican City|BiH|EMEA|Eurasian|ex-USSR|RuNet|UK|EU|Nordics|Eastern Europe|Europe|Antigua and Barbuda|Bahamas|Barbados|Belize|Canada|Costa Rica|Cuba|Dominica|Dominican Republic|El Salvador|Grenada|Guatemala|Haiti|Honduras|Jamaica|Mexico|Nicaragua|Panama|Saint Kitts and Nevis|Saint Lucia|Saint Vincent and the Grenadines|Trinidad and Tobago|United States|North America|Antigua and Barbuda|Bahamas|Barbados|Belize|Canada|Costa Rica|Cuba|Dominica|Dominican Republic|El Salvador|Grenada|Guatemala|Haiti|Honduras|Jamaica|Mexico|Nicaragua|Panama|Saint Kitts and Nevis|Saint Lucia|Saint Vincent and the Grenadines|Trinidad and Tobago|United States|North America|South America|Argentina|Bolivia|Brazil|Chile|Colombia|Ecuador|Guyana|Paraguay|Peru|Suriname|Uruguay|Venezuela|Latin America|LATAM|APAC|EMEA|Eurasian|Eurasian|Nordics")) expert_africa_df <- expert_africa_df |> ungroup() |> select(respID, expertise_region) |> mutate(no.africa.expert = n_distinct(respID))|> mutate(percentage.africa=formattable::percent(no.africa.expert/92)) |> mutate(africa="africa")|> distinct(respID, .keep_all = T) expert_namer_df <- expert_namer_df |> ungroup() |> select(respID, expertise_region) |> mutate(no.namer.expert = n_distinct(respID))|> mutate(percentage.namer=formattable::percent(no.namer.expert/92)) |> mutate(northamerica="northamerica") |> distinct(respID, .keep_all = T) expert_europe_df <- expert_europe_df |> ungroup() |> select(respID, expertise_region) |> mutate(no.europe.expert = n_distinct(respID))|> mutate(percentage.europe=formattable::percent(no.europe.expert/92)) |> mutate(europe="europe")|> distinct(respID, .keep_all = T) expert_samer_df <- expert_samer_df |> ungroup() |> select(respID, expertise_region) |> mutate(no.samer.expert = n_distinct(respID))|> mutate(percentage.samer=formattable::percent(no.samer.expert/92)) |> mutate(southamerica="southamerica")|> distinct(respID, .keep_all = T) expert_asiapacific_df <- expert_asiapacific_df |> ungroup() |> select(respID, expertise_region) |> mutate(no.asiapacific.expert = n_distinct(respID))|> mutate(percentage.asiapacific=formattable::percent(no.asiapacific.expert/92)) |> mutate(asiapacific="asiapacific")|> distinct(respID, .keep_all = T) expert_other_region_df <- expert_other_region_df |> ungroup() |> select(respID, expertise_region) |> mutate(no.otherregion.expert = n_distinct(respID))|> mutate(percentage.otherregion=formattable::percent(no.otherregion.expert/92)) |> mutate(otherregion="otherregion")|> distinct(respID, .keep_all = T) expertise_none_region<- expertise_bind |> select(respID,`expertise_region`) |> distinct(respID, .keep_all = T) |> filter(is.na(`expertise_region`))|> mutate(no.noneregion.expert = n_distinct(expertise_none_region$respID))|> mutate(percentage.noneregion=formattable::percent(no.noneregion.expert/92)) |> mutate(noneregion="noneregion")|> distinct(respID, .keep_all = T) expertise_region_bind <- left_join(expertise_bind, expert_africa_df) expertise_region_bind <- left_join(expertise_region_bind, expert_namer_df, by="respID") expertise_region_bind <- left_join(expertise_region_bind, expert_samer_df, by="respID") expertise_region_bind <- left_join(expertise_region_bind, expert_asiapacific_df, by="respID") expertise_region_bind <- left_join(expertise_region_bind, expert_europe_df, by="respID") expertise_region_bind <- left_join(expertise_region_bind, expertise_none_region, by="respID") expertise_region_bind <- left_join(expertise_region_bind, expert_other_region_df, by="respID") view(expertise_region_bind) expert.africa <- expertise_region_bind |> distinct(respID) |> summarise("no.africa.experts"=sum(!is.na(expertise_region_bind$no.africa.expert))) expert.asiapacific <- expertise_region_bind |> distinct(respID) |> summarise("no.asiapacific.experts"=sum(!is.na(expertise_region_bind$no.asiapacific.expert))) expert.europe <- expertise_region_bind |> distinct(respID) |> summarise("no.europe.experts"=sum(!is.na(expertise_region_bind$no.europe.expert))) expert.namer <- expertise_region_bind |> distinct(respID) |> summarise("no.namer.expert"=sum(!is.na(expertise_region_bind$no.namer.expert))) expert.samer <- expertise_region_bind |> distinct(respID) |> summarise("no.samer.experts"=sum(!is.na(expertise_region_bind$no.samer.expert))) expert.noneregion<- expertise_region_bind |> distinct(respID) |> summarise("no.noneregion.experts"=sum(!is.na(expertise_region_bind$no.noneregion.expert))) expert.otherregion<- expertise_region_bind |> distinct(respID) |> summarise("no.otherregion.experts"=sum(!is.na(expertise_region_bind$no.otherregion.expert))) expert.regions.all <- bind_rows(expert.asiapacific, expert.africa, expert.europe, expert.namer, expert.noneregion, expert.samer, expert.otherregion) expert.regions.all <- expert.regions.all |> select(no.asiapacific.experts:no.otherregion.experts) expert.regions.all <- pivot_longer(expert.regions.all, cols = c(no.asiapacific.experts:no.otherregion.experts), names_to ="expertise", values_to = "number.experts" ) expertise.regions.freqpercent <- expert.regions.all |> drop_na(number.experts) |> mutate(Percentage.of.respondents=formattable::percent(number.experts/92)) |> mutate(Percentage.of.experts=formattable::percent(number.experts/50)) |> arrange(expertise) expertise.regions.freqpercent[expertise.regions.freqpercent == "no.africa.experts"] <- "Africa" expertise.regions.freqpercent[expertise.regions.freqpercent == "no.asiapacific.experts"] <- "Asia Pacific" expertise.regions.freqpercent[expertise.regions.freqpercent == "no.europe.experts"] <- "Europe" expertise.regions.freqpercent[expertise.regions.freqpercent == "no.namer.expert"] <- "North America" expertise.regions.freqpercent[expertise.regions.freqpercent == "no.noneregion.experts"] <- "None" expertise.regions.freqpercent[expertise.regions.freqpercent == "no.samer.experts"] <- "South America" expertise.regions.freqpercent[expertise.regions.freqpercent == "no.otherregion.experts"] <- "Other" colnames(expertise.regions.freqpercent) <- c("Expertise", "No.", "% respondent pool", "% expert pool") expertise.regions.freqpercent <- expertise.regions.freqpercent |> arrange(desc("No.")) expertise.regions.freqpercent <- expertise.regions.freqpercent[c("Expertise", "No.", "% expert pool", "% respondent pool")] |> distinct(Expertise, .keep_all = T) ##Expert boxplots## #expert_tech_df<- group_by(index_long, expertise_crimetype) |> #filter(str_detect(expertise_crimetype, "Technical")) #expert_tech_df$crimetype <- "Technical" #expert_attack_df<- group_by(index_long, expertise_crimetype) |> #filter(str_detect(expertise_crimetype, "Attack")) #expert_scam_df<- group_by(index_long, expertise_crimetype) |> #filter(str_detect(expertise_crimetype, "Scam")) #expert_data_df<- group_by(index_long, expertise_crimetype) |> #filter(str_detect(expertise_crimetype, "Data")) #expert_cash_df<- group_by(index_long, expertise_crimetype) |> #filter(str_detect(expertise_crimetype, "Cash")) #expert_other_crimetype_df<- group_by(index_long, expertise_crimetype) |> #filter(!str_detect(expertise_crimetype, "Attack|Scam|Cash|Data|Technical")) #plots #expert_tech_plot <- #ggplot(expert_tech_df, #aes(x=crimetype, # y=rating)) #expert_tech_plot+ # geom_boxplot(aes(fill=crimetype))+ # coord_flip()+ # labs(title="How technical experts rated different crime types", y="Rating", x="Crime type")+ #theme(legend.position = "none") #expert_attack_plot <- ggplot(expert_attack_df, # aes(x=crimetype, # y=rating)) #expert_attack_plot+ # geom_boxplot(aes(fill=crimetype))+ # coord_flip()+ # labs(title="How attack experts rated different crime types", y="Rating", x="Crime type")+ # theme(legend.position="none") #expert_data_plot <- ggplot(expert_data_df, # aes(x=crimetype, # y=rating)) #expert_data_plot+ # geom_boxplot(aes(fill=crimetype))+ # coord_flip()+ # labs(title="How data experts rated different crime types", y="Rating", x="Crime type")+ # theme(legend.position="none") #expert_scam_plot <- ggplot(expert_scam_df, # aes(x=crimetype, # y=rating)) #expert_scam_plot+ # geom_boxplot(aes(fill=crimetype))+ # coord_flip()+ # labs(title="How scam experts rated different crime types", y="Rating", x="Crime type")+ # theme(legend.position="none") #expert_cash_plot <- ggplot(expert_cash_df, # aes(x=crimetype, # y=rating)) #expert_cash_plot+ # geom_boxplot(aes(fill=crimetype))+ # coord_flip()+ # labs(title="How cashing out experts rated different crime types", y="Rating", x="Crime type")+ # theme(legend.position="none") #expert_other_plot <- ggplot(expert_other_crimetype_df, # aes(x=crimetype, # y=rating)) #expert_other_plot+ # geom_violin(aes(fill=crimetype))+ # coord_flip()+ # labs(title="How other experts rated different crime types", y="Rating", x="Crime type")+ # theme(legend.position="none") #plots for region expertise #expert_europe_plot <- # ggplot(na.omit(expert_europe_df), # aes(x=country_region, # y=rating)) #expert_europe_plot+ # geom_violin(aes(fill=country_region))+ # coord_flip()+ # labs(title="How European experts rated different regions", y="Rating", x=NULL)+ # theme(legend.position = "none") #expert_africa_plot <- ggplot(na.omit(expert_africa_df), # aes(x=country_region, # y=rating)) #expert_africa_plot+ #geom_violin(aes(fill=country_region))+ # coord_flip()+ # labs(title="How African experts rated different regions", y="Rating", x=NULL)+ # theme(legend.position="none") #expert_asiapacific_plot <- ggplot(na.omit(expert_asiapacific_df), # aes(x=country_region, # y=rating)) #expert_asiapacific_plot+ # geom_violin(aes(fill=country_region))+ # coord_flip()+ # labs(title="How Asia Pacific experts rated different regions", y="Rating", x=NULL)+ # theme(legend.position="none") #expert_namer_plot <- ggplot(expert_namer_df, # aes(x=country_region, # y=rating)) #expert_namer_plot+ # geom_violin(aes(fill=country_region))+ # coord_flip()+ # labs(title="How North American experts rated different\n regions", y="Rating", x=NULL)+ # scale_x_discrete(limits=c("North America", "Europe", "Asia Pacific", "Africa", "South America"))+ # theme(legend.position="none") #expert_samer_plot <- ggplot(expert_samer_df, # aes(x=country_region, # y=rating)) #expert_samer_plot+ # geom_violin(aes(fill=country_region))+ # coord_flip()+ # labs(title="How South American experts rated different\n regions", y="Rating", x=NULL)+ # scale_x_discrete(limits=c("North America", "Europe", "Asia Pacific", "Africa", "South America"))+ # theme(legend.position="none")+ # geom_jitter(position=position_jitter(0.02)) #testing survey fatigue #do overall/average scores for each country increase/decrease as survey progressed? #row bind in survey order, then only keep countries that are repeated in each crimetype, then Mann-Kendall & Sen slope test fatigue_df <- bind_rows(tech_index, attack_index) fatigue_df <- bind_rows(fatigue_df, data_index) fatigue_df <- bind_rows(fatigue_df, scams_index) fatigue_df <- bind_rows(fatigue_df, cash_index) fatigue_df <- fatigue_df |> group_by(Country) |> filter( n() > 1 ) |> view() fatigue_df <- as.data.frame(fatigue_df) #MK trend & sens slope test install.packages("trend") library(trend) mk.test(fatigue_df$Overall) sens.slope(fatigue_df$Overall) #russia trend test, as an experiment #fatigue_df_russ <- fatigue_df |> # filter(str_detect(Country, "Russia")) #mk.test(fatigue_df_russ$Overall) #sens.slope(fatigue_df_russ$Overall) #fatigue test all countries, not just ones nominated more than once fatigue_df_all <- bind_rows(tech_index, attack_index) fatigue_df_all <- bind_rows(fatigue_df_all, data_index) fatigue_df_all <- bind_rows(fatigue_df_all, scams_index) fatigue_df_all <- bind_rows(fatigue_df_all, cash_index) mk.test(fatigue_df_all$Overall) sens.slope(fatigue_df_all$Overall) #insufficient effort response test install.packages("careless") library(careless) careless_df <- index_import |> select(where(is.numeric)) careless_long <- longstring(careless_df, avg = FALSE) careless_avg <- longstring(careless_df, avg = TRUE) boxplot(careless_avg$longstr) #produce a boxplot of the longstring index boxplot(careless_avg$avgstr) boxplot(careless_avg$longstr)$stats boxplot(careless_avg$avgstr)$stats #intra-individual response variability IRV test # calculate the irv over all items irv_total <- irv(careless_df) #calculate the irv over all items + calculate the irv for each quarter of the questionnaire irv_split <- irv(careless_df, split = TRUE, num.split = 5) irv_split$respID <- 1:92 boxplot(irv_total)$stats boxplot(irv_split$irv5) #produce a boxplot of the IRV for the fifth category boxplot(irv_split$irv5)$stats boxplot(irv_split$irv4)$stats boxplot(irv_split$irv3)$stats boxplot(irv_split$irv2)$stats boxplot(irv_split$irv1)$stats boxplot(irv_split$irv1, irv_split$irv2, irv_split$irv3, irv_split$irv4, irv_split$irv5)$stats #ggplot(irv_split_long, aes(x = reorder(irv, progression, FUN = median), y = irv)) + geom_boxplot() #irv_split_long <- pivot_longer(irv_split, cols=2:6, names_to = "progression", values_to = "irv") #irv_lm <- lm(irv ~ progression, data = irv_split_long) #summary(irv_lm) #ggplot(irv_split_long, aes("progression", "irv"))+ # geom_point()+ # geom_smooth(method="lm", formula = irv~progression, se=FALSE)+ # labs(title="IRV as survey progressed", x="Survey progression", y="IRV") #ggplot(irv_split_long, aes(x = progression, y = irv)) + # geom_point() + # geom_abline(slope = coef(irv_lm), # intercept = coef(irv_lm)[["(Intercept)"]]) #standard deviation of all ratings #rating_sd <- unlist(index_long[,17]) #sd(rating_sd, na.rm=TRUE) #check list of dependencies# #install.packages("renv") #renv::dependencies() |> # view() save.image(file = "wci_df_reprex.Rdata")