Dissertation Chapter: Aspirations to Post-Graduate Education of Hispanics & Social Capital
Paulina Cano McC
January/February 2020
This data set uses the ELS public database. Filters out any non-Hispanic student.
library(ggplot2)## Warning: package 'ggplot2' was built under R version 3.6.2library(haven)## Warning: package 'haven' was built under R version 3.6.2library(stargazer)##
## Please cite as:## Hlavac, Marek (2018). stargazer: Well-Formatted Regression and Summary Statistics Tables.## R package version 5.2.2. https://CRAN.R-project.org/package=stargazerlibrary(survey)## Loading required package: grid## Loading required package: Matrix## Loading required package: survival## Warning: package 'survival' was built under R version 3.6.2##
## Attaching package: 'survey'## The following object is masked from 'package:graphics':
##
## dotchartlibrary(questionr)
load("C:/Users/PCMcC/Documents/DISSERTATION/CHAPTER 1 ASPIRATIONS & SOCIAL CAPITAL/Data/els_02_12_byf3pststu_v1_0.rdata")
data1<-as.data.frame(els_02_12_byf3pststu_v1_0)
rm(els_02_12_byf3pststu_v1_0)
options(scipen = 999)#Rename columns
library(dplyr)## Warning: package 'dplyr' was built under R version 3.6.2##
## Attaching package: 'dplyr'## The following objects are masked from 'package:stats':
##
## filter, lag## The following objects are masked from 'package:base':
##
## intersect, setdiff, setequal, uniondata2<-data1 %>% select(STU_ID,psu,bystuwt,F3QWT,F3BYPNLWT,STRAT_ID,BYS14, byrace,bypared,byparacr, bystexp, BYP85, F3ATTAINMENT, F3BYEDEXPFF, F3A17, F3FURTHERED, F3STEXP,F3HIIMPACT,F3A14A,F3A14B,F3A14C,F3A14D,F3A14E,F3A14F,F3D39,F3D40,F3D42)
#bystuwt represents student weight at baseline
#F3QWT represents weight at third round of survey
#psu representents primary sampling unit
#STRAT_ID representes the sampling strata
# F3BYPNLWT this is a panel weight used when your analysis spans more than one round of data collection. Need to clarify if this is more appropriate considering some of my control variables are coming from baseline survey like parent's income, parent's race.
#Rename Columns
names(data2)[names(data2) == "STU_ID"]<-"studentID"
names(data2)[names(data2)=="BYS14"]<-"sex"
names(data2)[names(data2)=="byrace"]<-"race"
names(data2)[names(data2)=="bypared"]<-"parentsed"
names(data2)[names(data2)=="bystexp"]<-"firstExpectEdu"
names(data2)[names(data2)=="BYP85"]<-"familyInc"
names(data2)[names(data2)=="F3ATTAINMENT"]<-"eduAttain"
names(data2)[names(data2)=="F3BYEDEXPFF"]<-"expectEduStatus"
#names(data2)[names(data2)=="F3A17"]<-"expectEduLevel"
names(data2)[names(data2)=="F3FURTHERED"]<-"expectEduAttained"
names(data2)[names(data2)=="F3STEXP"]<-"thirdExpectEdu"
names(data2)[names(data2)=="F3HIIMPACT"]<-"ImpactActivities"
names(data2)[names(data2)=="F3A14A"]<-"internship"
names(data2)[names(data2)=="F3A14B"]<-"research"
names(data2)[names(data2)=="F3A14C"]<-"abroad"
names(data2)[names(data2)=="F3A14D"]<-"community"
names(data2)[names(data2)=="F3A14E"]<-"seniorExperience"
names(data2)[names(data2)=="F3A14F"]<-"mentoring"
names(data2)[names(data2)=="F3D39"]<-"voted"
names(data2)[names(data2)=="F3D40"]<-"unpaidWork"
names(data2)[names(data2)=="F3D42"]<-"volunteeringFreq"
names(data2)[names(data2)=="STRAT_ID"]<-"strataID"
names(data2)[names(data2)=="F3QWT"]<-"studentWeight"
names(data2)[names(data2)=="F3BYPNLWT"]<-"panelStudentWeight"
#Filter data set to keep only hispanic students (4=Hispanic, no race specified, 5=Hispanic, race specified)
data2<-filter(data2, race %in% c("4", "5"))
#Recode Variables
library(car)## Warning: package 'car' was built under R version 3.6.2## Loading required package: carData## Warning: package 'carData' was built under R version 3.6.1##
## Attaching package: 'car'## The following object is masked from 'package:dplyr':
##
## recode#sex
data2$Female <-Recode(data2$sex, recodes = "1=0; 2=1; else =NA")
data2$sex2 <- Recode(data2$sex, recodes = "1='Male'; 2='Female'; else =NA", as.factor=T)
#parents education
data2$parentsed2 <-Recode(data2$parentsed, recodes = "1='LessthanHS';2='HighSchool'; 3='LessthanAssociate'; 4='Associate'; 5='LessthanBach';6='Bachelors'; 7:8='Graduate'; else=NA ", as.factor=T)
data2$parentsBachelorOrMore<-Recode(data2$parentsed, recodes = "1:5=0; 6:8=1; else=NA ", as.factor=T)
#first survey academic expectation
data2$firstExpectEdu2 <-Recode(data2$firstExpectEdu, recodes = "1='LessthanHS';2='HighSchool'; 3='Associate'; 4='SomeCollege'; 5='Bachelors'; 6:7='Graduate'; else=NA ", as.factor=T)
#family income during first survey
data2$familyInc2 <-Recode(data2$familyInc, recodes = "1:6='Lessthan20k';7:9:='20kto50k'; 10:11='50kto100k'; 12:13='Morethan100k'; else=NA ", as.factor=T)
#education attainment during F3 survey
data2$eduAttain2 <-Recode(data2$eduAttain, recodes = "1='LessthanHS';2='HighSchool'; 3='SomeCollege'; 4='Certificate'; 5='Associates'; 6='Bachelors'; 7='PostBacCert'; 8:10='Graduate';else= NA", as.factor=T)
#Fulfillment of BY educational expectations as of F3
data2$expectEduStatus2 <-Recode(data2$expectEduStatus, recodes = "1:2='MetExpectations';3='DidNotMeetExpectations'; 4:5='NotApplicable';else='NA' ", as.factor=T)
#Highest level of education respondent expects to complete - composite (DEPENDENT VARIABLE)
data2$thirdExpectEdu2 <-Recode(data2$thirdExpectEdu, recodes = "1='LessthanHS';2:3='HighSchool'; 4='Certificate'; 5='Associates'; 6='Bachelors'; 7:10='Graduate'; 11='dontknow';else= NA", as.factor=T)
data2$thirdExpectEdu2<-relevel(data2$thirdExpectEdu2, ref = "HighSchool")
data2$GradExpect <-Recode(data2$thirdExpectEdu, recodes = "1:6=0; 7:10=1; 11=0; else=NA ")
#High-impact PS activities: Internship/co-op/field experience/student teaching/clinical assignment
data2$internship2 <-Recode(data2$internship, recodes = "1=1; 0=0; else=NA ")
# High-impact PS activities: Research project with faculty member outside course/program requirements
data2$research2 <-Recode(data2$research, recodes = "1=1; 0=0; else=NA ")
#High-impact PS activities: Study abroad
data2$abroad2 <-Recode(data2$abroad, recodes = "1=1; 0=0; else=NA ")
#High-impact PS activities: Community-based project
data2$community2<-Recode(data2$community, recodes = "1=1; 0=0; else= NA ")
#High-impact PS activities: Culminating senior experience
data2$seniorExperience2<-Recode(data2$seniorExperience, recodes = "1=1; 0=0; else= NA ")
#High-impact PS activities: Mentoring
data2$mentoring2<-Recode(data2$mentoring, recodes = "1=1; 0=0; else=NA ")
#Sum of High-Impact Activities
data2$ImpactActivities2<-ifelse(data2$ImpactActivities<0, NA,data2$ImpactActivities )
#Whether voted in any local/state/national election during 2009, 2010, or 2011
data2$voted2<-Recode(data2$voted, recodes = "1=1; 0=0; else= NA ")
#Performed unpaid volunteer work during the last 2 years
data2$unpaidWork2<-Recode(data2$unpaidWork, recodes = "1=1; 0=0; else=NA ")
#Frequency of volunteer service
data2$volunteeringFreq2<-Recode(data2$volunteeringFreq, recodes = "1='lessthanonceamonth'; 2='onceamonth'; 3='onceaweek';else=NA ", as.factor=T)
data2$volunteering<-Recode(data2$volunteeringFreq, recodes = "1:3=1;else=0 ")SELECT RECODED VARIABLES AND KEEP COMPLETE CASES
data3<-data2 %>% select(studentID, psu,studentWeight,panelStudentWeight,strataID,Female,sex2, parentsed2, familyInc2,thirdExpectEdu2,GradExpect,internship2,research2,abroad2,community2,seniorExperience2,mentoring2,ImpactActivities2,voted2,volunteering,parentsBachelorOrMore) %>% filter(complete.cases(.))
data3<-data3 %>% filter(complete.cases(.))
#Filter out students who have not indicated clearly what their education expectations were
#data3<-filter(data3,thirdExpectEdu2!="NA")CREATE SURVEY DESIGN
options(survey.lonely.psu="adjust")
des<-svydesign(ids= ~psu, strata = ~strataID, weights= ~studentWeight, data= data3, nest=T)DESCRIPTIVE STATISTICS BY SEX AND EXPECTED LEVEL OF EDUCATION
Percent of men and women by expected level of education.Standard errors are adjusted to survey design
library(survey)
sv.table<-as.data.frame(svyby(formula = ~thirdExpectEdu2, by= ~sex2, design=des, FUN=svymean, na.rm=T))
t(sv.table)## Female Male
## sex2 "Female" "Male"
## thirdExpectEdu2HighSchool "0.08700098" "0.09732707"
## thirdExpectEdu2Associates "0.10602566" "0.08119032"
## thirdExpectEdu2Bachelors "0.2726824" "0.3331056"
## thirdExpectEdu2Certificate "0.03994399" "0.07473933"
## thirdExpectEdu2dontknow "0.06774006" "0.07054863"
## thirdExpectEdu2Graduate "0.4240965" "0.3430890"
## thirdExpectEdu2LessthanHS "0.002510407" "0.000000000"
## se.thirdExpectEdu2HighSchool "0.01451037" "0.01765306"
## se.thirdExpectEdu2Associates "0.01677427" "0.01551446"
## se.thirdExpectEdu2Bachelors "0.02232773" "0.02608747"
## se.thirdExpectEdu2Certificate "0.01035604" "0.01399850"
## se.thirdExpectEdu2dontknow "0.01534381" "0.01550985"
## se.thirdExpectEdu2Graduate "0.02440294" "0.02428952"
## se.thirdExpectEdu2LessthanHS "0.002507866" "0.000000000"DESCRIPTIVE STATISTICS BY EXPECTATION OF GRADUATE SCHOOL OR NOT AND NUMBER OF IMPACT ACTIVITIES
Percent of men and women by expected level of education.Standard errors are adjusted to survey design
library(survey)
sv.table2<-as.data.frame(svyby(formula = ~GradExpect, by= ~ImpactActivities2, design=des, FUN=svymean, na.rm=T))
t(sv.table2)## 0 1 2 3 4
## ImpactActivities2 0.00000000 1.00000000 2.00000000 3.00000000 4.00000000
## GradExpect 0.25012207 0.45847323 0.58051790 0.64972493 0.68846406
## se 0.01916117 0.04404977 0.05742597 0.06696498 0.07324312
## 5 6
## ImpactActivities2 5.0000000 6.0000000
## GradExpect 0.6559168 0.8849742
## se 0.1879861 0.1128552GENERATE PLOTS FROM MY DATA AND CHI-SQUARE ANALYSES
https://uc-r.github.io/descriptives_categorical
Post-Graduate Expectations by Sex
library(ggplot2)
library(ggthemes)## Warning: package 'ggthemes' was built under R version 3.6.2library(viridis)## Warning: package 'viridis' was built under R version 3.6.2## Loading required package: viridisLitechi1<-svyby(formula = ~sex2, by = ~GradExpect, design = des, FUN = svymean, na.rm=T)
svychisq(~sex2+GradExpect, design = des)##
## Pearson's X^2: Rao & Scott adjustment
##
## data: svychisq(~sex2 + GradExpect, design = des)
## F = 4.7378, ndf = 1, ddf = 123, p-value = 0.03142qplot(x=chi1$GradExpect ,y=chi1$sex2Female, data=chi1 ,xlab="test", ylab="females" )+
geom_errorbar(aes(x=GradExpect, ymin=sex2Female,ymax=sex2Female), width=.25)+
ggtitle(label = "% of Females with Post-Graduate Expectations")
Post-Graduate Expectations by Voting Status
The Chisquared tests of association for survey data shows that students who voted versus those that did not ARE statistically different in their expectations to attend post-graduate school.
chi2<-svyby(formula = ~voted2, by = ~GradExpect, design = des, FUN = svymean, na.rm=T)
svychisq(~voted2+GradExpect, design = des)##
## Pearson's X^2: Rao & Scott adjustment
##
## data: svychisq(~voted2 + GradExpect, design = des)
## F = 10.614, ndf = 1, ddf = 123, p-value = 0.001451qplot(x=chi2$GradExpect ,y=chi2$voted2, data=chi2 ,xlab="Post-Graduate Expectations 1=yes, 0=no", ylab="percent of participants who voted" )+
geom_errorbar(aes(x=GradExpect, ymin=voted2,ymax=voted2), width=.25)+
ggtitle(label = "% of Participants who Voted and Post-Graduate Expectations")
Post-Graduate Expectations by Volunteering Status
chi3<-svyby(formula = ~volunteering, by = ~GradExpect, design = des, FUN = svymean, na.rm=T)
svychisq(~volunteering+GradExpect, design = des)##
## Pearson's X^2: Rao & Scott adjustment
##
## data: svychisq(~volunteering + GradExpect, design = des)
## F = 45.008, ndf = 1, ddf = 123, p-value = 0.0000000006374qplot(x=chi3$GradExpect ,y=chi3$volunteering, data=chi3 ,xlab="Post-Graduate Expectations 1=yes, 0=no", ylab="percent of participants who volunteered" )+
geom_errorbar(aes(x=GradExpect, ymin=volunteering,ymax=volunteering), width=.25)+
ggtitle(label = "% of Participants who Volunteered and Post-Graduate Expectations")
Post-Graduate Expectations by having received Mentoring
chi4<-svyby(formula = ~mentoring2, by = ~GradExpect, design = des, FUN = svymean, na.rm=T)
svychisq(~mentoring2+GradExpect, design = des)##
## Pearson's X^2: Rao & Scott adjustment
##
## data: svychisq(~mentoring2 + GradExpect, design = des)
## F = 19.276, ndf = 1, ddf = 123, p-value = 0.00002409qplot(x=chi4$GradExpect ,y=chi4$mentoring2, data=chi4 ,xlab="Post-Graduate Expectations 1=yes, 0=no", ylab="percent of participants who received mentoring" )+
geom_errorbar(aes(x=GradExpect, ymin=mentoring2,ymax=mentoring2), width=.25)+
ggtitle(label = "% of Participants who Received Mentoring and Post-Graduate Expectations")
## Post-Graduate Expectations by Parent’s Education (greater than a bacherlor’s degree)
chi5<-svyby(formula = ~parentsBachelorOrMore, by = ~GradExpect, design = des, FUN = svymean, na.rm=T)
svychisq(~parentsBachelorOrMore+GradExpect, design = des)##
## Pearson's X^2: Rao & Scott adjustment
##
## data: svychisq(~parentsBachelorOrMore + GradExpect, design = des)
## F = 29.709, ndf = 1, ddf = 123, p-value = 0.0000002629qplot(x=chi5$GradExpect ,y=chi5$parentsBachelorOrMore1, data=chi5 ,xlab="Survey Participant's Post-Graduate Expectations 1=yes, 0=no", ylab="Percent of Parents with a Bachelors or More" )+
geom_errorbar(aes(x=GradExpect, ymin=parentsBachelorOrMore1 , ymax=parentsBachelorOrMore1), width=.25)+
ggtitle(label = "% of Parents with a Bachelors or More and Participant's Post-Graduate Expectations")
Post-Graduate Expectations and ALL Impact Activities
chi6<-svyby(formula = ~ImpactActivities2, by = ~GradExpect, design = des, FUN = svymean, na.rm=T)
svychisq(~ImpactActivities2+GradExpect, design = des)##
## Pearson's X^2: Rao & Scott adjustment
##
## data: svychisq(~ImpactActivities2 + GradExpect, design = des)
## F = 14.957, ndf = 5.2665, ddf = 647.7797, p-value = 0.00000000000001723qplot(x=chi6$GradExpect ,y=chi6$ImpactActivities2, data=chi6 ,xlab="Survey Participant's Post-Graduate Expectations 1=yes, 0=no", ylab="Percent of Parents with a Bachelors or More" )+
geom_errorbar(aes(x=GradExpect, ymin=ImpactActivities2 , ymax=ImpactActivities2), width=.25)+
ggtitle(label = "Impact Activities and Participant's Post-Graduate Expectations")
## Post-Graduate Expectations and Parent’s Income
chi7<-svyby(formula = ~familyInc2, by = ~GradExpect, design = des, FUN = svymean, na.rm=T)
svychisq(~familyInc2+GradExpect, design = des)##
## Pearson's X^2: Rao & Scott adjustment
##
## data: svychisq(~familyInc2 + GradExpect, design = des)
## F = 9.8226, ndf = 2.9101, ddf = 357.9370, p-value = 0.000004031qplot(x=chi7$GradExpect, y=chi7$familyInc2Morethan100k, data=chi7 ,xlab="Survey Participant's Post-Graduate Expectations 1=yes, 0=no", ylab="Family Income" )+
geom_errorbar(aes(x=GradExpect, ymin=familyInc2Morethan100k , ymax=familyInc2Morethan100k), width=.25)+
ggtitle(label = "Parent's Income > 100k at Baseline and Post-Graduate Expectations")
ADDITIONAL DESCRIPTIVE INFORMATION
Impact Activities
prop.table(questionr::wtd.table(data3$ImpactActivities2,weights=data3$studentWeight))*100## 0 1 2 3 4 5 6
## 54.5999218 19.4464412 13.3472575 6.5446368 4.8153862 0.6541071 0.5922493ggplot(data3, aes(x = `ImpactActivities2`)) +
geom_bar(aes(y = (..count..)/sum(..count..))) +
xlab("Number of Impact Activities") +
scale_y_continuous(labels = scales::percent, name = "Percent of Respondents") +
theme(axis.text.x = element_text(angle = 45, hjust = 1)) + theme_hc()+
ggtitle("Percent of Participants by Number of Impact Activities")
Parents Education
prop.table(questionr::wtd.table(data3$parentsed2,weights=data3$studentWeight))*100## Associate Bachelors Graduate HighSchool
## 11.03330 16.69473 10.43277 19.14218
## LessthanAssociate LessthanBach LessthanHS
## 12.12562 12.39418 18.17722ggplot(data3, aes(x = `parentsed2`)) +
geom_bar(aes(y = (..count..)/sum(..count..))) +
xlab("Participants Parents Education (ordered by size)") +
scale_y_continuous(labels = scales::percent, name = "Percent of Respondents") +
theme(axis.text.x = element_text(angle = 45, hjust = 1)) + theme_hc()+
ggtitle("Percent of Parent's Educational Attainment of Survey Respondents")
## Education Expectations
prop.table(questionr::wtd.table(data3$thirdExpectEdu2,weights=data3$studentWeight))*100## HighSchool Associates Bachelors Certificate dontknow Graduate
## 9.1549656 9.5085599 29.9299075 5.5271465 6.8977248 38.8412393
## LessthanHS
## 0.1404564reorder_size <- function(x) {
factor(x, levels = names(sort(table(x), decreasing = TRUE)))
}
ggplot(data3, aes(x = reorder_size(`thirdExpectEdu2`))) +
geom_bar(aes(y = (..count..)/sum(..count..))) +
xlab("Education Expectations (ordered by size)") +
scale_y_continuous(labels = scales::percent, name = "Percent of Respondents") +
theme(axis.text.x = element_text(angle = 45, hjust = 1)) + theme_hc() +
ggtitle("Percent of Survey Respondents by Educational Expectations")
## Educational Expectation by Sex
prop.table(questionr::wtd.table(data3$thirdExpectEdu2, data3$sex2,weights=data3$studentWeight))*100## Female Male
## HighSchool 4.8676727 4.2872929
## Associates 5.9320967 3.5764633
## Bachelors 15.2564825 14.6734250
## Certificate 2.2348514 3.2922951
## dontknow 3.7900318 3.1076930
## Graduate 23.7280424 15.1131969
## LessthanHS 0.1404564 0.0000000ggplot(data3, aes(thirdExpectEdu2, (..count..)/sum(..count..)*100 )) + geom_bar(aes(fill = sex2), position = "dodge")+ xlab("Education Expectations") + ylab("Percent of Respondents") +
ggtitle("Percent of Survey Respondents by Educational Expectations by Sex")
prop.table(questionr::wtd.table(data3$GradExpect, data3$sex2,weights=data3$studentWeight))*100## Female Male
## 0 32.22159 28.93717
## 1 23.72804 15.11320ggplot(data3, aes(GradExpect, (..count..)/sum(..count..)*100 )) + geom_bar(aes(fill = sex2), position = "dodge") + xlab("Post-Graduate Expectations") + ylab("Percent of Respondents") +
ggtitle("Percent of Survey Respondents by Post-Graduate vs. Non Post-Graduate Expectations by Sex")
CLEAN UP ENV.
rm(chi1, chi2, chi3, chi4, chi5, chi6, chi7, sv.table, sv.table2)REGRESSION ANALYSIS
The outcome of interest is whether a survey participant is expected to enroll in graduate school (masters, phd, doctoral) or not. A logistic regression is appropriate to use in this case.
Logit model
library(survey)
#Average of Participants Expecting to go to Graduate School (1)
svymean(~GradExpect, design=des)## mean SE
## GradExpect 0.38841 0.0162#LOGIT MODEL
fit.logit<-svyglm(as.numeric(GradExpect) ~ Female + seniorExperience2 + community2 + research2 + internship2 + volunteering + voted2 + mentoring2 + parentsBachelorOrMore + familyInc2, design = des, family = binomial)## Warning in eval(family$initialize): non-integer #successes in a binomial glm!summary(fit.logit)##
## Call:
## svyglm(formula = as.numeric(GradExpect) ~ Female + seniorExperience2 +
## community2 + research2 + internship2 + volunteering + voted2 +
## mentoring2 + parentsBachelorOrMore + familyInc2, design = des,
## family = binomial)
##
## Survey design:
## svydesign(ids = ~psu, strata = ~strataID, weights = ~studentWeight,
## data = data3, nest = T)
##
## Coefficients:
## Estimate Std. Error t value Pr(>|t|)
## (Intercept) -1.74605 0.17882 -9.764 < 0.0000000000000002 ***
## Female 0.17107 0.17232 0.993 0.323013
## seniorExperience2 0.48684 0.25865 1.882 0.062424 .
## community2 0.13993 0.31787 0.440 0.660645
## research2 0.68328 0.24011 2.846 0.005279 **
## internship2 0.76278 0.22458 3.396 0.000948 ***
## volunteering 0.67384 0.18481 3.646 0.000407 ***
## voted2 0.41344 0.16995 2.433 0.016585 *
## mentoring2 0.15738 0.25722 0.612 0.541888
## parentsBachelorOrMore1 0.63658 0.19625 3.244 0.001559 **
## familyInc250kto100k -0.03141 0.22949 -0.137 0.891377
## familyInc2Lessthan20k 0.22649 0.24171 0.937 0.350782
## familyInc2Morethan100k 1.38415 0.33683 4.109 0.0000762 ***
## ---
## Signif. codes: 0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1
##
## (Dispersion parameter for binomial family taken to be 1.021202)
##
## Number of Fisher Scoring iterations: 4#Short version of logit model
#LOGIT MODEL
fit.logit2<-svyglm(as.numeric(GradExpect) ~ ImpactActivities2 + familyInc2, design = des, family = binomial)## Warning in eval(family$initialize): non-integer #successes in a binomial glm!summary(fit.logit2)##
## Call:
## svyglm(formula = as.numeric(GradExpect) ~ ImpactActivities2 +
## familyInc2, design = des, family = binomial)
##
## Survey design:
## svydesign(ids = ~psu, strata = ~strataID, weights = ~studentWeight,
## data = data3, nest = T)
##
## Coefficients:
## Estimate Std. Error t value Pr(>|t|)
## (Intercept) -1.12975 0.12905 -8.754 0.0000000000000164 ***
## ImpactActivities2 0.53486 0.06583 8.125 0.0000000000004749 ***
## familyInc250kto100k 0.11895 0.22601 0.526 0.600
## familyInc2Lessthan20k 0.09553 0.22838 0.418 0.676
## familyInc2Morethan100k 1.55779 0.30825 5.054 0.0000015849870031 ***
## ---
## Signif. codes: 0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1
##
## (Dispersion parameter for binomial family taken to be 1.009485)
##
## Number of Fisher Scoring iterations: 4TEST FOR MULTICOLLINEARITY
The variance inflation factor (VIF) quantifies the extent of correlation between one predictor and the other predictors in a model. It is used for diagnosing collinearity/multicollinearity. Higher values signify that it is difficult to impossible to assess accurately the contribution of predictors to a model.
library(rms)## Warning: package 'rms' was built under R version 3.6.2## Loading required package: Hmisc## Warning: package 'Hmisc' was built under R version 3.6.2## Loading required package: lattice## Loading required package: Formula##
## Attaching package: 'Hmisc'## The following objects are masked from 'package:dplyr':
##
## src, summarize## The following objects are masked from 'package:questionr':
##
## describe, wtd.mean, wtd.table, wtd.var## The following object is masked from 'package:survey':
##
## deff## The following objects are masked from 'package:base':
##
## format.pval, units## Loading required package: SparseM## Warning: package 'SparseM' was built under R version 3.6.2##
## Attaching package: 'SparseM'## The following object is masked from 'package:base':
##
## backsolve##
## Attaching package: 'rms'## The following objects are masked from 'package:car':
##
## Predict, vif## The following object is masked from 'package:survey':
##
## calibratevif(fit.logit)## Female seniorExperience2 community2
## 1.290482 1.251171 1.432380
## research2 internship2 volunteering
## 1.169167 1.605373 1.335499
## voted2 mentoring2 parentsBachelorOrMore1
## 1.297728 1.382488 1.218255
## familyInc250kto100k familyInc2Lessthan20k familyInc2Morethan100k
## 1.350884 1.407225 1.349944