library(psych) #for describe()
library(corrplot) #for corrplot()
library(nFactors) #for factor analysis
library(naniar) #for gg_miss_upset()
library(mclust) #for LPA
library(kableExtra) #for tables
library(tidyr) #for graphs
library(ggplot2) #for graphs
library(Rmisc) #for summarySE()
library(reshape2) #for data transformation
library(fmsb) #for spider plot
library(dplyr) #for data transformation
library(stringr)
library(sjPlot)import <- read.csv(file="finaldata.csv")
# limited to finished and doctoral students
import2 <- subset(import, import$Q71 == 1 & Finished == 1)
d <- subset(import2, select=c(X, grep("Q3.1", colnames(import)),
grep("Q4.1", colnames(import)),
grep("Q5.1", colnames(import))))
rm(import, import2)mcar_test(d[-1])# A tibble: 1 x 4
statistic df p.value missing.patterns
<dbl> <dbl> <dbl> <int>
1 544. 514 0.178 14alpha_results <- list(
psych::alpha(subset(d, select = c(grep("Q3",colnames(d))))),
psych::alpha(subset(d, select = c(grep("Q5",colnames(d))))),
psych::alpha(subset(d, select = c(grep("Q4",colnames(d))))))
sapply(alpha_results, function(x) round(x$total$raw_alpha, 3))[1] 0.913 0.935 0.943d$s <- (d$Q3.1_2 + d$Q3.1_3 + d$Q3.1_4 + d$Q3.1_5 + d$Q3.1_6 + d$Q3.1_7 +
d$Q3.1_11 + d$Q3.1_12 + d$Q3.1_13 + d$Q3.1_14 + d$Q3.1_15 +
d$Q3.1_8 + d$Q3.1_9 + d$Q3.1_10)/14
d$r <- (d$Q5.1_2 + d$Q5.1_3 + d$Q5.1_4 + d$Q5.1_5 + d$Q5.1_6 + d$Q5.1_7 +
d$Q5.1_16 + d$Q5.1_12 + d$Q5.1_13 + d$Q5.1_14 + d$Q5.1_15 +
d$Q5.1_8 + d$Q5.1_9 + d$Q5.1_10 + d$Q5.1_11)/15
d$e <- (d$Q4.1_2 + d$Q4.1_3 + d$Q4.1_4 + d$Q4.1_6 + d$Q4.1_7 +
d$Q4.1_11 + d$Q4.1_12 + d$Q4.1_13 + d$Q4.1_14 + d$Q4.1_10 +
d$Q4.1_8 + d$Q4.1_9 + d$Q4.1_5)/17d1 <- na.omit(subset(d, select=c(1, s, r, e)))
m_dist <- mahalanobis(d1[-c(1)], colMeans(d1[-c(1)]), cov(d1[-c(1)]))
d1$MD <- round(m_dist, 1)
plot(d1$MD)
describe(m_dist) vars n mean sd median trimmed mad min max range skew kurtosis se
X1 1 1160 3 3.47 2.09 2.32 1.59 0.02 33.43 33.41 3.67 19.63 0.1cut <- qchisq(.999, df=(ncol(d1)-1))
abline(a=cut, b=0, col="red")
d1$outlier <- F
d1$outlier[d1$MD > cut] <- T
table(d1$outlier)
FALSE TRUE
1149 11 d1 <- subset(d1, outlier == F)
d2 <- subset(d, X %in% d1$X)d2$s_std <- as.vector(scale(d2$s, center=T, scale=T))
d2$r_std <- as.vector(scale(d2$r, center=T, scale=T))
d2$e_std <- as.vector(scale(d2$e, center=T, scale=T))load(file="LPAv6.RData")
d_std <- subset(d2, select=c(X, 68:70))
d_raw <- subset(d2, select=c(X, 59:61))#entropy plot - elbow indicates number of classes that should be considered
d <- subset(d_std, select=-c(1))
# output <- clustCombi(data = d)
output2 <- clustCombiOptim(output, plot=TRUE)
summary(output)----------------------------------------------------
Combining Gaussian mixture components for clustering
----------------------------------------------------
Mclust model name: VII
Number of components: 8
Combining steps:
Step | Classes combined at this step | Class labels after this step
-------|-------------------------------|-----------------------------
0 | --- | 1 2 3 4 5 6 7 8
1 | 1 & 6 | 1 2 3 4 5 7 8
2 | 1 & 5 | 1 2 3 4 7 8
3 | 1 & 3 | 1 2 4 7 8
4 | 1 & 2 | 1 4 7 8
5 | 1 & 7 | 1 4 8
6 | 1 & 4 | 1 8
7 | 1 & 8 | 1 entPlot(output$MclustOutput$z, output$combiM, abc = c("normalized"))
entPlot(output$MclustOutput$z, output$combiM, abc = c("standard"))
#calculate and plot LRT scores
#BLRT - bootstrap likelihood ratio test - compares same model with diff number of clases, sig indicates that model with more classes fits better than one with fewer - should be sig
# lrt <- mclustBootstrapLRT(d, mclust.options("emModelNames"))
print(lrt)-------------------------------------------------------------
Bootstrap sequential LRT for the number of mixture components
-------------------------------------------------------------
Model = EII
Replications = 999
LRTS bootstrap p-value
1 vs 2 4.504357e+02 0.001
2 vs 3 7.954769e+01 0.001
3 vs 4 1.784418e+02 0.001
4 vs 5 7.649047e+01 0.001
5 vs 6 5.237173e+01 0.001
6 vs 7 5.490722e+01 0.001
7 vs 8 4.971794e+01 0.001
8 vs 9 1.254765e-03 0.803#calculate and plot BIC scores
#sample adjusted BIC - in the literature, lowest BIC equals best fitting model (specific to Mplus), in R higher BIC is better
# BIC <- mclustBIC(d, G = 1:9, mclust.options("emModelNames"))
plot.mclustBIC(BIC)
#calculate ICL scores
#BIC penalized by estimated mean entropy, not great for datasets with overlapping, higher BIC is better
# icl <- mclustICL(d, G = 1:9, mclust.options("emModelNames"))
plot.mclustICL(icl)
# # calculate stuff for table
# n <- 1
# p <- 1
# q <- 1
# for (i in 1:9) {
# #run model
# varname <- paste("m",n,sep="")
# m <- Mclust(d, G = n, mclust.options("emModelNames"))
# assign(varname,m)
# #extract ll
# varname3 <- paste("logl",n,sep="")
# ll <- eval(parse(text=paste("m",n,"$loglik",sep = "")))
# assign(varname3,ll)
# #calc free parameters
# #dimension of the data = number of original variables in model
# #EII 1
# #VII G
# #EEI p
# #free covariance parameters, VEI p + G - 1
# #EVI pG - G + 1
# #VVI pG
# #EEE p(p + 1)/2
# #free covariance parameters, EEV Gp(p + 1)/2 - (G - 1)p
# #free covariance parameters, VEV Gp(p + 1)/2 - (G - 1)(p - 1)
# #VVV Gp(p + 1)/2
# #EVE p(p + 1)/2 + (G - 1)(p - 1)
# #VVE p(p + 1)/2 + (G - 1)p
# #VEE p(p + 1)/2 + (G - 1)
# #EVV Gp(p + 1)/2 - (G - 1)
# if (m$modelName == "EII") {
# freepar <- 1
# } else if (m$modelName == "VII") {
# freepar <- m$G
# } else if (m$modelName == "EEI") {
# freepar <- m$d
# } else if (m$modelName == "VEI") {
# freepar <- m$d + m$G - 1
# } else if (m$modelName == "EVI") {
# freepar <- m$d*m$G - m$G + 1
# } else if (m$modelName == "VVI") {
# freepar <- m$d*m$G
# } else if (m$modelName == "EEE") {
# freepar <- m$d(m$d + 1)/2
# } else if (m$modelName == "EEV") {
# freepar <- m$G*m$d*(m$d + 1)/2 - (m$G - 1)*m$d
# } else if (m$modelName == "VEV") {
# freepar <- m$G*m2$d*(m$d + 1)/2 - (m$G - 1)*(m$d - 1)
# } else if (m$modelName == "VVV") {
# freepar <- m$G*m$d*(m$d + 1)/2
# } else if (m$modelName == "XXX") {
# freepar <- "-"
# } else if (m$modelName == "EVE") {
# freepar <- m$d*(m$d + 1)/2 + (m$G - 1)*(m$d - 1)
# } else if (m$modelName == "VVE") {
# freepar <- m$d*(m$d + 1)/2 + (m$G - 1)*m$d
# } else if (m$modelName == "VEE") {
# freepar <- m$d*(m$d + 1)/2 + (m$G - 1)
# } else if (m$modelName == "EVV") {
# freepar <- m$G*m$d*(m$d + 1)/2 - (m$G - 1)
# } else {
# freepar <- 999
# }
# varname4 <- paste("freep",n,sep="")
# assign(varname4,freepar)
# #cal post prob
# m <- eval(parse(text=paste("m",n,sep="")))
# uncer <- m$uncertainty
# posterior_prob <- 1 - round(uncer, 5)
# #cal entropy
# ent <- round(mean(posterior_prob), 3)
# varname2 <- paste("ent",n,sep="")
# assign(varname2, ent)
# #AIC <- (2 * mcm$df - 2 * mcm$loglik)
# varname5 <- paste("aic",n,sep="")
# aic <- (2 * m$df - 2 * m$loglik)
# assign(varname5,aic)
# #CAIC <- (((log(mcm$n) + 1) * mcm$df) - 2 * mcm$loglik)
# varname6 <- paste("caic",n,sep="")
# caic <- (((log(m$n) + 1) * m$df) - 2 * m$loglik)
# assign(varname6,caic)
# #SABIC <- (mcm$df * log((mcm$n + 2)/24) - 2 * mcm$loglik)
# varname7 <- paste("sabic",n,sep="")
# sabic <- (m$df * log((m$n + 2)/24) - 2 * m$loglik)
# assign(varname7, sabic)
# #iterate
# n <- n+1
# p <- p+1
# #grab group sizes
# grps <- table(eval(parse(text=paste("m",q,"$classification",sep=""))))
# tg <- as.list(sort(unique(eval(parse(text=paste("m",q,"$classification",sep=""))))))
# t <- 1
# #calculate group percents
# for (i in tg) {
# perc <- grps[[t]] / (eval(parse(text=paste("m",i,"$n[1]",sep=""))))
# varname <- paste("m",q,"perc",t,sep="")
# assign(varname,perc)
# t <- t + 1
# }
# #combine all group percents into one vector
# mp <- mapply(get, grep(paste("m",q,"perc",sep=""), ls(), value=T))
# varname <- paste("allperc",q,sep="")
# assign(varname,mp)
# q <- q + 1
# }
#throw stuff together for the table
proflist <- sprintf("Profiles %d", 1:9)
modlist <- rbind(m1$modelName, m2$modelName, m3$modelName, m2$modelName, m2$modelName, m2$modelName, m2$modelName, m2$modelName, m2$modelName)
#fix BLRT columns
blank <- "-"
lrtobs <- "-"
lrtobs <- append(lrtobs, round(lrt$obs[1:7], digits = 2))
lrtobs <- append(lrtobs, blank)
lrtp <- "-"
lrtp <- append(lrtp, lrt$p.value[1:7])
lrtp <- append(lrtp, blank)
bicval <- round(BIC[1:9], digits = 2)
iclval <- round(icl[1:9], digits = 2)
listent <- round(rbind(ent1,ent2,ent3,ent4,ent5,ent6,ent7,ent8,ent9), digits = 2)
listlog <- round(rbind(logl1,logl2,logl3,logl4,logl5,logl6,logl7,logl8,logl9), digits = 2)
listfreep <- rbind(freep1,freep2,freep3,freep4,freep5,freep6,freep7,freep8,freep9)
listaic <- round(rbind(aic1,aic2,aic3,aic4,aic5,aic6,aic7,aic8,aic9), digits = 2)
listcaic <- round(rbind(caic1,caic2,caic3,caic4,caic5,caic6,caic7,caic8,caic9), digits = 2)
listsabic <- round(rbind(sabic1,sabic2,sabic3,sabic4,sabic5,sabic6,sabic7,sabic8,sabic9), digits = 2)
listperc <- rbind(
sum(allperc1 < .05),
sum(allperc2 < .05),
sum(allperc3 < .05),
sum(allperc4 < .05),
sum(allperc5 < .05),
sum(allperc6 < .05),
sum(allperc7 < .05),
sum(allperc8 < .05),
sum(allperc9 < .05))
indicestab <- cbind(modlist,listlog,listfreep,listaic,listcaic,bicval,listsabic,iclval,lrtobs,lrtp,listent,listperc)
colnames(indicestab) <- c("Model","LL","#FP","AIC","CAIC","BIC","SABIC","ICL","BLRTS","BLRTS p","Entropy","<5%")
rownames(indicestab) <- proflist
indicestab %>%
kbl() %>%
kable_styling(font_size = 11)| Model | LL | #FP | AIC | CAIC | BIC | SABIC | ICL | BLRTS | BLRTS p | Entropy | <5% | |
|---|---|---|---|---|---|---|---|---|---|---|---|---|
| Profiles 1 | XXX | -4519.04 | - | 9056.07 | 9110.37 | -9671.07 | 9072.78 | -9671.07 | - | - | 1 | 0 |
| Profiles 2 | VEV | -4285.58 | 10 | 8605.16 | 8707.71 | -9248.77 | 8636.72 | -9541 | 450.44 | 0.001 | 0.98 | 0 |
| Profiles 3 | VEV | -4218.65 | 14 | 8487.3 | 8638.12 | -9197.35 | 8533.71 | -9535.1 | 79.55 | 0.001 | 0.84 | 0 |
| Profiles 4 | VEV | -4128.84 | 18 | 8323.68 | 8522.75 | -9047.04 | 8384.94 | -9529.21 | 178.44 | 0.001 | 0.77 | 0 |
| Profiles 5 | VEV | -4123.64 | 5 | 8295.29 | 8440.07 | -8998.68 | 8339.84 | -9456.41 | 76.49 | 0.001 | 0.85 | 1 |
| Profiles 6 | VEV | -4099.49 | 6 | 8256.98 | 8431.93 | -8974.44 | 8310.82 | -9407.92 | 52.37 | 0.001 | 0.84 | 2 |
| Profiles 7 | VEV | -4080.04 | 7 | 8228.07 | 8433.18 | -8947.66 | 8291.19 | -9448.65 | 54.91 | 0.001 | 0.79 | 2 |
| Profiles 8 | VEV | -4045.21 | 8 | 8168.42 | 8403.69 | -8926.07 | 8240.82 | -9508.31 | 49.72 | 0.001 | 0.8 | 3 |
| Profiles 9 | VEV | -4009.91 | 46 | 8181.82 | 8670.46 | -8954.2 | 8332.18 | -9648.95 | - | - | 0.76 | 2 |
# save.image(file="LPAv6.RData")viz2 <- function(orig){
mcm <- orig
print(summary.Mclust(mcm))
d_std$class <- mcm$classification
df_long <- gather(d_std, key = "variable", value = "value", s_std, r_std, e_std)
summ <- summarySE(df_long, measurevar="value", groupvars = c("variable","class"), na.rm = T) #the summarySE() command creates a table with the information we need for our plot
# Create a barplot of means for each column by class
means <- ggplot(summ, aes(x = class, y = value, fill = variable)) +
geom_bar(stat = "summary", fun = "mean", position = "dodge") +
geom_errorbar(aes(ymin=value-se, ymax=value+se),
width=.2,
position=position_dodge(.9))
print(means)
df_long$class <- as.factor(df_long$class)
dist <- df_long %>%
ggplot(aes(x=value, fill=class)) +
geom_density(alpha=0.4, bw=.20) +
geom_vline(xintercept=0, size=.5, color="black") +
facet_wrap(~variable) +
xlim(-4.5,2)
print(dist)
vars <- unique(df_long$variable)
n <- 1
for (i in 1:length(vars)) {
print(vars[n])
print(subset(df_long, variable == vars[n]) %>%
ggplot(aes(x=value, fill=class)) +
geom_density(alpha=0.4, bw=.20) +
geom_vline(xintercept=0, size=.5, color="black") +
facet_wrap(~class) +
xlim(-4.5,2))
n <- n + 1
}
n <- 1
for (i in 1:length(unique(df_long$class))) {
df_long$class_spec <- 0
df_long$class_spec[d_std$class == n] <- 1
df_long$class_spec <- as.factor(df_long$class_spec)
scatter <- ggplot(df_long, aes(x = variable, y = value, color = class_spec)) +
geom_jitter()
print(paste("Class",n))
print(scatter)
n <- n + 1
}
class <- mcm$classification
uncer <- mcm$uncertainty
posterior_prob <- 1 - round(uncer, 5)
density <- plot(density(posterior_prob))
print(density)
}viz2(m5)----------------------------------------------------
Gaussian finite mixture model fitted by EM algorithm
----------------------------------------------------
Mclust VII (spherical, varying volume) model with 5 components:
log-likelihood n df BIC ICL
-4123.645 1133 24 -8416.073 -8812.255
Clustering table:
1 2 3 4 5
226 179 630 65 33 
[1] "s_std"
[1] "r_std"
[1] "e_std"
[1] "Class 1"
[1] "Class 2"
[1] "Class 3"
[1] "Class 4"
[1] "Class 5"
NULLviz2(m6)----------------------------------------------------
Gaussian finite mixture model fitted by EM algorithm
----------------------------------------------------
Mclust VII (spherical, varying volume) model with 6 components:
log-likelihood n df BIC ICL
-4099.492 1133 29 -8402.93 -8845.879
Clustering table:
1 2 3 4 5 6
231 198 554 78 39 33 
[1] "s_std"
[1] "r_std"
[1] "e_std"
[1] "Class 1"
[1] "Class 2"
[1] "Class 3"
[1] "Class 4"
[1] "Class 5"
[1] "Class 6"
NULL# classes <- cbind.data.frame(
# subset(d_fin, select=c(X, s_std,r_std,e_std)),
# subset(d_raw, select=c(s,r,e)),
# subset(d_fin, select=c(class,class_fin,pp)))
#
# write.csv(classes, file="classes.csv", row.names = F)
# write.csv(classes, file="classes2.csv", row.names = F)
rm(list = ls())
df <- read.csv(file="auxvar.csv", header=T)
lpa_df <- subset(df, select=c(R_IDs, s_std, r_std, e_std, class_fin))df_long <- gather(lpa_df, key = "variable", value = "value", s_std, r_std, e_std)
df_long$variable[df_long$variable == "s_std"] <- "Scientist Identity"
df_long$variable[df_long$variable == "r_std"] <- "Researcher Identity"
df_long$variable[df_long$variable == "e_std"] <- "Engineering Identity"
df_long$class_fin[df_long$class_fin == 1] <- "Dispersed GEI, n = 231"
df_long$class_fin[df_long$class_fin == 2] <- "Strong GEI, n = 198"
df_long$class_fin[df_long$class_fin == 3] <- "Moderate GEI, n = 554"
df_long$class_fin[df_long$class_fin == 4] <- "Mixed, Primarily Scientist, n = 78"
df_long$class_fin[df_long$class_fin == 5] <- "Mixed, Primarily Engineer, n = 39"
df_long$class_fin[df_long$class_fin == 6] <- "Maximum GEI, n = 33"
df_long$class_fin <- factor(df_long$class_fin, levels = c("Dispersed GEI, n = 231",
"Moderate GEI, n = 554",
"Strong GEI, n = 198",
"Mixed, Primarily Scientist, n = 78",
"Mixed, Primarily Engineer, n = 39",
"Maximum GEI, n = 33"))
lpa <- ggplot(df_long, aes(x = variable, y = value, color = factor(variable))) +
geom_violin(fill = "transparent", color = "black", alpha = 0, linewidth = .25) +
stat_summary(fun = mean, geom = "crossbar", width = 0.5,
middle.linewidth = 2, color = "black", linewidth = .4) +
geom_jitter(width = .1, height = .1, alpha = .25, size = 1.5) +
labs(y = "Identity Score", color = "Identity Domain:") +
facet_wrap(
~ class_fin,
nrow = 1,
labeller = label_wrap_gen(width = 12)
) +
xlab(NULL) +
theme_bw(base_size = 10) +
theme(
axis.text.x = element_blank(),
axis.ticks.x = element_blank(),
axis.title.y = element_text(size = 9),
axis.text.y = element_text(size = 7),
# legend at bottom, horizontal
legend.position = "bottom",
legend.direction = "horizontal",
legend.box = "horizontal",
legend.title = element_text(size = 9, lineheight = 0.9),
legend.text = element_text(size = 8, lineheight = 0.9),
legend.key.height = unit(0.9, "lines"),
legend.key.width = unit(1.2, "lines"),
strip.text = element_text(size = 8, lineheight = 0.9),
panel.grid.minor = element_blank()
) +
guides(
color = guide_legend(
override.aes = list(size = 3, alpha = 1),
nrow = 1
)
)
lpa <- lpa +
scale_color_viridis_d(option = "D", end = 0.9) +
scale_y_continuous(
breaks = function(x) seq(floor(min(x)), ceiling(3), by = 1)
)
lpa <- lpa +
theme(
axis.title.y = element_text(size = 11),
axis.text.y = element_text(size = 9),
legend.title = element_text(size = 11, lineheight = 0.9),
legend.text = element_text(size = 10, lineheight = 0.9),
strip.text = element_text(size = 10, lineheight = 0.9),
plot.background = element_rect(fill = "transparent", color = NA),
legend.background = element_rect(fill = "transparent", color = NA)
)
lpa
# ggsave("pres/lpa_vio.png", plot = lpa, bg = "transparent",
# width = 8, height = 5.5, dpi = 300)# order of sub-constructs within each participant
domain_order <- c("Engineering Identity", "Scientist Identity", "Researcher Identity")
# prep data: dispersed group only, ordered by engineering identity
dispersed <- df_long[df_long$class_fin == "Dispersed GEI, n = 231", ]
eng_order <- dispersed[dispersed$variable == "Engineering Identity",
c("R_IDs", "value")]
eng_order <- eng_order[order(eng_order$value), ]
eng_order$participant_rank <- seq_len(nrow(eng_order))
dispersed <- merge(dispersed, eng_order[, c("R_IDs", "participant_rank")], by = "R_IDs")
dispersed$variable <- factor(dispersed$variable, levels = domain_order)
dispersed$reveal_order <- (dispersed$participant_rank - 1) * 3 +
as.integer(dispersed$variable)
library(reshape)
dispersed_wide <- reshape(dispersed[, c("R_IDs", "variable", "value")],
idvar = "R_IDs", timevar = "variable", direction = "wide")
dispersed_wide$range <- apply(dispersed_wide[, -1], 1, function(x) max(x) - min(x))
# sort descending
dispersed_wide <- dispersed_wide[order(-dispersed_wide$range), ]
head(dispersed_wide) R_IDs value.Scientist Identity value.Engineering Identity
157 31714 1.559984 -3.7043428
457 69101 1.321765 -3.2481202
625 90025 -1.060419 -0.6248398
238 43265 1.083547 -3.2481202
391 61935 -5.110133 -1.4232295
448 68305 1.202656 -3.1340645
value.Researcher Identity range
157 1.208237 5.264327
457 1.208237 4.569886
625 -5.127389 4.502549
238 1.208237 4.456357
391 -0.757992 4.352141
448 1.099002 4.336721dispersed_wide$group <- cut(rank(-dispersed_wide$range),
breaks = 6,
labels = c("Group 1 (Largest Differences)",
"Group 2", "Group 3",
"Group 4", "Group 5",
"Group 6 (Smallest Differences)"))
dispersed_sub <- dispersed[dispersed$R_IDs %in% dispersed_wide$R_IDs, ]
dispersed_sub <- merge(dispersed_sub, dispersed_wide[, c("R_IDs", "group", "range")], by = "R_IDs")
dispersed_sub$variable <- factor(dispersed_sub$variable,
levels = c("Engineering Identity",
"Researcher Identity",
"Scientist Identity"))
# identify top 5 most and least dispersed within the full dispersed group
top5_ids <- head(dispersed_wide$R_IDs, 5)
bot5_ids <- tail(dispersed_wide$R_IDs, 5)
dispersed_sub$highlight <- ifelse(dispersed_sub$R_IDs %in% top5_ids, "high",
ifelse(dispersed_sub$R_IDs %in% bot5_ids, "low", "none"))
disp <- ggplot(dispersed_sub, aes(x = variable, y = value, color = factor(variable))) +
geom_point(alpha = .6, size = 2) +
geom_line(aes(group = R_IDs, alpha = highlight, linewidth = highlight), color = "grey30") +
scale_alpha_manual(values = c("high" = .9, "low" = .9, "none" = .25), guide = "none") +
scale_linewidth_manual(values = c("high" = .8, "low" = .8, "none" = .2), guide = "none") +
facet_wrap(~ group, nrow = 2) +
labs(y = "Identity Score", color = "Identity Domain:") +
xlab(NULL) +
scale_color_viridis_d(option = "D", end = 0.9) +
scale_y_continuous(
breaks = function(x) seq(floor(min(x)), ceiling(3), by = 1)
) +
theme_bw(base_size = 10) +
theme(
axis.text.x = element_blank(),
axis.ticks.x = element_blank(),
axis.title.y = element_text(size = 11),
axis.text.y = element_text(size = 9),
legend.position = "bottom",
legend.direction = "horizontal",
plot.background = element_rect(fill = "transparent", color = NA),
legend.background = element_rect(fill = "transparent", color = NA),
panel.grid.minor = element_blank(),
legend.title = element_text(size = 11, lineheight = 0.9),
legend.text = element_text(size = 10, lineheight = 0.9),
strip.text = element_text(size = 10, lineheight = 0.9),
) +
labs(title = "Differences in Domain Scores Among Dispersed GEI Participants")
disp
# ggsave("pres/disp.png", plot = disp, bg = "transparent",
# width = 8, height = 5.5, dpi = 300)rm(list = ls())
import <- read.csv(file="finaldata.csv", header=T, na.strings = c(""," ","NA"))
df <- subset(import, select=c(R_IDs, Major, Start.Year.Month, Q19.3, Q19.16, Q19.16_7_TEXT, Q19.13, Q19.13_8_TEXT, Q13.1_1, Q13.1_2, Q19.11))
colnames(df) <- c("R_IDs","Major","start","progress","gender","gendertxt","race","racetxt","dadp","itp","intl")
classes <- read.csv(file="classes.csv", header=T)
classes <- merge(classes, subset(import, select=c(R_IDs, X)), by="X")
classes$class_bu <- classes$class
classes$class[classes$class_bu == "2"] <- "3"
classes$class[classes$class_bu == "3"] <- "2"table(df$Major, useNA = "always")
Acoustic Engineering
2
Aerospace Engineering
50
Agricultural and Biological Engineering
47
Biological Engineering
2
Biomedical
128
Biomedical Engineering
3
Chemical and Biomolecular Engineering
18
Chemical and Materials Engineering
2
Chemical Engineering
170
Civil and Environmental Engineering
18
Civil Engineering
142
Computer Engineering
33
Computer Science
93
Computer Science and Engineering
21
Computer Science Engineering
4
Electrical & Computer Engineering
62
Electrical and Electronic Engineering
2
Electrical Engineering
200
Electrical Engineering and Computer Science
2
Engineering
2
Engineering Education
1
Engineering Management
4
Engineering Physics
5
Environmental
51
Environmental Engineering
1
Geo
6
Industrial
40
Material Science and Engineering
185
Mechanical Engineering
260
Nuclear Engineering
59
Ocean Engineering
11
Other Engineering
3
Petroleum
17
Systems Engineering
10
<NA>
100 df$maj_fin <- df$Major
df$maj_fin[df$Major == "Acoustic Engineering"] <- "Other Engineering" #n = 2
df$maj_fin[df$Major == "Engineering"] <- "Other Engineering" #n = 2
df$maj_fin[df$Major == "Engineering Education"] <- "Other Engineering" #n = 1
df$maj_fin[df$Major == "Engineering Management"] <- "Other Engineering" #n = 4
df$maj_fin[df$Major == "Engineering Physics"] <- "Other Engineering" #n = 5
df$maj_fin[df$Major == "Geo"] <- "Other Engineering" #n = 6
df$maj_fin[df$Major == "Other Engineering"] <- "Other Engineering" #n = 3
df$maj_fin[df$Major == "Biological Engineering"] <- "Bio Engineering"
df$maj_fin[df$Major == "Biomedical"] <- "Bio Engineering"
df$maj_fin[df$Major == "Biomedical Engineering"] <- "Bio Engineering"
df$maj_fin[df$Major == "Chemical and Biomolecular Engineering"] <- "Chemical Engineering"
df$maj_fin[df$Major == "Chemical and Materials Engineering"] <- "Chemical Engineering"
df$maj_fin[df$Major == "Civil and Environmental Engineering"] <- "Civil Engineering"
df$maj_fin[df$Major == "Computer Science Engineering"] <- "Computer Science and Engineering"
df$maj_fin[df$Major == "Electrical & Computer Engineering"] <- "Elec Engineering"
df$maj_fin[df$Major == "Electrical and Electronic Engineering"] <- "Elec Engineering"
df$maj_fin[df$Major == "Electrical Engineering"] <- "Elec Engineering"
df$maj_fin[df$Major == "Electrical Engineering and Computer Science"] <- "Elec Engineering"
df$maj_fin[df$Major == "Environmental"] <- "Environmental Engineering"
df$maj_fin[df$Major == "Industrial"] <- "Industrial Engineering"
df$maj_fin[df$Major == "Ocean Engineering"] <- "Other Engineering"
df$maj_fin[df$Major == "Petroleum"] <- "Other Engineering"
df$maj_fin[df$Major == "Systems Engineering"] <- "Other Engineering"
table(df$Major, df$maj_fin, useNA = "always")
Aerospace Engineering
Acoustic Engineering 0
Aerospace Engineering 50
Agricultural and Biological Engineering 0
Biological Engineering 0
Biomedical 0
Biomedical Engineering 0
Chemical and Biomolecular Engineering 0
Chemical and Materials Engineering 0
Chemical Engineering 0
Civil and Environmental Engineering 0
Civil Engineering 0
Computer Engineering 0
Computer Science 0
Computer Science and Engineering 0
Computer Science Engineering 0
Electrical & Computer Engineering 0
Electrical and Electronic Engineering 0
Electrical Engineering 0
Electrical Engineering and Computer Science 0
Engineering 0
Engineering Education 0
Engineering Management 0
Engineering Physics 0
Environmental 0
Environmental Engineering 0
Geo 0
Industrial 0
Material Science and Engineering 0
Mechanical Engineering 0
Nuclear Engineering 0
Ocean Engineering 0
Other Engineering 0
Petroleum 0
Systems Engineering 0
<NA> 0
Agricultural and Biological Engineering
Acoustic Engineering 0
Aerospace Engineering 0
Agricultural and Biological Engineering 47
Biological Engineering 0
Biomedical 0
Biomedical Engineering 0
Chemical and Biomolecular Engineering 0
Chemical and Materials Engineering 0
Chemical Engineering 0
Civil and Environmental Engineering 0
Civil Engineering 0
Computer Engineering 0
Computer Science 0
Computer Science and Engineering 0
Computer Science Engineering 0
Electrical & Computer Engineering 0
Electrical and Electronic Engineering 0
Electrical Engineering 0
Electrical Engineering and Computer Science 0
Engineering 0
Engineering Education 0
Engineering Management 0
Engineering Physics 0
Environmental 0
Environmental Engineering 0
Geo 0
Industrial 0
Material Science and Engineering 0
Mechanical Engineering 0
Nuclear Engineering 0
Ocean Engineering 0
Other Engineering 0
Petroleum 0
Systems Engineering 0
<NA> 0
Bio Engineering
Acoustic Engineering 0
Aerospace Engineering 0
Agricultural and Biological Engineering 0
Biological Engineering 2
Biomedical 128
Biomedical Engineering 3
Chemical and Biomolecular Engineering 0
Chemical and Materials Engineering 0
Chemical Engineering 0
Civil and Environmental Engineering 0
Civil Engineering 0
Computer Engineering 0
Computer Science 0
Computer Science and Engineering 0
Computer Science Engineering 0
Electrical & Computer Engineering 0
Electrical and Electronic Engineering 0
Electrical Engineering 0
Electrical Engineering and Computer Science 0
Engineering 0
Engineering Education 0
Engineering Management 0
Engineering Physics 0
Environmental 0
Environmental Engineering 0
Geo 0
Industrial 0
Material Science and Engineering 0
Mechanical Engineering 0
Nuclear Engineering 0
Ocean Engineering 0
Other Engineering 0
Petroleum 0
Systems Engineering 0
<NA> 0
Chemical Engineering
Acoustic Engineering 0
Aerospace Engineering 0
Agricultural and Biological Engineering 0
Biological Engineering 0
Biomedical 0
Biomedical Engineering 0
Chemical and Biomolecular Engineering 18
Chemical and Materials Engineering 2
Chemical Engineering 170
Civil and Environmental Engineering 0
Civil Engineering 0
Computer Engineering 0
Computer Science 0
Computer Science and Engineering 0
Computer Science Engineering 0
Electrical & Computer Engineering 0
Electrical and Electronic Engineering 0
Electrical Engineering 0
Electrical Engineering and Computer Science 0
Engineering 0
Engineering Education 0
Engineering Management 0
Engineering Physics 0
Environmental 0
Environmental Engineering 0
Geo 0
Industrial 0
Material Science and Engineering 0
Mechanical Engineering 0
Nuclear Engineering 0
Ocean Engineering 0
Other Engineering 0
Petroleum 0
Systems Engineering 0
<NA> 0
Civil Engineering
Acoustic Engineering 0
Aerospace Engineering 0
Agricultural and Biological Engineering 0
Biological Engineering 0
Biomedical 0
Biomedical Engineering 0
Chemical and Biomolecular Engineering 0
Chemical and Materials Engineering 0
Chemical Engineering 0
Civil and Environmental Engineering 18
Civil Engineering 142
Computer Engineering 0
Computer Science 0
Computer Science and Engineering 0
Computer Science Engineering 0
Electrical & Computer Engineering 0
Electrical and Electronic Engineering 0
Electrical Engineering 0
Electrical Engineering and Computer Science 0
Engineering 0
Engineering Education 0
Engineering Management 0
Engineering Physics 0
Environmental 0
Environmental Engineering 0
Geo 0
Industrial 0
Material Science and Engineering 0
Mechanical Engineering 0
Nuclear Engineering 0
Ocean Engineering 0
Other Engineering 0
Petroleum 0
Systems Engineering 0
<NA> 0
Computer Engineering
Acoustic Engineering 0
Aerospace Engineering 0
Agricultural and Biological Engineering 0
Biological Engineering 0
Biomedical 0
Biomedical Engineering 0
Chemical and Biomolecular Engineering 0
Chemical and Materials Engineering 0
Chemical Engineering 0
Civil and Environmental Engineering 0
Civil Engineering 0
Computer Engineering 33
Computer Science 0
Computer Science and Engineering 0
Computer Science Engineering 0
Electrical & Computer Engineering 0
Electrical and Electronic Engineering 0
Electrical Engineering 0
Electrical Engineering and Computer Science 0
Engineering 0
Engineering Education 0
Engineering Management 0
Engineering Physics 0
Environmental 0
Environmental Engineering 0
Geo 0
Industrial 0
Material Science and Engineering 0
Mechanical Engineering 0
Nuclear Engineering 0
Ocean Engineering 0
Other Engineering 0
Petroleum 0
Systems Engineering 0
<NA> 0
Computer Science
Acoustic Engineering 0
Aerospace Engineering 0
Agricultural and Biological Engineering 0
Biological Engineering 0
Biomedical 0
Biomedical Engineering 0
Chemical and Biomolecular Engineering 0
Chemical and Materials Engineering 0
Chemical Engineering 0
Civil and Environmental Engineering 0
Civil Engineering 0
Computer Engineering 0
Computer Science 93
Computer Science and Engineering 0
Computer Science Engineering 0
Electrical & Computer Engineering 0
Electrical and Electronic Engineering 0
Electrical Engineering 0
Electrical Engineering and Computer Science 0
Engineering 0
Engineering Education 0
Engineering Management 0
Engineering Physics 0
Environmental 0
Environmental Engineering 0
Geo 0
Industrial 0
Material Science and Engineering 0
Mechanical Engineering 0
Nuclear Engineering 0
Ocean Engineering 0
Other Engineering 0
Petroleum 0
Systems Engineering 0
<NA> 0
Computer Science and Engineering
Acoustic Engineering 0
Aerospace Engineering 0
Agricultural and Biological Engineering 0
Biological Engineering 0
Biomedical 0
Biomedical Engineering 0
Chemical and Biomolecular Engineering 0
Chemical and Materials Engineering 0
Chemical Engineering 0
Civil and Environmental Engineering 0
Civil Engineering 0
Computer Engineering 0
Computer Science 0
Computer Science and Engineering 21
Computer Science Engineering 4
Electrical & Computer Engineering 0
Electrical and Electronic Engineering 0
Electrical Engineering 0
Electrical Engineering and Computer Science 0
Engineering 0
Engineering Education 0
Engineering Management 0
Engineering Physics 0
Environmental 0
Environmental Engineering 0
Geo 0
Industrial 0
Material Science and Engineering 0
Mechanical Engineering 0
Nuclear Engineering 0
Ocean Engineering 0
Other Engineering 0
Petroleum 0
Systems Engineering 0
<NA> 0
Elec Engineering
Acoustic Engineering 0
Aerospace Engineering 0
Agricultural and Biological Engineering 0
Biological Engineering 0
Biomedical 0
Biomedical Engineering 0
Chemical and Biomolecular Engineering 0
Chemical and Materials Engineering 0
Chemical Engineering 0
Civil and Environmental Engineering 0
Civil Engineering 0
Computer Engineering 0
Computer Science 0
Computer Science and Engineering 0
Computer Science Engineering 0
Electrical & Computer Engineering 62
Electrical and Electronic Engineering 2
Electrical Engineering 200
Electrical Engineering and Computer Science 2
Engineering 0
Engineering Education 0
Engineering Management 0
Engineering Physics 0
Environmental 0
Environmental Engineering 0
Geo 0
Industrial 0
Material Science and Engineering 0
Mechanical Engineering 0
Nuclear Engineering 0
Ocean Engineering 0
Other Engineering 0
Petroleum 0
Systems Engineering 0
<NA> 0
Environmental Engineering
Acoustic Engineering 0
Aerospace Engineering 0
Agricultural and Biological Engineering 0
Biological Engineering 0
Biomedical 0
Biomedical Engineering 0
Chemical and Biomolecular Engineering 0
Chemical and Materials Engineering 0
Chemical Engineering 0
Civil and Environmental Engineering 0
Civil Engineering 0
Computer Engineering 0
Computer Science 0
Computer Science and Engineering 0
Computer Science Engineering 0
Electrical & Computer Engineering 0
Electrical and Electronic Engineering 0
Electrical Engineering 0
Electrical Engineering and Computer Science 0
Engineering 0
Engineering Education 0
Engineering Management 0
Engineering Physics 0
Environmental 51
Environmental Engineering 1
Geo 0
Industrial 0
Material Science and Engineering 0
Mechanical Engineering 0
Nuclear Engineering 0
Ocean Engineering 0
Other Engineering 0
Petroleum 0
Systems Engineering 0
<NA> 0
Industrial Engineering
Acoustic Engineering 0
Aerospace Engineering 0
Agricultural and Biological Engineering 0
Biological Engineering 0
Biomedical 0
Biomedical Engineering 0
Chemical and Biomolecular Engineering 0
Chemical and Materials Engineering 0
Chemical Engineering 0
Civil and Environmental Engineering 0
Civil Engineering 0
Computer Engineering 0
Computer Science 0
Computer Science and Engineering 0
Computer Science Engineering 0
Electrical & Computer Engineering 0
Electrical and Electronic Engineering 0
Electrical Engineering 0
Electrical Engineering and Computer Science 0
Engineering 0
Engineering Education 0
Engineering Management 0
Engineering Physics 0
Environmental 0
Environmental Engineering 0
Geo 0
Industrial 40
Material Science and Engineering 0
Mechanical Engineering 0
Nuclear Engineering 0
Ocean Engineering 0
Other Engineering 0
Petroleum 0
Systems Engineering 0
<NA> 0
Material Science and Engineering
Acoustic Engineering 0
Aerospace Engineering 0
Agricultural and Biological Engineering 0
Biological Engineering 0
Biomedical 0
Biomedical Engineering 0
Chemical and Biomolecular Engineering 0
Chemical and Materials Engineering 0
Chemical Engineering 0
Civil and Environmental Engineering 0
Civil Engineering 0
Computer Engineering 0
Computer Science 0
Computer Science and Engineering 0
Computer Science Engineering 0
Electrical & Computer Engineering 0
Electrical and Electronic Engineering 0
Electrical Engineering 0
Electrical Engineering and Computer Science 0
Engineering 0
Engineering Education 0
Engineering Management 0
Engineering Physics 0
Environmental 0
Environmental Engineering 0
Geo 0
Industrial 0
Material Science and Engineering 185
Mechanical Engineering 0
Nuclear Engineering 0
Ocean Engineering 0
Other Engineering 0
Petroleum 0
Systems Engineering 0
<NA> 0
Mechanical Engineering
Acoustic Engineering 0
Aerospace Engineering 0
Agricultural and Biological Engineering 0
Biological Engineering 0
Biomedical 0
Biomedical Engineering 0
Chemical and Biomolecular Engineering 0
Chemical and Materials Engineering 0
Chemical Engineering 0
Civil and Environmental Engineering 0
Civil Engineering 0
Computer Engineering 0
Computer Science 0
Computer Science and Engineering 0
Computer Science Engineering 0
Electrical & Computer Engineering 0
Electrical and Electronic Engineering 0
Electrical Engineering 0
Electrical Engineering and Computer Science 0
Engineering 0
Engineering Education 0
Engineering Management 0
Engineering Physics 0
Environmental 0
Environmental Engineering 0
Geo 0
Industrial 0
Material Science and Engineering 0
Mechanical Engineering 260
Nuclear Engineering 0
Ocean Engineering 0
Other Engineering 0
Petroleum 0
Systems Engineering 0
<NA> 0
Nuclear Engineering
Acoustic Engineering 0
Aerospace Engineering 0
Agricultural and Biological Engineering 0
Biological Engineering 0
Biomedical 0
Biomedical Engineering 0
Chemical and Biomolecular Engineering 0
Chemical and Materials Engineering 0
Chemical Engineering 0
Civil and Environmental Engineering 0
Civil Engineering 0
Computer Engineering 0
Computer Science 0
Computer Science and Engineering 0
Computer Science Engineering 0
Electrical & Computer Engineering 0
Electrical and Electronic Engineering 0
Electrical Engineering 0
Electrical Engineering and Computer Science 0
Engineering 0
Engineering Education 0
Engineering Management 0
Engineering Physics 0
Environmental 0
Environmental Engineering 0
Geo 0
Industrial 0
Material Science and Engineering 0
Mechanical Engineering 0
Nuclear Engineering 59
Ocean Engineering 0
Other Engineering 0
Petroleum 0
Systems Engineering 0
<NA> 0
Other Engineering <NA>
Acoustic Engineering 2 0
Aerospace Engineering 0 0
Agricultural and Biological Engineering 0 0
Biological Engineering 0 0
Biomedical 0 0
Biomedical Engineering 0 0
Chemical and Biomolecular Engineering 0 0
Chemical and Materials Engineering 0 0
Chemical Engineering 0 0
Civil and Environmental Engineering 0 0
Civil Engineering 0 0
Computer Engineering 0 0
Computer Science 0 0
Computer Science and Engineering 0 0
Computer Science Engineering 0 0
Electrical & Computer Engineering 0 0
Electrical and Electronic Engineering 0 0
Electrical Engineering 0 0
Electrical Engineering and Computer Science 0 0
Engineering 2 0
Engineering Education 1 0
Engineering Management 4 0
Engineering Physics 5 0
Environmental 0 0
Environmental Engineering 0 0
Geo 6 0
Industrial 0 0
Material Science and Engineering 0 0
Mechanical Engineering 0 0
Nuclear Engineering 0 0
Ocean Engineering 11 0
Other Engineering 3 0
Petroleum 17 0
Systems Engineering 10 0
<NA> 0 100table(df$maj_fin, useNA = "always")
Aerospace Engineering Agricultural and Biological Engineering
50 47
Bio Engineering Chemical Engineering
133 190
Civil Engineering Computer Engineering
160 33
Computer Science Computer Science and Engineering
93 25
Elec Engineering Environmental Engineering
266 52
Industrial Engineering Material Science and Engineering
40 185
Mechanical Engineering Nuclear Engineering
260 59
Other Engineering <NA>
61 100 # survey closed 7/25/2018
spl <- colsplit(df$start," ",c("year", "month"))
spl$year <- as.numeric(spl$year)
spl$years <- 2018-spl$year
df$years <- spl$yearstable(df$progress, useNA="always")
1 1,2 1,2,3 1,3 1,4 2 2,3 2,4 3 4 <NA>
398 254 41 1 3 36 4 3 9 803 202 # VALUES: 1 through 4, multiples possible; 1 = “Comprehensive Exam (Written or Oral)
# 2 = “Dissertation Proposal”, 3 = “Dissertation Defense”, 4 = “Not applicable”
df$prog_fin <- NA
df$prog_fin[df$progress == "2,4"] <- 2
df$prog_fin[df$progress == "2,3"] <- 3
df$prog_fin[df$progress == "1,4"] <- 1
df$prog_fin[df$progress == "1,3"] <- 3
df$prog_fin[df$progress == "1,2,3"] <- 3
df$prog_fin[df$progress == "1,2"] <- 2
df$prog_fin[df$progress == "1"] <- 1
df$prog_fin[df$progress == "2"] <- 2
df$prog_fin[df$progress == "3"] <- 3
df$prog_fin[df$progress == "4"] <- 0
table(df$prog_fin, useNA = "always")
0 1 2 3 <NA>
803 401 293 55 202 table(df$gender, useNA = "always")
1 1,2,3,4,5,6,7 1,3 1,3,5 1,4
541 1 3 1 1
1,5 1,6 2 2,3 2,3,5
1 23 987 5 1
2,5 2,6 2,7 3 5
1 33 6 2 1
6 7 <NA>
1 12 134 # VALUES: 1 through 7, multiples possible; 1 = “Female”, 2 = “Male”, 3 = “Genderqueer”
# 4 = “Agender”, 5 = “Transgender”, 6 = “Cisgender”, 7 = “A gender not listed” (Write in response optional)
df$gen_fin <- NA
df$gen_fin[df$gender == 1 | df$gender == "1,6" | df$gender == "1,5"] <- "female"
df$gen_fin[df$gender == 2 | df$gender == "2,6" | df$gender == "2,5"] <- "male"
df$gen_fin[df$gender == "1,3,5"] <- "gq"
df$gen_fin[df$gender == "2,3,5"] <- "gq"
df$gen_fin[df$gender == 3 | df$gender == 4 | df$gender == 7 | df$gender == "2,3" | df$gender == "1,3" | df$gender == "2,7" | df$gender == "1,4"] <- "gq"
table(df$gen_fin, df$gender, useNA = "always")
1 1,2,3,4,5,6,7 1,3 1,3,5 1,4 1,5 1,6 2 2,3 2,3,5 2,5 2,6 2,7 3
female 541 0 0 0 0 1 23 0 0 0 0 0 0 0
gq 0 0 3 1 1 0 0 0 5 1 0 0 6 2
male 0 0 0 0 0 0 0 987 0 0 1 33 0 0
<NA> 0 1 0 0 0 0 0 0 0 0 0 0 0 0
5 6 7 <NA>
female 0 0 0 0
gq 0 0 12 0
male 0 0 0 0
<NA> 1 1 0 134table(df$gen_fin, useNA = "always")
female gq male <NA>
565 31 1021 137 table(df$gendertxt)
A zorplon from Sagitarius-7
1
Alpha Male
1
An ether sniffing robot sex toy on wheels
1
Attack helicopter
1
Attack Helicopter
1
Hairdresser
1
HOKIE
1
Non-binary
1
Not relevant
1
Not Relevant
1
questioning
1
Sexist question
1
there are only two genders
1
there are only two genders fuck off
1
This is a silly option. Your biology determines this. That's science, that's fact. A better option is "what I feel my gender is".
1
Turtle
1
XX, XY, XO, XXY, XXX, or XYY the rest is made up
1 # legitimate responses: "Non-binary" - already recoded as gq, all else as NAtable(df$race, useNA = "always")
1,2 1,3 1,3,4 1,3,7 1,4,5,7,8 1,4,7 1,7 2
1 1 1 1 1 3 4 474
2,3,5,7 2,4 2,4,6,7 2,4,7 2,5,7 2,6 2,6,7 2,7
1 3 1 3 2 1 1 30
2,8 3 3,4 3,6 3,7 3,8 4 4,7
1 37 2 1 2 1 53 36
4,7,8 4,8 5 5,7 6 6,7 7 7,8
1 2 49 21 1 1 846 2
8 <NA>
25 145 # 1 = “American Indian or Alaska Native”, 2 = “Asian”, 3 = “Black or African American”
# 4 = “Hispanic, Latino/Latina/Latinx, or Spanish origin”, 5 = “Middle Eastern or North African”
# 6 = “Native Hawaiian or Other Pacific Islander”, 7 = “White”, 8 = “Another race or ethnicity not listed above” (Write in response optional)
df$race_fin <- "Bi/Multiracial"
df$race_fin[df$race == "2"] <- c("Asian")
df$race_fin[df$race == "3"] <- c("Black or African American")
df$race_fin[df$race == "4"] <- c("Hispanic, Latino/Latina/Latinx, or Spanish origin")
df$race_fin[df$race == "5"] <- c("Middle Eastern or North African")
df$race_fin[df$race == "7"] <- c("White")
df$race_fin[df$race == "6"] <- c("Native Hawaiian or Other Pacific Islander")
df$race_fin[df$race == "8"] <- c("Another race or ethnicity not listed above")
df$race_fin[is.na(df$race)] <- NA
# df$race_fin_temp <- 1
# df$race_fin_temp[!is.na(df$race_fin)] <- 2
# table(df$race, df$race_fin_temp)
table(df$racetxt)
--
1
AS THE DRIVEN SNOW
1
Asian
1
Basque
1
Black African/French
1
Caucasian (Pakistani)
1
Demographics are racist
1
Earthling
1
HOKIE
1
I highlighted groups that are closely related to mine throughout history/migration. If places are going to put "white", put european american as well. It's only being consistent. Speaking of which, hispanics are europeans, caucasian, white etc. This is all silly. Race/ethnicity is a personal identification, not a tool or weapon which this question aims to use it as.
1
I identify as a race and I don't want to divulge that info
1
Indian
2
Iranian
1
Israelite
1
Nepalese
1
Not relevant
1
South American Indian
1
south asian
1
South asian
1
South Asian
2
southern asian
1
this question is racist
1
Unaffiliated
1
unknown
1 # legitimate entries: "Basque", "Black African/French", "Caucasian (Pakistani)", "Indian", "Iranian", "Israelite", "Nepalese", "South American Indian", "south asian", "South asian", "South Asian", "southern asian"
df$race_fin[df$racetxt == "Indian"] <- c("Asian")
df$race_fin[df$racetxt == "Nepalese"] <- c("Asian")
df$race_fin[df$racetxt == "south asian"] <- c("Asian")
df$race_fin[df$racetxt == "South asian"] <- c("Asian")
df$race_fin[df$racetxt == "South Asian"] <- c("Asian")
df$race_fin[df$racetxt == "southern asian"] <- c("Asian")
table(df$race_fin, useNA = "always")
Another race or ethnicity not listed above
18
Asian
482
Bi/Multiracial
123
Black or African American
37
Hispanic, Latino/Latina/Latinx, or Spanish origin
53
Middle Eastern or North African
49
Native Hawaiian or Other Pacific Islander
1
White
846
<NA>
145 df$race_fin2 <- df$race_fin
df$race_fin2[df$race_fin == "Another race or ethnicity not listed above"] <- "URM"
df$race_fin2[df$race_fin == "Native Hawaiian or Other Pacific Islander"] <- "URM"
df$race_fin2[df$race_fin == "Black or African American"] <- "URM"
df$race_fin2[df$race_fin == "Hispanic, Latino/Latina/Latinx, or Spanish origin"] <- "URM"
df$race_fin2[df$race_fin == "Middle Eastern or North African"] <- "URM"
table(df$race_fin2, useNA = "always")
Asian Bi/Multiracial URM White <NA>
482 123 158 846 145 table(df$intl, useNA = "always")
1 2 <NA>
1073 573 108 df$intl_fin <- "international"
df$intl_fin[df$intl == 1] <- "domestic"
df$intl_fin[is.na(df$intl)] <- NA
table(df$intl_fin, useNA = "always")
domestic international <NA>
1073 573 108 table(df$dadp, useNA = "always")
1 2 3 4 5 <NA>
271 441 382 402 222 36 table(df$itp, useNA = "always")
1 2 3 4 5 <NA>
18 22 69 221 1389 35 corr.test(df$dadp, df$itp)Call:corr.test(x = df$dadp, y = df$itp)
Correlation matrix
[1] -0.11
Sample Size
[1] 1718
These are the unadjusted probability values.
The probability values adjusted for multiple tests are in the p.adj object.
[1] 0
To see confidence intervals of the correlations, print with the short=FALSE optiondf$itp_rc[df$itp == 1] <- 5
df$itp_rc[df$itp == 2] <- 4
df$itp_rc[df$itp == 3] <- 3
df$itp_rc[df$itp == 4] <- 2
df$itp_rc[df$itp == 5] <- 1
df$uncer <- (df$dadp + df$itp_rc)/2names(df) [1] "R_IDs" "Major" "start" "progress" "gender" "gendertxt"
[7] "race" "racetxt" "dadp" "itp" "intl" "maj_fin"
[13] "years" "prog_fin" "gen_fin" "race_fin" "race_fin2" "intl_fin"
[19] "itp_rc" "uncer" df2 <- subset(df, select=c(R_IDs, maj_fin, years, prog_fin, gen_fin, race_fin, race_fin2, intl_fin, dadp, itp, uncer))
d <- merge(df2, classes, by = "R_IDs")
rm(import, spl, df, df2, classes)
# write.csv(d, file="auxvar.csv", row.names = F)table(d$prog_fin, useNA = "always")
0 1 2 3 <NA>
324 373 286 51 99 prop.table(table(d$prog_fin, useNA = "always"))
0 1 2 3 <NA>
0.28596646 0.32921447 0.25242718 0.04501324 0.08737864 table(d$gen_fin, useNA = "always")
female gq male <NA>
398 15 692 28 prop.table(table(d$gen_fin, useNA = "always"))
female gq male <NA>
0.35127979 0.01323919 0.61076787 0.02471315 table(d$race_fin, useNA = "always")
Another race or ethnicity not listed above
10
Asian
315
Bi/Multiracial
90
Black or African American
23
Hispanic, Latino/Latina/Latinx, or Spanish origin
36
Middle Eastern or North African
44
Native Hawaiian or Other Pacific Islander
1
White
581
<NA>
33 prop.table(table(d$race_fin, useNA = "always"))
Another race or ethnicity not listed above
0.0088261253
Asian
0.2780229479
Bi/Multiracial
0.0794351280
Black or African American
0.0203000883
Hispanic, Latino/Latina/Latinx, or Spanish origin
0.0317740512
Middle Eastern or North African
0.0388349515
Native Hawaiian or Other Pacific Islander
0.0008826125
White
0.5127978817
<NA>
0.0291262136 table(d$race_fin2, useNA = "always")
Asian Bi/Multiracial URM White <NA>
315 90 114 581 33 table(d$intl_fin, useNA = "always")
domestic international <NA>
718 405 10 prop.table(table(d$intl_fin, useNA = "always"))
domestic international <NA>
0.633715799 0.357458076 0.008826125 table(d$maj_fin, useNA = "always")
Aerospace Engineering Agricultural and Biological Engineering
28 27
Bio Engineering Chemical Engineering
102 149
Civil Engineering Computer Engineering
100 19
Computer Science Computer Science and Engineering
55 21
Elec Engineering Environmental Engineering
157 28
Industrial Engineering Material Science and Engineering
29 162
Mechanical Engineering Nuclear Engineering
165 44
Other Engineering <NA>
34 13 d$maj_fin2 <- d$maj_fin
d$maj_fin2[d$maj_fin == "Computer Engineering"] <- "Computer Science and Engineering"
d$maj_fin2[d$maj_fin == "Computer Science"] <- "Computer Science and Engineering"
d$maj_fin2[d$maj_fin == "Agricultural and Biological Engineering"] <- "Bio Engineering"
d$maj_fin2[d$maj_fin == "Aerospace Engineering"] <- "Other Engineering"
d$maj_fin2[d$maj_fin == "Environmental Engineering"] <- "Other Engineering"
d$maj_fin2[d$maj_fin == "Industrial Engineering"] <- "Other Engineering"
d$maj_fin2[d$maj_fin == "Nuclear Engineering"] <- "Other Engineering"
d$maj_fin2[is.na(d$maj_fin)] <- "Other Engineering"
table(d$maj_fin, d$maj_fin2, useNA = "always")
Bio Engineering Chemical Engineering
Aerospace Engineering 0 0
Agricultural and Biological Engineering 27 0
Bio Engineering 102 0
Chemical Engineering 0 149
Civil Engineering 0 0
Computer Engineering 0 0
Computer Science 0 0
Computer Science and Engineering 0 0
Elec Engineering 0 0
Environmental Engineering 0 0
Industrial Engineering 0 0
Material Science and Engineering 0 0
Mechanical Engineering 0 0
Nuclear Engineering 0 0
Other Engineering 0 0
<NA> 0 0
Civil Engineering
Aerospace Engineering 0
Agricultural and Biological Engineering 0
Bio Engineering 0
Chemical Engineering 0
Civil Engineering 100
Computer Engineering 0
Computer Science 0
Computer Science and Engineering 0
Elec Engineering 0
Environmental Engineering 0
Industrial Engineering 0
Material Science and Engineering 0
Mechanical Engineering 0
Nuclear Engineering 0
Other Engineering 0
<NA> 0
Computer Science and Engineering
Aerospace Engineering 0
Agricultural and Biological Engineering 0
Bio Engineering 0
Chemical Engineering 0
Civil Engineering 0
Computer Engineering 19
Computer Science 55
Computer Science and Engineering 21
Elec Engineering 0
Environmental Engineering 0
Industrial Engineering 0
Material Science and Engineering 0
Mechanical Engineering 0
Nuclear Engineering 0
Other Engineering 0
<NA> 0
Elec Engineering
Aerospace Engineering 0
Agricultural and Biological Engineering 0
Bio Engineering 0
Chemical Engineering 0
Civil Engineering 0
Computer Engineering 0
Computer Science 0
Computer Science and Engineering 0
Elec Engineering 157
Environmental Engineering 0
Industrial Engineering 0
Material Science and Engineering 0
Mechanical Engineering 0
Nuclear Engineering 0
Other Engineering 0
<NA> 0
Material Science and Engineering
Aerospace Engineering 0
Agricultural and Biological Engineering 0
Bio Engineering 0
Chemical Engineering 0
Civil Engineering 0
Computer Engineering 0
Computer Science 0
Computer Science and Engineering 0
Elec Engineering 0
Environmental Engineering 0
Industrial Engineering 0
Material Science and Engineering 162
Mechanical Engineering 0
Nuclear Engineering 0
Other Engineering 0
<NA> 0
Mechanical Engineering
Aerospace Engineering 0
Agricultural and Biological Engineering 0
Bio Engineering 0
Chemical Engineering 0
Civil Engineering 0
Computer Engineering 0
Computer Science 0
Computer Science and Engineering 0
Elec Engineering 0
Environmental Engineering 0
Industrial Engineering 0
Material Science and Engineering 0
Mechanical Engineering 165
Nuclear Engineering 0
Other Engineering 0
<NA> 0
Other Engineering <NA>
Aerospace Engineering 28 0
Agricultural and Biological Engineering 0 0
Bio Engineering 0 0
Chemical Engineering 0 0
Civil Engineering 0 0
Computer Engineering 0 0
Computer Science 0 0
Computer Science and Engineering 0 0
Elec Engineering 0 0
Environmental Engineering 28 0
Industrial Engineering 29 0
Material Science and Engineering 0 0
Mechanical Engineering 0 0
Nuclear Engineering 44 0
Other Engineering 34 0
<NA> 13 0table(d$maj_fin2, useNA = "always")
Bio Engineering Chemical Engineering
129 149
Civil Engineering Computer Science and Engineering
100 95
Elec Engineering Material Science and Engineering
157 162
Mechanical Engineering Other Engineering
165 176
<NA>
0 d$maj_fin2 <- as.factor(d$maj_fin2)
chiout <- chisq.test(d$maj_fin2, d$class)
chiout
Pearson's Chi-squared test
data: d$maj_fin2 and d$class
X-squared = 83.451, df = 35, p-value = 7.788e-06chiout$observed d$class
d$maj_fin2 1 2 3 4 5 6
Bio Engineering 23 63 26 12 1 4
Chemical Engineering 27 66 35 12 2 7
Civil Engineering 19 59 13 2 6 1
Computer Science and Engineering 20 46 16 8 4 1
Elec Engineering 36 71 28 6 9 7
Material Science and Engineering 34 80 19 26 2 1
Mechanical Engineering 33 76 37 0 11 8
Other Engineering 39 93 24 12 4 4chiout$stdres d$class
d$maj_fin2 1 2 3
Bio Engineering -0.76632958 -0.01436740 0.85125079
Chemical Engineering -0.73720274 -1.20570018 2.07434150
Civil Engineering -0.36089782 2.11671464 -1.23429794
Computer Science and Engineering 0.16790022 -0.09690145 -0.16990277
Elec Engineering 0.85165847 -0.99218269 0.12750361
Material Science and Engineering 0.20451787 0.13366525 -2.08075393
Mechanical Engineering -0.13395615 -0.78846217 1.81086296
Other Engineering 0.63444081 1.13895670 -1.45937256
d$class
d$maj_fin2 4 5 6
Bio Engineering 1.15224878 -1.76503822 0.13499964
Chemical Engineering 0.60491992 -1.50868838 1.39063837
Civil Engineering -2.02036770 1.46933969 -1.19116080
Computer Science and Engineering 0.61803981 0.42915979 -1.12632892
Elec Engineering -1.63306619 1.69597826 1.24113665
Material Science and Engineering 4.97681996 -1.66485458 -1.87666165
Mechanical Engineering -3.77868178 2.45791909 1.59981798
Other Engineering -0.03774010 -0.92595575 -0.54928788sr_di <- data.frame(chiout$stdres)
colnames(sr_di) <- c("discipline","class","sr")d$race_fin2 <- as.factor(d$race_fin2)
d$race_fin2 <- relevel(d$race_fin2, ref = "White")
# d$race_fin3 <- d$race_fin2
# d$race_fin3[d$race_fin == "Bi/Multiracial"] <- "URM"
# d$race_fin3 <- droplevels(d$race_fin3)
chiout <- chisq.test(table(d$race_fin2, d$class))
chiout
Pearson's Chi-squared test
data: table(d$race_fin2, d$class)
X-squared = 23.989, df = 15, p-value = 0.06529chiout$observed
1 2 3 4 5 6
White 102 286 110 50 15 18
Asian 80 151 41 17 17 9
Bi/Multiracial 15 47 16 7 2 3
URM 25 58 22 4 4 1chiout$stdres
1 2 3 4 5
White -2.29585560 -0.03316766 1.62893368 2.07117563 -1.67706174
Asian 2.72986170 -0.56152680 -2.32024014 -1.38666511 2.23443253
Bi/Multiracial -0.86710142 0.58409099 0.15641425 0.26494185 -0.66806724
URM 0.49115918 0.36192342 0.63272688 -1.57388305 0.03348577
6
White 0.59356084
Asian 0.04946184
Bi/Multiracial 0.30818740
URM -1.32268471sr_re <- data.frame(chiout$stdres)
colnames(sr_re) <- c("raceethnicity","class","sr")table(d$gen_fin)
female gq male
398 15 692 d$gen_fin2 <- d$gen_fin
d$gen_fin2[d$gen_fin == "gq"] <- "fgq"
d$gen_fin2[d$gen_fin == "female"] <- "fgq"
d$gen_fin2 <- as.factor(d$gen_fin2)
table(d$gen_fin2)
fgq male
413 692 d$gen_fin2 <- as.factor(d$gen_fin2)
d$gen_fin2 <- relevel(d$gen_fin2, ref = "male")
chiout <- chisq.test(table(d$gen_fin2, d$class))
chiout
Pearson's Chi-squared test
data: table(d$gen_fin2, d$class)
X-squared = 5.5919, df = 5, p-value = 0.348chiout$observed
1 2 3 4 5 6
male 137 333 130 45 25 22
fgq 87 212 60 33 12 9chiout$stdres
1 2 3 4 5 6
male -0.5071185 -1.0326826 1.8149150 -0.9339242 0.6321686 0.9739402
fgq 0.5071185 1.0326826 -1.8149150 0.9339242 -0.6321686 -0.9739402sr_ge <- data.frame(chiout$stdres)
colnames(sr_ge) <- c("gender","class","sr")table(d$prog_fin)
0 1 2 3
324 373 286 51 d$prog_fin <- as.factor(d$prog_fin)
chiout <- chisq.test(table(d$prog_fin, d$class))
chiout
Pearson's Chi-squared test
data: table(d$prog_fin, d$class)
X-squared = 28.679, df = 15, p-value = 0.01769chiout$observed
1 2 3 4 5 6
0 62 173 55 14 11 9
1 89 172 63 23 17 9
2 53 139 48 32 5 9
3 5 24 11 4 5 2chiout$stdres
1 2 3 4 5 6
0 -0.58253899 1.85338691 -0.08228287 -2.32267728 -0.32325106 -0.03534427
1 2.19415971 -1.45775355 -0.14615610 -0.84279759 1.13311296 -0.57317542
2 -0.83244866 -0.21007906 -0.17671955 3.20493329 -2.03623281 0.41212118
3 -1.89841666 -0.30338656 0.86543503 0.22393674 2.38587841 0.49548634sr_pr <- data.frame(chiout$stdres)
colnames(sr_pr) <- c("degreeprog","class","sr")table(d$intl_fin, useNA = "always")
domestic international <NA>
718 405 10 d$intl_fin <- as.factor(d$intl_fin)
chiout <- chisq.test(table(d$intl_fin, d$class))
chiout
Pearson's Chi-squared test
data: table(d$intl_fin, d$class)
X-squared = 19.381, df = 5, p-value = 0.001632chiout$observed
1 2 3 4 5 6
domestic 131 359 124 63 18 23
international 98 193 69 15 20 10chiout$stdres
1 2 3 4 5
domestic -2.37730893 0.75502373 0.09945113 3.20951548 -2.16377895
international 2.37730893 -0.75502373 -0.09945113 -3.20951548 2.16377895
6
domestic 0.69956457
international -0.69956457sr_in <- data.frame(chiout$stdres)
colnames(sr_in) <- c("intl_status","class","sr")
chiout <- chisq.test(table(d$intl_fin, d$maj_fin2))
chiout
Pearson's Chi-squared test
data: table(d$intl_fin, d$maj_fin2)
X-squared = 122.59, df = 7, p-value < 2.2e-16chiout$observed
Bio Engineering Chemical Engineering Civil Engineering
domestic 100 111 46
international 28 37 54
Computer Science and Engineering Elec Engineering
domestic 47 66
international 48 91
Material Science and Engineering Mechanical Engineering
domestic 140 87
international 21 76
Other Engineering
domestic 121
international 50chiout$stdres
Bio Engineering Chemical Engineering Civil Engineering
domestic 3.551637 3.008327 -3.913496
international -3.551637 -3.008327 3.913496
Computer Science and Engineering Elec Engineering
domestic -3.068171 -6.160828
international 3.068171 6.160828
Material Science and Engineering Mechanical Engineering
domestic 6.572379 -3.037176
international -6.572379 3.037176
Other Engineering
domestic 2.018462
international -2.018462chiout <- chisq.test(table(d$intl_fin, d$race_fin2))
chiout
Pearson's Chi-squared test
data: table(d$intl_fin, d$race_fin2)
X-squared = 473.22, df = 3, p-value < 2.2e-16chiout$observed
White Asian Bi/Multiracial URM
domestic 528 63 65 48
international 53 252 25 66chiout$stdres
White Asian Bi/Multiracial URM
domestic 19.649575 -19.258756 1.695918 -5.144093
international -19.649575 19.258756 -1.695918 5.144093d <- subset(d, prog_fin != 3)
d$maj_fin2[d$maj_fin2 == "Other Engineering"] <- NA
d$maj_fin2 <- relevel(d$maj_fin2, ref = "Mechanical Engineering")
d$race_fin2 <- relevel(d$race_fin2, ref = "White")
d$intl_fin <- relevel(d$intl_fin, ref = "domestic")
d$gen_fin2 <- relevel(d$gen_fin2, ref = "male")
d$prog_fin <- relevel(d$prog_fin, ref = "0")
d$class_fin[d$class == "1"] <- "disorganized"
d$class_fin[d$class == "2"] <- "moderate"
d$class_fin[d$class == "3"] <- "strong"
d$class_fin[d$class == "4"] <- "mle"
d$class_fin[d$class == "5"] <- "mls"
d$class_fin[d$class == "6"] <- "max"
d$class_fin <- as.factor(d$class_fin)
d$class_fin <- relevel(d$class_fin, ref = "moderate")
regout1 <- lm(uncer ~ maj_fin2 + gen_fin2 + race_fin2 + intl_fin + prog_fin, data = d, na.action = na.exclude)
plot_model(regout1, type="diag")[[1]]
[[2]]
[[3]]
[[4]]
qqnorm(d$uncer)
qqline(d$uncer)
plot(regout1, 4)
plot(regout1, 5)
plot(regout1, 3)
summary(regout1)
Call:
lm(formula = uncer ~ maj_fin2 + gen_fin2 + race_fin2 + intl_fin +
prog_fin, data = d, na.action = na.exclude)
Residuals:
Min 1Q Median 3Q Max
-1.56448 -0.60809 0.03452 0.54057 2.66023
Coefficients:
Estimate Std. Error t value Pr(>|t|)
(Intercept) 2.144976 0.083162 25.793 < 2e-16
maj_fin2Bio Engineering 0.048421 0.096266 0.503 0.615104
maj_fin2Chemical Engineering 0.089741 0.091700 0.979 0.328054
maj_fin2Civil Engineering 0.014319 0.101946 0.140 0.888337
maj_fin2Computer Science and Engineering 0.136315 0.102758 1.327 0.185029
maj_fin2Elec Engineering 0.025970 0.089742 0.289 0.772365
maj_fin2Material Science and Engineering 0.163696 0.094330 1.735 0.083057
gen_fin2fgq -0.003499 0.054631 -0.064 0.948947
race_fin2Asian 0.040529 0.075925 0.534 0.593623
race_fin2Bi/Multiracial 0.268740 0.102873 2.612 0.009159
race_fin2URM 0.059797 0.092915 0.644 0.520038
intl_fininternational -0.245993 0.072755 -3.381 0.000757
prog_fin1 0.154261 0.061821 2.495 0.012784
prog_fin2 -0.050636 0.067105 -0.755 0.450723
(Intercept) ***
maj_fin2Bio Engineering
maj_fin2Chemical Engineering
maj_fin2Civil Engineering
maj_fin2Computer Science and Engineering
maj_fin2Elec Engineering
maj_fin2Material Science and Engineering .
gen_fin2fgq
race_fin2Asian
race_fin2Bi/Multiracial **
race_fin2URM
intl_fininternational ***
prog_fin1 *
prog_fin2
---
Signif. codes: 0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1
Residual standard error: 0.7362 on 809 degrees of freedom
(160 observations deleted due to missingness)
Multiple R-squared: 0.05074, Adjusted R-squared: 0.03548
F-statistic: 3.326 on 13 and 809 DF, p-value: 5.613e-05tab_model(regout1, show.ci = F, show.se = T, show.stat = T)| uncer | ||||
| Predictors | Estimates | std. Error | Statistic | p |
| (Intercept) | 2.14 | 0.08 | 25.79 | <0.001 |
| maj fin2 [Bio Engineering] |
0.05 | 0.10 | 0.50 | 0.615 |
| maj fin2 [Chemical Engineering] |
0.09 | 0.09 | 0.98 | 0.328 |
| maj fin2 [Civil Engineering] |
0.01 | 0.10 | 0.14 | 0.888 |
| maj fin2 [Computer Science and Engineering] |
0.14 | 0.10 | 1.33 | 0.185 |
| maj fin2 [Elec Engineering] |
0.03 | 0.09 | 0.29 | 0.772 |
| maj fin2 [Material Science and Engineering] |
0.16 | 0.09 | 1.74 | 0.083 |
| gen fin2 [fgq] | -0.00 | 0.05 | -0.06 | 0.949 |
| race fin2 [Asian] | 0.04 | 0.08 | 0.53 | 0.594 |
| race fin2 [Bi/Multiracial] |
0.27 | 0.10 | 2.61 | 0.009 |
| race fin2 [URM] | 0.06 | 0.09 | 0.64 | 0.520 |
| intl fin [international] | -0.25 | 0.07 | -3.38 | 0.001 |
| prog fin [1] | 0.15 | 0.06 | 2.50 | 0.013 |
| prog fin [2] | -0.05 | 0.07 | -0.75 | 0.451 |
| Observations | 823 | |||
| R2 / R2 adjusted | 0.051 / 0.035 | |||
plot_model(regout1)
regout2 <- lm(uncer ~ class_fin + maj_fin2 + gen_fin2 + race_fin2 + intl_fin + prog_fin, data = d, na.action = na.exclude)
plot_model(regout2, type="diag")[[1]]
[[2]]
[[3]]
[[4]]
plot(regout2, 4)
plot(regout2, 5)
plot(regout2, 3)
summary(regout2)
Call:
lm(formula = uncer ~ class_fin + maj_fin2 + gen_fin2 + race_fin2 +
intl_fin + prog_fin, data = d, na.action = na.exclude)
Residuals:
Min 1Q Median 3Q Max
-1.67864 -0.49771 0.02514 0.52788 2.33980
Coefficients:
Estimate Std. Error t value Pr(>|t|)
(Intercept) 2.219025 0.083063 26.715 < 2e-16
class_findisorganized 0.309858 0.064940 4.771 2.17e-06
class_finmax -0.331029 0.158673 -2.086 0.037272
class_finmle -0.352733 0.101492 -3.475 0.000537
class_finmls -0.203343 0.138255 -1.471 0.141743
class_finstrong -0.367676 0.068329 -5.381 9.73e-08
maj_fin2Bio Engineering 0.065662 0.092259 0.712 0.476853
maj_fin2Chemical Engineering 0.108568 0.087724 1.238 0.216223
maj_fin2Civil Engineering -0.010904 0.097353 -0.112 0.910846
maj_fin2Computer Science and Engineering 0.123246 0.098208 1.255 0.209864
maj_fin2Elec Engineering 0.005229 0.085579 0.061 0.951295
maj_fin2Material Science and Engineering 0.137365 0.091699 1.498 0.134524
gen_fin2fgq -0.014320 0.052132 -0.275 0.783631
race_fin2Asian -0.013215 0.072759 -0.182 0.855927
race_fin2Bi/Multiracial 0.280523 0.098037 2.861 0.004327
race_fin2URM 0.035375 0.088487 0.400 0.689426
intl_fininternational -0.253635 0.069372 -3.656 0.000273
prog_fin1 0.144530 0.059044 2.448 0.014585
prog_fin2 -0.036032 0.064008 -0.563 0.573639
(Intercept) ***
class_findisorganized ***
class_finmax *
class_finmle ***
class_finmls
class_finstrong ***
maj_fin2Bio Engineering
maj_fin2Chemical Engineering
maj_fin2Civil Engineering
maj_fin2Computer Science and Engineering
maj_fin2Elec Engineering
maj_fin2Material Science and Engineering
gen_fin2fgq
race_fin2Asian
race_fin2Bi/Multiracial **
race_fin2URM
intl_fininternational ***
prog_fin1 *
prog_fin2
---
Signif. codes: 0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1
Residual standard error: 0.7005 on 804 degrees of freedom
(160 observations deleted due to missingness)
Multiple R-squared: 0.1458, Adjusted R-squared: 0.1267
F-statistic: 7.623 on 18 and 804 DF, p-value: < 2.2e-16tab_model(regout2, show.ci = F, show.se = T, show.stat = T)| uncer | ||||
| Predictors | Estimates | std. Error | Statistic | p |
| (Intercept) | 2.22 | 0.08 | 26.71 | <0.001 |
| class fin [disorganized] | 0.31 | 0.06 | 4.77 | <0.001 |
| class fin [max] | -0.33 | 0.16 | -2.09 | 0.037 |
| class fin [mle] | -0.35 | 0.10 | -3.48 | 0.001 |
| class fin [mls] | -0.20 | 0.14 | -1.47 | 0.142 |
| class fin [strong] | -0.37 | 0.07 | -5.38 | <0.001 |
| maj fin2 [Bio Engineering] |
0.07 | 0.09 | 0.71 | 0.477 |
| maj fin2 [Chemical Engineering] |
0.11 | 0.09 | 1.24 | 0.216 |
| maj fin2 [Civil Engineering] |
-0.01 | 0.10 | -0.11 | 0.911 |
| maj fin2 [Computer Science and Engineering] |
0.12 | 0.10 | 1.25 | 0.210 |
| maj fin2 [Elec Engineering] |
0.01 | 0.09 | 0.06 | 0.951 |
| maj fin2 [Material Science and Engineering] |
0.14 | 0.09 | 1.50 | 0.135 |
| gen fin2 [fgq] | -0.01 | 0.05 | -0.27 | 0.784 |
| race fin2 [Asian] | -0.01 | 0.07 | -0.18 | 0.856 |
| race fin2 [Bi/Multiracial] |
0.28 | 0.10 | 2.86 | 0.004 |
| race fin2 [URM] | 0.04 | 0.09 | 0.40 | 0.689 |
| intl fin [international] | -0.25 | 0.07 | -3.66 | <0.001 |
| prog fin [1] | 0.14 | 0.06 | 2.45 | 0.015 |
| prog fin [2] | -0.04 | 0.06 | -0.56 | 0.574 |
| Observations | 823 | |||
| R2 / R2 adjusted | 0.146 / 0.127 | |||
plot_model(regout2)
anova(regout1,regout2)Analysis of Variance Table
Model 1: uncer ~ maj_fin2 + gen_fin2 + race_fin2 + intl_fin + prog_fin
Model 2: uncer ~ class_fin + maj_fin2 + gen_fin2 + race_fin2 + intl_fin +
prog_fin
Res.Df RSS Df Sum of Sq F Pr(>F)
1 809 438.48
2 804 394.58 5 43.907 17.893 < 2.2e-16 ***
---
Signif. codes: 0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1rm(list = ls())
df <- read.csv(file="auxvar.csv", header=T)
lpa_df <- subset(df, select=c(R_IDs, s_std, r_std, e_std, class_fin))
un_df <- read.csv(file="un_df2.csv", header=T)
d <- df
table(d$maj_fin, useNA = "always")
Aerospace Engineering Agricultural and Biological Engineering
28 27
Bio Engineering Chemical Engineering
102 149
Civil Engineering Computer Engineering
100 19
Computer Science Computer Science and Engineering
55 21
Elec Engineering Environmental Engineering
157 28
Industrial Engineering Material Science and Engineering
29 162
Mechanical Engineering Nuclear Engineering
165 44
Other Engineering <NA>
34 13 d$maj_fin2 <- d$maj_fin
d$maj_fin2[d$maj_fin == "Computer Engineering"] <- "Computer Science and Engineering"
d$maj_fin2[d$maj_fin == "Computer Science"] <- "Computer Science and Engineering"
d$maj_fin2[d$maj_fin == "Agricultural and Biological Engineering"] <- "Bio Engineering"
d$maj_fin2[d$maj_fin == "Aerospace Engineering"] <- "Other Engineering"
d$maj_fin2[d$maj_fin == "Environmental Engineering"] <- "Other Engineering"
d$maj_fin2[d$maj_fin == "Industrial Engineering"] <- "Other Engineering"
d$maj_fin2[d$maj_fin == "Nuclear Engineering"] <- "Other Engineering"
d$maj_fin2[d$maj_fin == "Elec Engineering"] <- "Electrical Engineering"
d$maj_fin2[is.na(d$maj_fin)] <- "Other Engineering"
table(d$maj_fin, d$maj_fin2, useNA = "always")
Bio Engineering Chemical Engineering
Aerospace Engineering 0 0
Agricultural and Biological Engineering 27 0
Bio Engineering 102 0
Chemical Engineering 0 149
Civil Engineering 0 0
Computer Engineering 0 0
Computer Science 0 0
Computer Science and Engineering 0 0
Elec Engineering 0 0
Environmental Engineering 0 0
Industrial Engineering 0 0
Material Science and Engineering 0 0
Mechanical Engineering 0 0
Nuclear Engineering 0 0
Other Engineering 0 0
<NA> 0 0
Civil Engineering
Aerospace Engineering 0
Agricultural and Biological Engineering 0
Bio Engineering 0
Chemical Engineering 0
Civil Engineering 100
Computer Engineering 0
Computer Science 0
Computer Science and Engineering 0
Elec Engineering 0
Environmental Engineering 0
Industrial Engineering 0
Material Science and Engineering 0
Mechanical Engineering 0
Nuclear Engineering 0
Other Engineering 0
<NA> 0
Computer Science and Engineering
Aerospace Engineering 0
Agricultural and Biological Engineering 0
Bio Engineering 0
Chemical Engineering 0
Civil Engineering 0
Computer Engineering 19
Computer Science 55
Computer Science and Engineering 21
Elec Engineering 0
Environmental Engineering 0
Industrial Engineering 0
Material Science and Engineering 0
Mechanical Engineering 0
Nuclear Engineering 0
Other Engineering 0
<NA> 0
Electrical Engineering
Aerospace Engineering 0
Agricultural and Biological Engineering 0
Bio Engineering 0
Chemical Engineering 0
Civil Engineering 0
Computer Engineering 0
Computer Science 0
Computer Science and Engineering 0
Elec Engineering 157
Environmental Engineering 0
Industrial Engineering 0
Material Science and Engineering 0
Mechanical Engineering 0
Nuclear Engineering 0
Other Engineering 0
<NA> 0
Material Science and Engineering
Aerospace Engineering 0
Agricultural and Biological Engineering 0
Bio Engineering 0
Chemical Engineering 0
Civil Engineering 0
Computer Engineering 0
Computer Science 0
Computer Science and Engineering 0
Elec Engineering 0
Environmental Engineering 0
Industrial Engineering 0
Material Science and Engineering 162
Mechanical Engineering 0
Nuclear Engineering 0
Other Engineering 0
<NA> 0
Mechanical Engineering
Aerospace Engineering 0
Agricultural and Biological Engineering 0
Bio Engineering 0
Chemical Engineering 0
Civil Engineering 0
Computer Engineering 0
Computer Science 0
Computer Science and Engineering 0
Elec Engineering 0
Environmental Engineering 0
Industrial Engineering 0
Material Science and Engineering 0
Mechanical Engineering 165
Nuclear Engineering 0
Other Engineering 0
<NA> 0
Other Engineering <NA>
Aerospace Engineering 28 0
Agricultural and Biological Engineering 0 0
Bio Engineering 0 0
Chemical Engineering 0 0
Civil Engineering 0 0
Computer Engineering 0 0
Computer Science 0 0
Computer Science and Engineering 0 0
Elec Engineering 0 0
Environmental Engineering 28 0
Industrial Engineering 29 0
Material Science and Engineering 0 0
Mechanical Engineering 0 0
Nuclear Engineering 44 0
Other Engineering 34 0
<NA> 13 0table(d$maj_fin2, useNA = "always")
Bio Engineering Chemical Engineering
129 149
Civil Engineering Computer Science and Engineering
100 95
Electrical Engineering Material Science and Engineering
157 162
Mechanical Engineering Other Engineering
165 176
<NA>
0 d$maj_fin2 <- as.factor(d$maj_fin2)
d$class[d$class == 1] <- "Dispersed"
d$class[d$class == 2] <- "Moderate"
d$class[d$class == 3] <- "Strong"
d$class[d$class == 4] <- "Primarily Scientist"
d$class[d$class == 5] <- "Primarily Engineer"
d$class[d$class == 6] <- "Maximum"
d$class <- factor(d$class, levels = c("Dispersed",
"Moderate",
"Strong",
"Primarily Scientist",
"Primarily Engineer",
"Maximum"))
table(d$class)
Dispersed Moderate Strong Primarily Scientist
231 554 198 78
Primarily Engineer Maximum
39 33 chiout <- chisq.test(d$maj_fin2, d$class)
# standardized residuals
resid_mat <- chiout$stdres
df_resid <- as.data.frame(as.table(resid_mat))
names(df_resid) <- c("maj_fin2", "class", "std_resid")
df_resid_filt <- df_resid %>%
group_by(maj_fin2) %>%
filter(any(abs(std_resid) > 2)) %>%
ungroup()
plotout <- ggplot(df_resid_filt, aes(x = class, y = std_resid, fill = class)) +
geom_col(width = 0.7) +
geom_hline(
yintercept = c(-2, 2),
linetype = "dashed",
linewidth = 0.3
) +
facet_wrap(
~ maj_fin2,
ncol = 4,
labeller = label_wrap_gen(width = 20)
) +
labs(
x = NULL,
y = "Standardized Residuals",
fill = "Class"
) +
theme_bw() +
scale_fill_viridis_d(option = "D", end = 0.9) +
theme(
axis.text.x = element_blank(),
axis.ticks.x = element_blank(),
legend.position = "bottom",
legend.direction = "horizontal",
strip.text = element_text(size = 8, lineheight = 1),
legend.text = element_text(size = 8),
legend.title = element_text(size = 9),
legend.key.width = unit(0.7, "lines"),
legend.key.height = unit(0.7, "lines"),
axis.title.y = element_text(size = 9),
axis.text.y = element_text(size = 7),
) +
guides(fill = guide_legend(nrow = 1, override.aes = list(size = 2)))
disc <- plotout +
theme(
axis.title.y = element_text(size = 11),
axis.text.y = element_text(size = 9),
legend.title = element_text(size = 11, lineheight = 0.9),
legend.text = element_text(size = 10, lineheight = 0.9),
strip.text = element_text(size = 10, lineheight = 0.9),
plot.background = element_rect(fill = "transparent", color = NA),
legend.background = element_rect(fill = "transparent", color = NA)
)
disc
# ggsave("pres/disc.png", plot = disc, bg = "transparent",
# width = 8, height = 2.75, dpi = 300)d <- na.omit(subset(df, select=c(intl_fin, class)))
table(d$intl_fin, useNA = "always")
domestic international <NA>
718 405 0 d$intl_fin <- as.factor(d$intl_fin)
d$class[d$class == 1] <- "Dispersed"
d$class[d$class == 2] <- "Moderate"
d$class[d$class == 3] <- "Strong"
d$class[d$class == 4] <- "Primarily Scientist"
d$class[d$class == 5] <- "Primarily Engineer"
d$class[d$class == 6] <- "Maximum"
d$class <- factor(d$class, levels = c("Dispersed",
"Moderate",
"Strong",
"Primarily Scientist",
"Primarily Engineer",
"Maximum"))
table(d$class)
Dispersed Moderate Strong Primarily Scientist
229 552 193 78
Primarily Engineer Maximum
38 33 chiout <- chisq.test(d$intl_fin, d$class)
table(df$maj_fin, df$intl_fin)
domestic international
Aerospace Engineering 22 6
Agricultural and Biological Engineering 15 12
Bio Engineering 85 16
Chemical Engineering 111 37
Civil Engineering 46 54
Computer Engineering 6 13
Computer Science 28 27
Computer Science and Engineering 13 8
Elec Engineering 66 91
Environmental Engineering 19 9
Industrial Engineering 17 12
Material Science and Engineering 140 21
Mechanical Engineering 87 76
Nuclear Engineering 38 6
Other Engineering 20 14chiout2 <- chisq.test(df$maj_fin, df$intl_fin)
# chiout2
# chiout2$stdres
# standardized residuals
resid_mat <- chiout$stdres
df_resid <- as.data.frame(as.table(resid_mat))
names(df_resid) <- c("intl_fin", "class", "std_resid")
# df_resid_filt <- df_resid %>%
# group_by(maj_fin2) %>%
# filter(any(abs(std_resid) > 2)) %>%
# ungroup()
plotout <- ggplot(df_resid, aes(x = class, y = std_resid, fill = class)) +
geom_col(width = 0.7) +
geom_hline(
yintercept = c(-2, 2),
linetype = "dashed",
linewidth = 0.3
) +
facet_wrap(
~ intl_fin,
ncol = 4,
labeller = label_wrap_gen(width = 20)
) +
labs(
x = NULL,
y = "Standardized Residuals",
fill = "Class"
) +
theme_bw() +
scale_fill_viridis_d(option = "D", end = 0.9) +
theme(
axis.text.x = element_blank(),
axis.ticks.x = element_blank(),
legend.position = "bottom",
legend.direction = "horizontal",
strip.text = element_text(size = 8, lineheight = 1),
legend.text = element_text(size = 8),
legend.title = element_text(size = 9),
legend.key.width = unit(0.7, "lines"),
legend.key.height = unit(0.7, "lines"),
axis.title.y = element_text(size = 9),
axis.text.y = element_text(size = 7),
) +
guides(fill = guide_legend(nrow = 1, override.aes = list(size = 2)))
intl <- plotout +
theme(
axis.title.y = element_text(size = 11),
axis.text.y = element_text(size = 9),
legend.title = element_text(size = 11, lineheight = 0.9),
legend.text = element_text(size = 10, lineheight = 0.9),
strip.text = element_text(size = 10, lineheight = 0.9),
plot.background = element_rect(fill = "transparent", color = NA),
legend.background = element_rect(fill = "transparent", color = NA)
)
intl
# ggsave("pres/intl.png", plot = intl, bg = "transparent",
# width = 8, height = 2.75, dpi = 300)d <- df
table(d$prog_fin)
0 1 2 3
324 373 286 51 d$prog_fin[d$prog_fin == 0] <- "No Milestones"
d$prog_fin[d$prog_fin == 1] <- "First Milestone (Comprehensive Exam)"
d$prog_fin[d$prog_fin == 2] <- "Second Milestone (Dissertation Proposal)"
d$prog_fin[d$prog_fin == 3] <- "All Milestones (Dissertation Defense)"
d$prog_fin <- factor(d$prog_fin, levels = c("No Milestones",
"First Milestone (Comprehensive Exam)",
"Second Milestone (Dissertation Proposal)",
"All Milestones (Dissertation Defense)"))
d$class[d$class == 1] <- "Dispersed"
d$class[d$class == 2] <- "Moderate"
d$class[d$class == 3] <- "Strong"
d$class[d$class == 4] <- "Primarily Scientist"
d$class[d$class == 5] <- "Primarily Engineer"
d$class[d$class == 6] <- "Maximum"
d$class <- factor(d$class, levels = c("Dispersed",
"Moderate",
"Strong",
"Primarily Scientist",
"Primarily Engineer",
"Maximum"))
table(d$class)
Dispersed Moderate Strong Primarily Scientist
231 554 198 78
Primarily Engineer Maximum
39 33 chiout <- chisq.test(table(d$prog_fin, d$class))
# standardized residuals
resid_mat <- chiout$stdres
df_resid <- as.data.frame(as.table(resid_mat))
names(df_resid) <- c("pred", "class", "std_resid")
plotout <- ggplot(df_resid, aes(x = class, y = std_resid, fill = class)) +
geom_col(width = 0.7) +
geom_hline(
yintercept = c(-2, 2),
linetype = "dashed",
linewidth = 0.3
) +
facet_wrap(
~ pred,
ncol = 4,
labeller = label_wrap_gen(width = 25)
) +
labs(
x = NULL,
y = "Standardized Residuals",
fill = "Class"
) +
theme_bw() +
scale_fill_viridis_d(option = "D", end = 0.9) +
theme(
axis.text.x = element_blank(),
axis.ticks.x = element_blank(),
legend.position = "bottom",
legend.direction = "horizontal",
strip.text = element_text(size = 8, lineheight = 1),
legend.text = element_text(size = 8),
legend.title = element_text(size = 9),
legend.key.width = unit(0.7, "lines"),
legend.key.height = unit(0.7, "lines"),
axis.title.y = element_text(size = 9),
axis.text.y = element_text(size = 7),
) +
guides(fill = guide_legend(nrow = 1, override.aes = list(size = 2)))
mile <- plotout +
theme(
axis.title.y = element_text(size = 11),
axis.text.y = element_text(size = 9),
legend.title = element_text(size = 11, lineheight = 0.9),
legend.text = element_text(size = 10, lineheight = 0.9),
strip.text = element_text(size = 10, lineheight = 0.9),
plot.background = element_rect(fill = "transparent", color = NA),
legend.background = element_rect(fill = "transparent", color = NA)
)
mile
# ggsave("pres/mile.png", plot = mile, bg = "transparent",
# width = 8, height = 2.75, dpi = 300)un_df$class_fin[un_df$class_fin == "disorganized"] <- "Dispersed"
un_df$class_fin[un_df$class_fin == "strong"] <- "Strong"
un_df$class_fin[un_df$class_fin == "moderate"] <- "Moderate"
un_df$class_fin[un_df$class_fin == "mle"] <- "Primarily Scientist"
un_df$class_fin[un_df$class_fin == "mls"] <- "Primarily Engineer"
un_df$class_fin[un_df$class_fin == "max"] <- "Maximum"
un_df$class_fin <- factor(un_df$class_fin, levels = c("Dispersed",
"Moderate",
"Strong",
"Primarily Scientist",
"Primarily Engineer",
"Maximum"))
un_df2 <- na.omit(un_df)
un_df3 <- merge(un_df2, subset(df, select=c(R_IDs,prog_fin)), by = "R_IDs")
un_df4 <- subset(un_df3, prog_fin != 3)
uncer <- ggplot(un_df4, aes(x = class_fin, y = uncer, color = factor(class_fin))) +
geom_boxplot() +
geom_jitter(width = .20, height = .20, alpha = .20) +
labs(
y = "Degree Uncertainty",
x = NULL,
color = "GEI Profile"
) +
theme_bw(base_size = 12) +
scale_color_viridis_d(option = "D", end = 0.9) +
theme(
# x-axis labels back, angled
axis.text.x = element_text(
angle = 45, hjust = 1, vjust = 1, size = 8
),
axis.ticks.x = element_blank(),
# legend cleanup
legend.position = "none",
legend.direction = "horizontal",
legend.box = "horizontal",
legend.text = element_text(size = 8),
legend.title = element_text(size = 9),
legend.key.width = unit(0.7, "lines"),
legend.key.height = unit(0.7, "lines"),
axis.title.y = element_text(size = 9),
axis.text.y = element_text(size = 7)
) +
guides(
color = guide_legend(
nrow = 1,
override.aes = list(alpha = 1, size = 2)
)
)
uncer <- uncer +
theme(
axis.title.y = element_text(size = 12),
axis.text.x = element_text(size = 11),
legend.title = element_text(size = 12, lineheight = 0.9),
legend.text = element_text(size = 11, lineheight = 0.9),
strip.text = element_text(size = 11, lineheight = 0.9),
plot.background = element_rect(fill = "transparent", color = NA),
legend.background = element_rect(fill = "transparent", color = NA)
)
uncer
# ggsave("pres/uncer.png", plot = uncer, bg = "transparent",
# width = 6, height = 6, dpi = 300)
# describeBy(un_df2, group = "class_fin")