-
Notifications
You must be signed in to change notification settings - Fork 9
/
gdp.Rmd
266 lines (229 loc) · 9.94 KB
/
gdp.Rmd
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
89
90
91
92
93
94
95
96
97
98
99
100
101
102
103
104
105
106
107
108
109
110
111
112
113
114
115
116
117
118
119
120
121
122
123
124
125
126
127
128
129
130
131
132
133
134
135
136
137
138
139
140
141
142
143
144
145
146
147
148
149
150
151
152
153
154
155
156
157
158
159
160
161
162
163
164
165
166
167
168
169
170
171
172
173
174
175
176
177
178
179
180
181
182
183
184
185
186
187
188
189
190
191
192
193
194
195
196
197
198
199
200
201
202
203
204
205
206
207
208
209
210
211
212
213
214
215
216
217
218
219
220
221
222
223
224
225
226
227
228
229
230
231
232
233
234
235
236
237
238
239
240
241
242
243
244
245
246
247
248
249
250
251
252
253
254
255
256
257
258
259
260
261
262
263
264
265
266
---
title: "Gender and Diversity Project"
author: "Gokhan Ciflikli"
output: html_document
---
```{r warning=FALSE, error=FALSE, message=FALSE, echo=FALSE}
rm(list=ls())
#Loading libraries
library(foreign)
library(Hmisc)
library(dplyr)
library(stringr)
library(ggplot2)
library(RColorBrewer)
if(Sys.info()["user"]=="gokhan"){pathOUT="~/Dropbox/Projects/GDP/data/outputData";
pathIN="~/Dropbox/Projects/GDP/data/inputData";pathR="~/Dropbox/Projects/GDP/Rcode";
pathRep="~/Dropbox/Projects/GDP/Replication";pathM="~/Dropbox/Projects/GDP/manuscript"}
setwd(pathIN)
#Read in data
gender <- read.csv("gender.csv")
diversity <- read.csv("diversity.csv")
convener <- read.csv("convener.csv")
#Filter unique entries (remove duplicates)
#gender <- unique(gender)
#diversity <- unique(diversity)
#Subset the gender data
want.var <- c("Title","Author","Editor","AutGen","EdGen","AutM","AutF","EdM","EdF","Course","Type",
"Importance","Year","Publisher")
want <- which(colnames(gender) %in% want.var)
gender <- gender[,want]
#Merge the data
gender <- merge(gender,convener,by="Course",all.x=TRUE)
want.var <- c("Course","Cluster")
want <- which(colnames(convener) %in% want.var)
convener <- convener[,want]
diversity <- merge(diversity,convener,by="Course")
#Clean up and create new variables
gender$Code <- substr(gender$Course,3,5) #Extract course code
gender$Code <- as.integer(gender$Code)
gender$Level <- cut(gender$Code,
breaks=c(0,400,500,Inf),
labels=c("Undergrad","Masters","PhD")) #Split UG/MA/PhD levels
gender$Importance[gender$Importance==""] <- NA
gender$AutGen[gender$AutGen==""] <- NA
gender$Author[gender$Author==""] <- NA
gender <- gender[!is.na(gender$Author),]
gender <- gender[gender$AutF>0 | gender$AutM>0,]
gender <- gender[gender$Type=="Book" | gender$Type=="Article",] #Subset to books and articles only
gender$Weighted <- round(gender$AutF/(gender$AutF+gender$AutM),3) #Female involvement adjusted for no. of authors
gender$Senior <- ifelse(gender$Rank=="Associate" | gender$Rank=="Professor",1,0)
gender$Senior <- as.factor(gender$Senior)
#Transform variables for logistic regression
gender$Female <- ifelse(gender$AutF>0,1,0)
gender$Female <- as.factor(gender$Female)
```
```{r warning=FALSE, error=FALSE, message=FALSE, echo=FALSE}
#Graphs by date of publication
ggplot(gender[gender$Year>1965 & gender$Year<2017,],aes(x=Year,fill=Female)) +
geom_histogram(binwidth=.5,alpha=.5,position="identity") +
scale_fill_brewer(palette="Set1") +
scale_x_continuous(name="Date of Publication") +
scale_y_continuous(name="Times Included in Reading List")
#Frequency of female author publications by year
hist(gender$Year[gender$Year>1945 & gender$AutF>0])
#Co-authorship statistics
#Female co-authorship preferences
describe(gender$AutF[gender$AutF>0]) #with other females
describe(gender$AutM[gender$AutF>0]) #with males
```
```{r warning=FALSE, error=FALSE, message=FALSE}
#####Hypotheses Testing#####
#1a Male-Female Inclusion
t.test(gender$AutM,gender$AutF)
#1b Essential Readings
essential <- glm(gender$Importance=="Essential"~Female+Type+Year+Level+Convener+Senior,
data=gender,family="binomial")
summary(essential)
#2 Single-Author
gender$single.female <- ifelse(gender$AutF==1 & gender$AutM==0,1,0)
gender$single.male <- ifelse(gender$AutF==0 & gender$AutM==1,1,0)
gender$Single <- ifelse(gender$single.female==1 | gender$single.male==1,1,0)
single <- glm(gender$Single~Female+Type+Year+Level+Convener+Senior,
data=gender,family="binomial")
summary(single)
#3 Book
book <- glm(gender$Type=="Book"~Female+Importance+Year+Level+Convener+Senior,
data=gender,family="binomial")
summary(book)
#4 Top Journals
top.j <- c("International Organization","International security","American Political Science Review",
"International Studies Quarterly","Foreign Policy Analysis",
"European Journal of International Relations",
"Journal of Conflict Resolution","World Politics",
"Review of international political economy")
gender$Top <- ifelse(is.element(gender$Title,top.j),1,0)
gender$Top <- as.factor(gender$Top)
journal <- glm(Top~Female+Importance+Year+Level+Convener+Senior,
data=gender,family="binomial")
summary(journal)
#5 Top Publishers
top.uni <- c("Cambridge University Press","Routledge","Oxford University Press","Cornell University Press",
"Palgrave Macmillan","The MIT Press","Princeton University Press","Columbia University Press")
gender$Top.Press <- ifelse(is.element(gender$Publisher,top.uni),1,0)
gender$Top.Press <- as.factor(gender$Top.Press)
press <- glm(Top.Press~Female+Importance+Year+Level+Convener+Senior,
data=gender,family="binomial")
summary(press)
#6 First Author in Co-Authored Works
#Two Authors
gender$two <- ifelse(gender$AutM==1 & gender$AutF==1,1,0)
describe(gender$AutGen[gender$two==1])
gender2 <- gender[gender$two==1,]
gender2$fem <- ifelse(gender2$AutGen=="FM",1,0)
t.test(gender2$fem==1,gender2$fem==0)
#Three Authors
gender$three <- ifelse(gender$AutM>0 & gender$AutF>0 & gender$AutM+gender$AutF==3,1,0)
describe(gender$AutGen[gender$three==1])
gender3 <- gender[gender$three==1,]
gender3$fem <- ifelse(gender3$AutGen=="FFM" |
gender3$AutGen=="FMF" |
gender3$AutGen=="FMM",1,0)
t.test(gender3$fem==1,gender3$fem==0)
rm(gender2)
rm(gender3)
#8 Co-author gender
gender$female.comale <- ifelse(gender$AutF>0 & gender$AutM>0,1,0)
gender$female.cofemale <- ifelse(gender$AutF>1 & gender$AutM==0,1,0)
t.test(gender$female.comale,gender$female.cofemale)
#9 Co-authoring with men vs. single-author
t.test(gender$female.comale,gender$Single==1 & gender$AutM==0)
#10 Gender studies & #17a Time x Female
sex <- glm(Sexuality~Type+Importance+Year*Female+Level+Convener,
data=diversity,family="binomial",subset=diversity$Year>1945)
summary(sex)
#12 Male Convener & #13 Junior Faculty & #15 Course Level & #16 Non-Core Courses
core <- c("100","200","202","203","410","436","450","501","509")
gender$Core <- ifelse(is.element(gender$Code,core),1,0)
gender$Core <- as.factor(gender$Core)
#gender$Name <- as.character(gender$Name)
gender$Self <- mapply(function(x,y) all(x %in% y),
str_extract_all(gender$Author,"\\w+"),str_extract_all(gender$Name,"\\w+"))
logit.g <- glm(Female~Type+Importance+Year+Level+Convener+Senior+Cluster+Core+Top+Top.Press+Self,
data=gender,family="binomial")
summary(logit.g)
```
```{r}
options(width=80)
#Convert odds to percentages
odds.g <- round(exp(logit.g$coefficients),2)-1
odds.g #these mean "times more likely", i.e. 0 equals no effect, 1 means 100% more likely, -1 100% less
```
```{r warning=FALSE, error=FALSE, message=FALSE}
#17 Temporal Patterns
ols.time <- lm(Year~Female+Type+Importance+Year+Level+Convener+Senior+Single+Top+Top.Press+Self,
data=gender)
summary(ols.time)
```
```{r warning=FALSE, error=FALSE, message=FALSE, echo=FALSE}
#5 & #14 Publisher Tables and Yearly Inclusion
library(plyr)
gender$Female <- as.integer(gender$Female)-1
pub <- ddply(gender, .(Publisher), summarize, Total=length(Publisher), Female=sum(Female))
pub$Ratio <- round((pub$Female)/(pub$Total),2)
pub <- pub[pub$Total>9 & pub$Total<1000,] #Require a minimum of ten items
options(DT.options=list(pageLength=10,language=list(search='Filter:')))
DT::datatable(pub,caption="Table 1: Publisher Gender Breakdown (Minimum 10 Works)",options=list(
order=list(list(4,'desc'))
))
detach(package:plyr)
gender$male.comale <- ifelse(gender$AutM>1 & gender$AutF==0,1,0)
gender$male.cofemale <- ifelse(gender$AutM>0 & gender$AutF>0,1,0)
gender$Male <- ifelse(gender$Female==0,1,0)
gender$book <- ifelse(gender$Type=="Book",1,0)
gender$article <- ifelse(gender$book==1,0,1)
yearly <- gender %>% group_by(Year) %>%
summarise(Readings=length(Publisher),#total.female=sum(Female),total.male=sum(Male),
Female=sum(AutF),Male=sum(AutM),
SA=sum(Single),SA.F=sum(single.female),SA.M=sum(single.male),
FM=sum(female.comale),FF=sum(female.cofemale),
MM=sum(male.comale),
Weighted=sum(Weighted))
yearly$FM.Ratio <- round(yearly$Female/(yearly$Female+yearly$Male),3)
yearly$FM.Weighted <- round(yearly$Weighted/yearly$Readings,3)
yearly$Weighted <- NULL
yearly$Year[yearly$Year==""] <- NA
library(dygraphs)
authors <- cbind(yearly$Year,yearly$FM.Weighted)
authors <- authors[1:99,]
authors <- as.data.frame(authors)
authors$V1 <- paste(authors$V1,"-01-01",sep="")
authors$V1 <- as.Date(authors$V1)
authors$V3 <- 1-authors$V2
#authors <- authors[authors$V3<1,]
authors <- as.matrix(authors)
rownames(authors) = authors[,1]
authors <- authors[,2:3]
authors <- authors[43:99,]
dygraph(authors,main = "Reading List Inclusion Rates over Time") %>%
dyOptions(fillGraph = TRUE, fillAlpha = 0.1) %>%
dyLimit(.2, color = "red") %>%
dyLegend(width = 400) %>%
dyAxis("y", label = "Percentage of All Readings",valueRange = c(0,1.001)) %>%
dyAxis("x", label = "Date of Publication") %>%
dySeries("V2", label = "Female Inclusion") %>%
dySeries("V3", label = "Male Inclusion")
```
```{r echo=FALSE}
options(DT.options=list(pageLength=10,language=list(search='Filter:')))
DT::datatable(yearly,caption="Table 2: Yearly Breakdowns,
Gender/Single-Author/Co-Authored/Binary and Percentage Gender Ratio",
rownames=FALSE,options = list(
order = list(list(0, 'desc'))
))
course <- gender %>% group_by(Code) %>%
summarise(Readings=length(Publisher),
Female=sum(AutF),Male=sum(AutM),
Book=sum(book),Article=sum(article),
Weighted=sum(Weighted))
course$BA.Ratio <- round(course$Book/course$Readings,2)
course$FM.Ratio <- round(course$Female/(course$Female+course$Male),3)
course$FM.Weighted <- round(course$Weighted/course$Readings,3)
course$Weighted <- NULL
course$Code <- paste("IR",course$Code,sep="")
```
```{r echo=FALSE}
DT::datatable(course,caption="Table 3: Course Breakdown by Book/Article and Gender",rownames=FALSE,options=list(
order=list(list(7,'desc'))
))
```