Immigrant employment rates in Scandinavia: predictive validity of national IQs and Muslim percentages
Emil O. W. Kirkegaard
About
Data are from a recent Danish report published by the Danish statistics agency (DST). They collaborated with the Swedish and Norwegian agencies to gather comparable data for employment rates for a select group of 11 countries of origin. These data are combined with national IQs and Muslim %’s to see how predictable employment rates are.
Initialize
library(pacman)
p_load(readr, ggplot2, kirkegaard, dplyr, tidyr, lsr)
options(digits = 2, width = 300)d = read_csv("data/employtment_rates.csv")## Parsed with column specification:
## cols(
## Origin = col_character(),
## Host = col_character(),
## Employment_men = col_integer(),
## Employment_women = col_integer(),
## Employment_total = col_integer()
## )#translate
d$Origin %<>% pu_translate() %>% pu_translate(reverse = T)
d$Host %<>% pu_translate() %>% pu_translate(reverse = T)
#wide version
tidyr::spread(d[c("Origin", "Host", "Employment_total")], key = Host, value = Employment_total) %>%
select(-Origin) %>%
wtd.cors() %>%
MAT_half() %>%
averages()## arithmetic geometric harmonic mode median trimmed midrange
## 0.86 0.85 0.85 0.94 0.86 0.86 0.85#longer version for sex regression
d2 = d %>% select(-Employment_total) %>%
#gather sexs
gather(key = sex, value = Employment, Employment_men, Employment_women)
#fix text
d2$sex %<>% str_replace("Employment_", "")
#national data
mega = read_csv("data/Megadataset_v2.0p.csv") %>% mutate(iso = X1, IQ = LV2012estimatedIQ, Muslim = IslamPewResearch2010)## Parsed with column specification:
## cols(
## .default = col_double(),
## X1 = col_character(),
## EthnicHeterogenityVanhanen2012 = col_integer(),
## EthnicConflictVanhanen2012 = col_integer(),
## SlowTimePrefWangetal2011 = col_integer(),
## Math00Mean = col_integer(),
## Math00SD = col_integer(),
## Read00Mean = col_integer(),
## Read00SD = col_integer(),
## Sci00Mean = col_integer(),
## Sci00SD = col_integer(),
## Math03Mean = col_integer(),
## Math03SD = col_integer(),
## Read03Mean = col_integer(),
## Read03SD = col_integer(),
## Sci03Mean = col_integer(),
## Sci03SD = col_integer(),
## PS03Mean = col_integer(),
## PS03SD = col_integer(),
## Read09.Native = col_integer(),
## Read09.1g = col_integer()
## # ... with 102 more columns
## )## See spec(...) for full column specifications.mega_all = mega
#merge mega and employment data
mega = dplyr::left_join(d %>% mutate(Origin = pu_translate(Origin)) %>% rename(iso = Origin), mega[c("iso", "Muslim", "IQ")], by = "iso")## Warning: Column `iso` has different attributes on LHS and RHS of joinFigure of data
ggplot(d, aes(Origin, Employment_total/100, fill = Host)) +
geom_bar(stat = "identity", position = "dodge") +
theme_bw() +
theme(axis.text.x = element_text(angle = -20, hjust = 0)) +
scale_y_continuous("Employment rate (both sexes)", labels = scales::percent)
GG_save("figures/employment_rate.png")Regression
#sexs together
fit = lm(Employment_total ~ Origin + Host, data = d)
MOD_summary(fit, kfold = F)## The model data contains characters. These were automatically converteed but you should probably do this before calling this function.
## The model data contains characters. These were automatically converteed but you should probably do this before calling this function.##
## ---- Model summary ----
## Model coefficients
## Beta SE CI_lower CI_upper
## Origin: Afghanistan 0.00 NA NA NA
## Origin: Bosnia & Herzegovina 1.10 0.30 0.4799 1.72
## Origin: China 0.62 0.30 -0.0015 1.23
## Origin: Denmark 2.00 0.43 1.1077 2.89
## Origin: Germany 1.35 0.30 0.7303 1.97
## Origin: Iran 0.62 0.30 -0.0015 1.23
## Origin: Iraq -0.13 0.30 -0.7525 0.48
## Origin: Norway 1.74 0.43 0.8451 2.63
## Origin: Poland 1.48 0.30 0.8651 2.10
## Origin: Somalia -0.89 0.30 -1.5036 -0.27
## Origin: Sweden 2.15 0.43 1.2600 3.05
## Origin: Syria -1.25 0.30 -1.8694 -0.63
## Origin: Thailand 1.10 0.30 0.4799 1.72
## Origin: Turkey 0.64 0.30 0.0177 1.25
## Host: Denmark 0.00 NA NA NA
## Host: Norway 0.55 0.15 0.2288 0.87
## Host: Sweden 0.25 0.15 -0.0705 0.57
##
##
## Model meta-data
## outcome N df R2 R2-adj. R2-cv
## 1 Employment_total 36 20 0.92 0.87 NA
##
##
## Etas from analysis of variance
## Eta Eta_partial
## Origin 0.94 0.96
## Host 0.22 0.62#sexs apart
fit_sex = lm(Employment ~ Origin + Host + sex, data = d2)
fit_sex_sum = MOD_summary(fit_sex, kfold = F) %T>% print()## The model data contains characters. These were automatically converteed but you should probably do this before calling this function.
## The model data contains characters. These were automatically converteed but you should probably do this before calling this function.
## The model data contains characters. These were automatically converteed but you should probably do this before calling this function.##
## ---- Model summary ----
## Model coefficients
## Beta SE CI_lower CI_upper
## Origin: Afghanistan 0.000 NA NA NA
## Origin: Bosnia & Herzegovina 1.143 0.215 0.712 1.57
## Origin: China 0.726 0.215 0.295 1.16
## Origin: Denmark 1.994 0.311 1.370 2.62
## Origin: Germany 1.397 0.215 0.966 1.83
## Origin: Iran 0.689 0.215 0.258 1.12
## Origin: Iraq -0.054 0.215 -0.485 0.38
## Origin: Norway 1.761 0.311 1.138 2.38
## Origin: Poland 1.478 0.215 1.047 1.91
## Origin: Somalia -0.744 0.215 -1.175 -0.31
## Origin: Sweden 2.095 0.311 1.472 2.72
## Origin: Syria -1.107 0.215 -1.538 -0.68
## Origin: Thailand 1.116 0.215 0.685 1.55
## Origin: Turkey 0.635 0.215 0.204 1.07
## Host: Denmark 0.000 NA NA NA
## Host: Norway 0.505 0.112 0.280 0.73
## Host: Sweden 0.252 0.112 0.027 0.48
## sex: men 0.000 NA NA NA
## sex: women -0.500 0.088 -0.676 -0.32
##
##
## Model meta-data
## outcome N df R2 R2-adj. R2-cv
## 1 Employment 72 55 0.89 0.86 NA
##
##
## Etas from analysis of variance
## Eta Eta_partial
## Origin 0.89 0.94
## Host 0.20 0.52
## sex 0.25 0.61fit_sex_sum %>% .[[1]] %>% write_clipboard()## Beta SE CI lower CI upper
## Origin: Afghanistan 0.00
## Origin: Bosnia & Herzegovina 1.14 0.22 0.71 1.57
## Origin: China 0.73 0.22 0.29 1.16
## Origin: Denmark 1.99 0.31 1.37 2.62
## Origin: Germany 1.40 0.22 0.97 1.83
## Origin: Iran 0.69 0.22 0.26 1.12
## Origin: Iraq -0.05 0.22 -0.49 0.38
## Origin: Norway 1.76 0.31 1.14 2.38
## Origin: Poland 1.48 0.22 1.05 1.91
## Origin: Somalia -0.74 0.22 -1.17 -0.31
## Origin: Sweden 2.10 0.31 1.47 2.72
## Origin: Syria -1.11 0.22 -1.54 -0.68
## Origin: Thailand 1.12 0.22 0.68 1.55
## Origin: Turkey 0.63 0.22 0.20 1.07
## Host: Denmark 0.00
## Host: Norway 0.50 0.11 0.28 0.73
## Host: Sweden 0.25 0.11 0.03 0.48
## sex: men 0.00
## sex: women -0.50 0.09 -0.68 -0.32#missing cells
length(unique(d$Origin))## [1] 14length(unique(d$Host))## [1] 3length(unique(d$Origin)) * length(unique(d$Host))## [1] 42nrow(d) / (length(unique(d$Origin)) * length(unique(d$Host)))## [1] 0.86PISA data comparison
#data
pisa = read_csv("data/pisa_2006.csv")## Warning: Duplicated column names deduplicated: 'Germany' => 'Germany_1' [10]## Parsed with column specification:
## cols(
## .default = col_integer(),
## Origin = col_character()
## )## See spec(...) for full column specifications.#long format
pisa_long = gather(pisa, key = Host, value = PISA, -Origin) %>% na.omit
#regress
fit_pisa = lm(PISA ~ Origin + Host, data = pisa_long)
fit_pisa %>% summary##
## Call:
## lm(formula = PISA ~ Origin + Host, data = pisa_long)
##
## Residuals:
## Min 1Q Median 3Q Max
## -39.05 -2.13 0.00 2.42 45.34
##
## Coefficients: (22 not defined because of singularities)
## Estimate Std. Error t value Pr(>|t|)
## (Intercept) 353.07 46.78 7.55 3.3e-09 ***
## OriginArgentina 37.93 53.33 0.71 0.48107
## OriginAustralia 94.28 34.17 2.76 0.00870 **
## OriginAustria 142.67 20.41 6.99 1.9e-08 ***
## OriginAzerbaijan 28.93 53.33 0.54 0.59050
## OriginBangladesh 76.02 38.34 1.98 0.05427 .
## OriginBelarus 125.52 46.35 2.71 0.00992 **
## OriginBelgium 157.04 25.46 6.17 2.7e-07 ***
## OriginBosnia Herzegovina 72.93 22.05 3.31 0.00200 **
## OriginBrazil 86.78 36.87 2.35 0.02357 *
## OriginBulgaria 80.93 53.33 1.52 0.13701
## OriginCanada 180.93 53.33 3.39 0.00157 **
## OriginCap Verde 7.10 32.49 0.22 0.82803
## OriginChile 84.93 53.33 1.59 0.11915
## OriginChina 105.01 24.35 4.31 0.00010 ***
## OriginColombia 34.93 53.33 0.65 0.51623
## OriginCroatia 70.33 23.89 2.94 0.00537 **
## OriginCzech Republic 189.68 30.18 6.29 1.9e-07 ***
## OriginDenmark 125.61 31.60 3.97 0.00029 ***
## OriginEstonia 27.04 42.77 0.63 0.53083
## OriginFinland 153.04 42.77 3.58 0.00092 ***
## OriginFrance 107.35 19.93 5.39 3.4e-06 ***
## OriginGermany 154.47 18.30 8.44 2.0e-10 ***
## OriginGreece 44.99 23.92 1.88 0.06731 .
## OriginHong Kong 188.93 53.33 3.54 0.00102 **
## OriginHungary 181.68 30.18 6.02 4.4e-07 ***
## OriginIceland 137.93 53.33 2.59 0.01345 *
## OriginIndia 128.48 32.26 3.98 0.00028 ***
## OriginIndonesia 39.93 53.33 0.75 0.45840
## OriginIreland 154.93 53.33 2.90 0.00596 **
## OriginIsrael 100.93 53.33 1.89 0.06568 .
## OriginItaly 59.11 19.51 3.03 0.00428 **
## OriginJapan 177.93 53.33 3.34 0.00184 **
## OriginJordan 68.93 53.33 1.29 0.20361
## OriginKyrgyzstan -31.07 53.33 -0.58 0.56348
## OriginLatvia 111.52 46.35 2.41 0.02085 *
## OriginLiechtenstein 132.66 23.32 5.69 1.3e-06 ***
## OriginLithuania 134.93 53.33 2.53 0.01545 *
## OriginLuxembourg 113.10 32.49 3.48 0.00122 **
## OriginMacau 157.93 53.33 2.96 0.00513 **
## OriginMacedonia 44.83 23.89 1.88 0.06786 .
## OriginMexico 56.93 53.33 1.07 0.29215
## OriginMorocco 86.97 32.25 2.70 0.01019 *
## OriginNetherlands 157.36 26.30 5.98 5.0e-07 ***
## OriginNew Zealand 75.78 34.17 2.22 0.03233 *
## OriginNorway 179.08 43.70 4.10 0.00020 ***
## OriginPakistan 44.57 29.01 1.54 0.13234
## OriginPoland 119.21 21.42 5.57 1.9e-06 ***
## OriginPortugal 72.55 20.37 3.56 0.00097 ***
## OriginQatar -4.07 53.33 -0.08 0.93958
## OriginRep. of Korea 77.78 34.17 2.28 0.02827 *
## OriginRomania 59.68 30.18 1.98 0.05491 .
## OriginRussia 117.52 28.94 4.06 0.00022 ***
## OriginSamoa -26.38 39.81 -0.66 0.51144
## OriginSerbia 45.25 19.07 2.37 0.02253 *
## OriginSlovakia 127.68 30.18 4.23 0.00013 ***
## OriginSlovenia 50.33 23.89 2.11 0.04144 *
## OriginSouth Africa 105.93 39.16 2.71 0.00998 **
## OriginSpain 114.66 23.32 4.92 1.5e-05 ***
## OriginSweden 134.56 36.22 3.71 0.00062 ***
## OriginSwitzerland 140.16 23.32 6.01 4.6e-07 ***
## OriginTaiwan 178.93 53.33 3.35 0.00175 **
## OriginThailand 67.93 53.33 1.27 0.21011
## OriginThe Congo 75.97 32.25 2.36 0.02346 *
## OriginThe Philippines 76.93 39.16 1.96 0.05642 .
## OriginTunesia 32.93 53.33 0.62 0.54042
## OriginTurkey 39.41 17.77 2.22 0.03234 *
## OriginUkraine 93.52 46.35 2.02 0.05039 .
## OriginUnited Kingdom 112.28 34.17 3.29 0.00212 **
## OriginUnited States 135.93 39.16 3.47 0.00126 **
## OriginUruguay 74.93 53.33 1.40 0.16774
## OriginVietnam 82.93 39.16 2.12 0.04044 *
## HostAustralia 82.00 36.21 2.26 0.02903 *
## HostAustria 26.25 44.63 0.59 0.55963
## HostAzerbaijan NA NA NA NA
## HostBelgium -2.04 46.49 -0.04 0.96520
## HostBrazil -49.85 58.18 -0.86 0.39671
## HostBulgaria NA NA NA NA
## HostCanada NA NA NA NA
## HostChile NA NA NA NA
## HostColombia NA NA NA NA
## HostCroatia 69.60 54.61 1.27 0.20983
## HostCzech Republic -29.75 57.47 -0.52 0.60760
## HostDenmark -5.20 46.37 -0.11 0.91129
## HostEstonia 150.89 64.82 2.33 0.02507 *
## HostFinland 56.89 53.77 1.06 0.29634
## HostFrance 34.58 53.43 0.65 0.52117
## HostGermany 17.94 45.60 0.39 0.69608
## HostGermany_1 8.46 52.69 0.16 0.87324
## HostGreece 77.93 50.22 1.55 0.12854
## HostHong Kong NA NA NA NA
## HostHungary -30.75 57.47 -0.53 0.59562
## HostIceland NA NA NA NA
## HostIndonesia NA NA NA NA
## HostIreland NA NA NA NA
## HostIsrael NA NA NA NA
## HostItaly 62.83 53.27 1.18 0.24517
## HostJapan NA NA NA NA
## HostJordan NA NA NA NA
## HostKyrgyzstan NA NA NA NA
## HostLatvia 25.42 56.91 0.45 0.65758
## HostLiechtenstein 24.63 45.54 0.54 0.59159
## HostLithuania NA NA NA NA
## HostLuxembourg 19.83 46.61 0.43 0.67281
## HostMacau NA NA NA NA
## HostMexico NA NA NA NA
## HostNetherlands 28.19 47.98 0.59 0.56023
## HostNew Zealand 98.31 39.57 2.48 0.01726 *
## HostNorway -45.15 53.99 -0.84 0.40792
## HostPoland 25.72 53.64 0.48 0.63422
## HostPortugal 24.15 45.54 0.53 0.59881
## HostQatar NA NA NA NA
## HostRep. of Korea 91.15 48.56 1.88 0.06781 .
## HostRomania 5.25 57.47 0.09 0.92760
## HostRussia 8.42 56.91 0.15 0.88317
## HostSerbia 25.68 53.02 0.48 0.63074
## HostSeychelles 46.91 42.94 1.09 0.28116
## HostSlovakia 7.25 57.47 0.13 0.90018
## HostSlovenia 115.60 54.61 2.12 0.04053 *
## HostSpain 20.27 55.02 0.37 0.71454
## HostSweden 15.37 60.47 0.25 0.80066
## HostSwitzerland 21.91 45.54 0.48 0.63314
## HostTaiwan NA NA NA NA
## HostThailand NA NA NA NA
## HostTunesia NA NA NA NA
## HostTurkey 31.53 52.27 0.60 0.54979
## HostUnited Kingdom 49.65 48.56 1.02 0.31268
## HostUnited States NA NA NA NA
## HostUruguay NA NA NA NA
## ---
## Signif. codes: 0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1
##
## Residual standard error: 26 on 40 degrees of freedom
## Multiple R-squared: 0.941, Adjusted R-squared: 0.784
## F-statistic: 6.01 on 106 and 40 DF, p-value: 4.22e-09#(fit_pisa_sum = MOD_summary(fit_pisa, kfold = F))
#TODO: enable NAs in output. Useful in a few cases, like this one!
fit_pisa %>% aov %>% lsr::etaSquared() %>% sqrt## eta.sq eta.sq.part
## Origin 0.64 0.94
## Host 0.33 0.80#missing cells
length(unique(pisa$Origin))## [1] 72length(unique(pisa_long$Host))## [1] 58length(unique(pisa$Origin)) * length(unique(pisa_long$Host))## [1] 4176nrow(pisa_long) / (length(unique(pisa$Origin)) * length(unique(pisa_long$Host)))## [1] 0.035Origin country characteristics models
As in all previous studies, we calculate the correlations between country of origin predictors – IQ and Muslim% – and outcomes. In this case, we can do it by sex as well, and have near perfect comparison data for 3 countries.
#correlations overall
wtd.cors(mega[c("Employment_men", "Employment_women", "Employment_total", "IQ", "Muslim")])## Employment_men Employment_women Employment_total IQ Muslim
## Employment_men 1.00 0.91 0.98 0.64 -0.63
## Employment_women 0.91 1.00 0.97 0.76 -0.78
## Employment_total 0.98 0.97 1.00 0.70 -0.71
## IQ 0.64 0.76 0.70 1.00 -0.84
## Muslim -0.63 -0.78 -0.71 -0.84 1.00#by country
plyr::dlply(mega, "Host", function(this) {
#browser()
cor_matrix(this[c("Employment_men", "Employment_women", "Employment_total", "IQ", "Muslim")], CI = .95)
})## $Denmark
## Employment_men Employment_women Employment_total IQ Muslim
## Employment_men "1" "0.92 [0.75 0.98]" "0.98 [0.93 1.00]" "0.71 [0.24 0.91]" "-0.73 [-0.92 -0.28]"
## Employment_women "0.92 [0.75 0.98]" "1" "0.98 [0.92 0.99]" "0.77 [0.35 0.93]" "-0.86 [-0.96 -0.55]"
## Employment_total "0.98 [0.93 1.00]" "0.98 [0.92 0.99]" "1" "0.74 [0.30 0.92]" "-0.80 [-0.94 -0.43]"
## IQ "0.71 [0.24 0.91]" "0.77 [0.35 0.93]" "0.74 [0.30 0.92]" "1" "-0.84 [-0.95 -0.51]"
## Muslim "-0.73 [-0.92 -0.28]" "-0.86 [-0.96 -0.55]" "-0.80 [-0.94 -0.43]" "-0.84 [-0.95 -0.51]" "1"
##
## $Norway
## Employment_men Employment_women Employment_total IQ Muslim
## Employment_men "1" "0.88 [0.61 0.96]" "0.97 [0.90 0.99]" "0.67 [0.15 0.90]" "-0.69 [-0.90 -0.18]"
## Employment_women "0.88 [0.61 0.96]" "1" "0.96 [0.87 0.99]" "0.82 [0.46 0.95]" "-0.85 [-0.96 -0.53]"
## Employment_total "0.97 [0.90 0.99]" "0.96 [0.87 0.99]" "1" "0.75 [0.30 0.92]" "-0.78 [-0.94 -0.37]"
## IQ "0.67 [0.15 0.90]" "0.82 [0.46 0.95]" "0.75 [0.30 0.92]" "1" "-0.84 [-0.95 -0.51]"
## Muslim "-0.69 [-0.90 -0.18]" "-0.85 [-0.96 -0.53]" "-0.78 [-0.94 -0.37]" "-0.84 [-0.95 -0.51]" "1"
##
## $Sweden
## Employment_men Employment_women Employment_total IQ Muslim
## Employment_men "1" "0.93 [0.77 0.98]" "0.98 [0.94 1.00]" "0.59 [0.03 0.87]" "-0.50 [-0.83 0.10]"
## Employment_women "0.93 [0.77 0.98]" "1" "0.98 [0.94 1.00]" "0.72 [0.24 0.91]" "-0.67 [-0.90 -0.15]"
## Employment_total "0.98 [0.94 1.00]" "0.98 [0.94 1.00]" "1" "0.66 [0.14 0.89]" "-0.58 [-0.87 -0.01]"
## IQ "0.59 [0.03 0.87]" "0.72 [0.24 0.91]" "0.66 [0.14 0.89]" "1" "-0.84 [-0.96 -0.52]"
## Muslim "-0.50 [-0.83 0.10]" "-0.67 [-0.90 -0.15]" "-0.58 [-0.87 -0.01]" "-0.84 [-0.96 -0.52]" "1"
##
## attr(,"split_type")
## [1] "data.frame"
## attr(,"split_labels")
## Host
## 1 Denmark
## 2 Norway
## 3 Sweden#plot
ggplot(mega, aes(IQ, Employment_total/100, color = Host, label = iso)) +
geom_point() +
geom_smooth(method = lm, se = F) +
ggrepel::geom_text_repel() +
scale_y_continuous("Employed%", labels = scales::percent) +
xlab("National IQ of home country (Lynn & Vanhanen)") +
theme_bw()
GG_save("figures/IQ_by_country.png")
ggplot(mega, aes(Muslim, Employment_total/100, color = Host, label = iso)) +
geom_point() +
geom_smooth(method = lm, se = F) +
ggrepel::geom_text_repel() +
scale_y_continuous("Employed%", labels = scales::percent) +
scale_x_continuous("Muslim% in home country (Pew Research)", labels = scales::percent) +
theme_bw()
GG_save("figures/Muslim_by_country.png")
#AOV model betas - unemployment
aov_fit_betas = fit_sex_sum %>%
.[[1]] %>%
.["Beta"] %>%
rownames_to_column(var = "Predictor") %>%
filter(str_detect(Predictor, "Origin")) %>%
mutate(iso = str_replace(Predictor, "Origin: ", "") %>% pu_translate, Beta = standardize(Beta)) %>%
left_join(mega_all[c("iso", "Names", "IQ", "Muslim")], by = "iso")
#plot
GG_scatter(aov_fit_betas, "IQ", "Beta", case_names = "Names", repel_names = T) +
xlab("National IQ of home country (Lynn & Vanhanen)")
GG_save("figures/IQ_beta.png")
GG_scatter(aov_fit_betas, "Muslim", "Beta", case_names = "Names", repel_names = T) +
scale_x_continuous("Muslim% in home country (Pew Research)", labels = scales::percent)
GG_save("figures/Muslim_beta.png")
#AOV model betas - pisa
#due to missing data, we have to use the regular lm fit
aov_fit_betas_pisa = fit_pisa %>%
broom::tidy() %>%
filter(str_detect(term, "Origin")) %>%
mutate(iso = str_replace(term, "Origin", "") %>% pu_translate, Beta = standardize(estimate)) %>%
left_join(mega_all[c("iso", "Names", "IQ", "Muslim")], by = "iso")## No exact match: Cap Verde## No exact match: Rep. of Korea## No exact match: The Congo## No exact match: The Philippines## No exact match: Tunesia## Best fuzzy match found: Cap Verde -> Cape Verde with distance 1.00## Best fuzzy match found: Rep. of Korea -> Republic of Korea with distance 5.00## Best fuzzy match found: The Congo -> DR Congo with distance 3.00## Best fuzzy match found: The Philippines -> Philippines with distance 4.00## Best fuzzy match found: Tunesia -> Tunisia with distance 1.00#plot
GG_scatter(aov_fit_betas_pisa, "IQ", "Beta", case_names = "Names", repel_names = T) +
xlab("National IQ of home country (Lynn & Vanhanen)")
GG_save("figures/IQ_beta_pisa.png")
GG_scatter(aov_fit_betas_pisa, "Muslim", "Beta", case_names = "Names", repel_names = T) +
scale_x_continuous("Muslim% in home country (Pew Research)", labels = scales::percent, limits = c(-.15, NA))
GG_save("figures/Muslim_beta_pisa.png")