load data and libraries
library(tidyverse)
## ── Attaching packages ─────────────────────────────────────── tidyverse 1.3.2 ──
## ✔ ggplot2 3.4.0 ✔ purrr 0.3.5
## ✔ tibble 3.1.8 ✔ dplyr 1.0.10
## ✔ tidyr 1.2.1 ✔ stringr 1.4.1
## ✔ readr 2.1.3 ✔ forcats 0.5.2
## ── Conflicts ────────────────────────────────────────── tidyverse_conflicts() ──
## ✖ dplyr::filter() masks stats::filter()
## ✖ dplyr::lag() masks stats::lag()
ppm <- read.csv("publishingpaidme_data.csv", header = TRUE)
view data
names(ppm)
## [1] "timestamp" "primary_genre" "subgenre_1" "subgenre_2"
## [5] "subgenre_3" "payment_usd" "rights" "sale_or_pub_year"
## [9] "publisher" "debut" "agented" "race"
## [13] "race_2" "race_3" "race_4" "gender"
## [17] "sexuality" "disability" "author" "title"
## [21] "comp_titles_yn" "comments"
#make categorical variables factors
ppm$debut <- as.factor(ppm$debut)
ppm$primary_genre <- as.factor(ppm$primary_genre)
ppm$rights <- as.factor(ppm$rights)
ppm$primary_genre <- as.factor(ppm$primary_genre)
ppm$subgenre_1 <- as.factor(ppm$subgenre_1)
ppm$subgenre_2 <- as.factor(ppm$subgenre_2)
ppm$subgenre_3 <- as.factor(ppm$subgenre_3)
ppm$publisher <- as.factor(ppm$publisher)
ppm$agented <- as.factor(ppm$agented)
ppm$race <- as.factor(ppm$race)
ppm$race_2 <- as.factor(ppm$race_2)
ppm$race_3 <- as.factor(ppm$race_3)
ppm$race_4 <- as.factor(ppm$race_4)
ppm$gender <- as.factor(ppm$gender)
ppm$sexuality <- as.factor(ppm$sexuality)
ppm$disability <- as.factor(ppm$disability)
ppm$comp_titles_yn <- as.factor(ppm$comp_titles_yn)
ppm$sale_or_pub_year <- as.numeric(ppm$sale_or_pub_year)
#hist(count(ppm$debut))
View the summary
summary(ppm)
## timestamp primary_genre
## Length:2856 YA :931
## Class :character Science Fiction and Fantasy:311
## Mode :character Literary Fiction :309
## MG :301
## Non-Fiction :298
## PB :219
## (Other) :487
## subgenre_1 subgenre_2
## :1865 :2774
## Science Fiction and Fantasy: 385 Thriller/Mystery : 10
## Memoir : 81 Science Fiction and Fantasy: 8
## Contemporary : 72 Historical : 7
## Romance : 59 LGBT : 7
## Historical : 54 Short Story : 7
## (Other) : 340 (Other) : 43
## subgenre_3 payment_usd rights
## :2851 Min. : -100 Worldwide :932
## BIPOC : 1 1st Qu.: 6000 :755
## LGBT : 1 Median : 18000 English Language Only:598
## Literary Fiction: 1 Mean : 49820 North America Only :514
## Non-Fiction : 1 3rd Qu.: 45000 Other Language Only : 4
## Thriller/Mystery: 1 Max. :3000000 UK and Commonwealth : 3
## (Other) : 50
## sale_or_pub_year publisher debut
## Min. :1990 Small/Indie :505 Yes :1355
## 1st Qu.:2014 :378 No : 941
## Median :2017 Penguin Random House (and Imprints):303 : 506
## Mean :2016 Large (Big 5) :269 YA Debut: 10
## 3rd Qu.:2019 Mid-Size (Scholastic, Disney, etc) :246 PB Debut: 8
## Max. :2024 HarperCollins (and Imprints) :243 MG Debut: 7
## NA's :462 (Other) :912 (Other) : 29
## agented race
## : 780 White :1952
## For 12 : 1 Black : 152
## No : 397 East Asian : 131
## Short Story collection contest: 1 Latin American: 89
## Yes :1677 Multiracial : 76
## (Other) : 175
## NA's : 281
## race_2 race_3 race_4
## :2775 :2819 :2851
## White : 49 Multiracial: 32 Multiracial: 5
## Multiracial : 12 White : 5
## Nonwhite : 5
## Latin American : 4
## Native American/First Nations: 3
## (Other) : 8
## gender sexuality
## Cis Woman :1810 Straight :1313
## Cis Man : 475 : 801
## None Reported : 448 Bi/Pansexual: 317
## Nonbinary/Agender: 77 Queer : 173
## Other : 27 Gay : 68
## Trans Man : 11 Lesbian : 58
## (Other) : 8 (Other) : 126
## disability author
## : 967 Length:2856
## Chronic illness : 1 Class :character
## No :1657 Mode :character
## not disabled; I do have a neurodivergence: 1
## PTSD : 2
## Yes : 228
##
## title comp_titles_yn comments
## Length:2856 :2203 Length:2856
## Class :character No : 323 Class :character
## Mode :character Yes: 330 Mode :character
##
##
##
##
summary(ppm$payment_usd)
## Min. 1st Qu. Median Mean 3rd Qu. Max.
## -100 6000 18000 49820 45000 3000000
summary(ppm$race)
## Arab/West Asian/Middle Eastern Black
## 28 152
## East Asian Indian/South Asian
## 131 66
## Latin American Multiracial
## 89 76
## Native American/First Nations Nonwhite
## 14 13
## Other Pacific Islander
## 3 6
## Southeast Asian White
## 45 1952
## NA's
## 281
summary(ppm$race_2)
## Arab/West Asian/Middle Eastern
## 2775 2
## East Asian Latin American
## 2 4
## Multiracial Native American/First Nations
## 12 3
## Nonwhite Pacific Islander
## 5 2
## South Asian Southeast Asian
## 1 1
## White
## 49
summary(ppm$gender)
## Cis Man Cis Woman Genderfluid/Multi-Gender
## 475 1810 4
## Nonbinary/Agender None Reported Other
## 77 448 27
## Trans Man Trans Woman
## 11 4
We need to clean up some variables: comp titles No/no/No. and Yes/yes need to be merged replace blanks with NA update agented to remove the two non-answer values change specific debut to yes/no *change disability answers to y/n
hist(ppm$payment_usd)
ggplot(ppm, aes(x = payment_usd)) +
geom_histogram(bins = 45)
#mathematical summary
advances <- ppm %>% summarise(
minimum = min(payment_usd),
average = mean(payment_usd),
median_payment = median(payment_usd),
stdev = sd(payment_usd)
)
View(advances)
#summary
summary(ppm$payment_usd)
## Min. 1st Qu. Median Mean 3rd Qu. Max.
## -100 6000 18000 49820 45000 3000000
# Min. 1st Qu. Median Mean 3rd Qu. Max.
# -100 6000 18000 49820 45000 3000000
#who had below $0 advance?
min_ppm <- ppm[ppm$payment_usd < 0, ]
min_ppm
## timestamp primary_genre subgenre_1 subgenre_2 subgenre_3
## 281 6/16/2020 7:55 Science Fiction and Fantasy
## payment_usd rights sale_or_pub_year publisher debut agented race
## 281 -100 Worldwide 2015 Small/Indie No No White
## race_2 race_3 race_4 gender sexuality disability author title
## 281 Cis Man Straight No
## comp_titles_yn
## 281
## comments
## 281 The negative advance is due to contributing authors paying up front toward production and marketing costs, but the royalty rate was high. So far the authors have each earned $13k in royalties.
max_ppm <- ppm[ppm$payment_usd > 1000000, ]
max_ppm
## timestamp primary_genre subgenre_1 subgenre_2 subgenre_3
## 407 6/8/2020 11:38 Thriller/Mystery
## 626 6/10/2020 12:10 Literary Fiction
## 852 12/12/2022 15:33 Commercial
## 2557 6/7/2020 22:46 Non-Fiction
## 2568 6/7/2020 22:45 Non-Fiction
## 2575 6/7/2020 9:31 Literary Fiction
## 2579 6/7/2020 22:46 Non-Fiction
## payment_usd rights sale_or_pub_year
## 407 1100000 Worldwide 2020
## 626 2500000 North America Only 2016
## 852 1150000 North America Only 2022
## 2557 3000000 2018
## 2568 3000000 2019
## 2575 1300000 North America Only 2020
## 2579 2000000 2020
## publisher debut agented race race_2 race_3
## 407 Large (Big 5) No Yes White
## 626 No Yes East Asian
## 852 Large (Big 5) Yes Yes Multiracial
## 2557 Flatiron <NA>
## 2568 Little/Brown (Hachette) <NA>
## 2575 Macmillan (and Imprints) <NA>
## 2579 Simon & Schuster (and Imprints) <NA>
## race_4 gender sexuality disability author
## 407 Cis Man Straight No
## 626 Cis Woman Straight No
## 852 Cis Woman Straight No
## 2557 None Reported James Comey
## 2568 None Reported Ronan Farrow
## 2575 None Reported Jeanine Cummins
## 2579 None Reported John Bolton
## title comp_titles_yn
## 407
## 626
## 852
## 2557 A Higher Loyalty
## 2568 Catch and Kill
## 2575 American Dirt
## 2579 The Room Where It Happened
## comments
## 407 I'm a well-known, best-selling author
## 626 award-winning author, already published bestseller
## 852 Auction
## 2557
## 2568
## 2575
## 2579
We’re going to remove the two outliers ($3M advance) from this analysis because they are James Comey and Ronan Farrow and that’s all that needs to be said about that.
ppm_corrected <- ppm[ppm$payment_usd < 3000000, ]
summary(ppm_corrected$payment_usd)
## Min. 1st Qu. Median Mean 3rd Qu. Max.
## -100 6000 18000 47753 45000 2500000
remove the additional outliers above $1M
ppm_corrected <- ppm[ppm$payment_usd < 1000000, ]
summary(ppm_corrected$payment_usd)
## Min. 1st Qu. Median Mean 3rd Qu. Max.
## -100 5500 18000 42659 45000 900000
#removing the outliers above $1M did not change the median payment, but does slightly lower the mean Min. 1st Qu. Median Mean 3rd Qu. Max. -100 5500 18000 42659 45000 900000
look at the relationships between gender, race, sexuality, debut y/n, and advance amount
racepay <- lm(payment_usd ~ race, data = ppm_corrected)
racepay
##
## Call:
## lm(formula = payment_usd ~ race, data = ppm_corrected)
##
## Coefficients:
## (Intercept) raceBlack
## 44866.51 16406.02
## raceEast Asian raceIndian/South Asian
## 11130.26 -7131.29
## raceLatin American raceMultiracial
## -4107.97 18564.66
## raceNative American/First Nations raceNonwhite
## 33562.06 95.02
## raceOther racePacific Islander
## -35533.18 -19283.01
## raceSoutheast Asian raceWhite
## 7638.73 -3803.41
race_pay <- ggplot(ppm_corrected, aes(x = race, y = payment_usd))+
geom_boxplot()+
geom_jitter(width = 0.1) +
theme_bw() +
theme(axis.text.x = element_text(angle=90, vjust = 0.3)) +
xlab("Race") +
ylab("Payment (USD)") +
ggtitle("Advance payments by self-reported racial groups \n (excluding outliers above $1M USD)")
plot(race_pay)
#add in the outliers
race_pay_2 <- ggplot(ppm, aes(x = race, y = payment_usd))+
geom_boxplot()+
geom_jitter(width = 0.1) +
theme_bw() +
theme(axis.text.x = element_text(angle=90, vjust = 0.3)) +
xlab("Race") +
ylab("Payment (USD)") +
ggtitle("Advance payments by self-reported ethnic groups \n (including outliers above $1M USD)") +
geom_text(x = 9, y = 2e+06, label = "Top outliers above $1M in the NA category: \n John Bolton \n James Comey \n Jeanine Cummins \n Ronan Farrow")
plot(race_pay_2)
look at relationship between gender and advance
genderpay <- lm(payment_usd ~ gender, data = ppm_corrected)
gender_pay <- ggplot(ppm_corrected, aes(x = gender, y = payment_usd))+
geom_boxplot()
plot(gender_pay)
#add in the outliers
gender_pay_2 <- ggplot(ppm, aes(x = gender, y = payment_usd))+
geom_boxplot() +
geom_jitter(width = 0.1) +
theme_bw() +
theme(axis.text.x = element_text(angle=90, vjust = 0.3)) +
xlab("Gender") +
ylab("Payment (USD") +
ggtitle("Advance payments reported in the Publishing Paid Me dataset \n grouped by self-reported Gender (USD)") +
geom_text(x = 7, y = 2e+06, label = "Top outliers above $1M \n in the None Reported category: \n John Bolton \n James Comey \n Ronan Farrow \n Several unidentified authors")
plot(gender_pay_2)
year and advance
lm(payment_usd ~ sale_or_pub_year, data = ppm_corrected)
##
## Call:
## lm(formula = payment_usd ~ sale_or_pub_year, data = ppm_corrected)
##
## Coefficients:
## (Intercept) sale_or_pub_year
## 64576.37 -11.28
year_pay <- ggplot(ppm_corrected, aes(x = sale_or_pub_year, y = payment_usd))+
geom_smooth() +
geom_point()
plot(year_pay)
## `geom_smooth()` using method = 'gam' and formula = 'y ~ s(x, bs = "cs")'
## Warning: Removed 460 rows containing non-finite values (`stat_smooth()`).
## Warning: Removed 460 rows containing missing values (`geom_point()`).
#add in the outliers
year_pay_2 <- ggplot(ppm, aes(x = sale_or_pub_year, y = payment_usd))+
geom_smooth() +
geom_point() +
geom_jitter(width = 0.1)
plot(year_pay_2)
## `geom_smooth()` using method = 'gam' and formula = 'y ~ s(x, bs = "cs")'
## Warning: Removed 462 rows containing non-finite values (`stat_smooth()`).
## Warning: Removed 462 rows containing missing values (`geom_point()`).
## Removed 462 rows containing missing values (`geom_point()`).
Plot a histogram of the overall pattern in payments
hist_payments <- ggplot(ppm, aes(x = payment_usd)) +
geom_histogram(bins = 75)+
theme_bw() +
theme(axis.text.x = element_text(angle=90, vjust = 0.3)) +
xlab("Payment (USD)") +
ylab("Count") +
ggtitle("Advance payments reported in the Publishing Paid Me dataset (USD)") +
geom_text(x = 2e+06, y = 1250, label = "Minimum: $-100 \n Mean: $49820 \n Median: $18000 \n Maximum: $3M")
hist_payments
Do proportions of those reporting for gender
#make a table with the count data for gender
gender_counts <- ppm[, c(16, 6)] %>%
group_by(gender) %>%
summarise(
gender = unique(gender),
count = n(),
gender.index = paste(gender, count, sep = "\n")
)
summary(gender_counts)
## gender count gender.index
## Cis Man :1 Min. : 4.00 Length:8
## Cis Woman :1 1st Qu.: 9.25 Class :character
## Genderfluid/Multi-Gender:1 Median : 52.00 Mode :character
## Nonbinary/Agender :1 Mean : 357.00
## None Reported :1 3rd Qu.: 454.75
## Other :1 Max. :1810.00
## (Other) :2
#import the treemap package
#install.packages("treemap")
library(treemap)
## Warning: package 'treemap' was built under R version 4.2.3
treemap(dtf = gender_counts, vSize = "count", index = "gender.index", fontsize.labels = 10, force.print.labels = TRUE, vColor = "count", type = "value")
Do proportions of those reporting for race
#make a table with the count data for race/ethnicity
race_counts <- ppm[, c(12, 6)] %>%
group_by(race) %>%
summarise(
race = unique(race),
count = n(),
race.index = paste(race, count, sep = "\n")
)
summary(race_counts)
## race count race.index
## Arab/West Asian/Middle Eastern:1 Min. : 3.0 Length:13
## Black :1 1st Qu.: 14.0 Class :character
## East Asian :1 Median : 66.0 Mode :character
## Indian/South Asian :1 Mean : 219.7
## Latin American :1 3rd Qu.: 131.0
## (Other) :7 Max. :1952.0
## NA's :1
treemap(dtf = race_counts, vSize = "count", index = "race.index", fontsize.labels = 10, force.print.labels = TRUE, vColor = "count", type = "value")
pub_count <- ppm[, c(9, 6)] %>%
group_by(publisher) %>%
summarise(
pub = unique(publisher),
counts = n(),
pub.index = paste(pub, counts, sep = "\n")
)
summary(pub_count)
## publisher pub counts pub.index
## : 1 : 1 Min. : 1.0 Length:218
## &NOW books : 1 &NOW books : 1 1st Qu.: 1.0 Class :character
## 10 Speed : 1 10 Speed : 1 Median : 1.0 Mode :character
## Abrams : 1 Abrams : 1 Mean : 13.1
## Abrams : 1 Abrams : 1 3rd Qu.: 3.0
## Abrams Kids: 1 Abrams Kids: 1 Max. :505.0
## (Other) :212 (Other) :212
bar_publishers <- ggplot(pub_count, aes(x = pub)) +
geom_bar()+
theme_bw() +
theme(axis.text.x = element_text(angle=90, vjust = 0.3)) +
xlab("") +
ylab("") +
ggtitle("Advance payments reported in the Publishing Paid Me dataset (USD)")
bar_publishers
agent_counts <- ppm[, c(11, 6)] %>%
group_by(agented) %>%
summarise(
agent = unique(agented),
count = n(),
agent.index = paste(agent, count, sep = "\n")
)
summary(agent_counts)
## agented agent
## :1 :1
## For 12 :1 For 12 :1
## No :1 No :1
## Short Story collection contest:1 Short Story collection contest:1
## Yes :1 Yes :1
##
## count agent.index
## Min. : 1.0 Length:5
## 1st Qu.: 1.0 Class :character
## Median : 397.0 Mode :character
## Mean : 571.2
## 3rd Qu.: 780.0
## Max. :1677.0
agented <- lm(payment_usd ~ agented, data = ppm)
summary(agented)
##
## Call:
## lm(formula = payment_usd ~ agented, data = ppm)
##
## Residuals:
## Min 1Q Median 3Q Max
## -58504 -46004 -28504 -3444 2946238
##
## Coefficients:
## Estimate Std. Error t value Pr(>|t|)
## (Intercept) 53762 4951 10.859 < 2e-16 ***
## agentedFor 12 -53762 138363 -0.389 0.698
## agentedNo -48118 8525 -5.644 1.82e-08 ***
## agentedShort Story collection contest -52762 138363 -0.381 0.703
## agentedYes 4742 5993 0.791 0.429
## ---
## Signif. codes: 0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1
##
## Residual standard error: 138300 on 2851 degrees of freedom
## Multiple R-squared: 0.01657, Adjusted R-squared: 0.01519
## F-statistic: 12.01 on 4 and 2851 DF, p-value: 1.122e-09