Worried about Academic Probation?
Angela Cao
03/04/2020
- Workstation Setup
- Get the dataset from desktop and re-define the main variables
- Descriptive statistics for the main data (2010)
- Code snippet 1: Counting the number of observations with X different from missing
- Code snippet 2: Counting the unique values of X
- Code snippet 3: Using rddensity
- Code snippet 4: Using rdrobust on hsgrade_pct
- Code snippet 5: (Figure 3.2)-Figure 2 in report: Using rdplot on hsgrade_pct
- Table 3.3 in A Practical Introduction to Regression Discontinuity Designs, Volume II
- Figure 3.3:Figure 3.3: RD Plots for Predetermined Covariates—LSO Application, 1.3 in the report: High School Grade Percentile
- Figure 3.3:Figure 3.3: RD Plots for Predetermined Covariates—LSO Application, 1.4 in the report: Total Credits in First Year
- Figure 3.3:Figure 3.3: RD Plots for Predetermined Covariates—LSO Application, 1.5 in the
- Figure 3.3:Figure 3.3: RD Plots for Predetermined Covariates—LSO Application, 1.6 in the report: English First Language Indicator
- Figure 3.3:Figure 3.3: RD Plots for Predetermined Covariates—LSO Application, 1.7 in the report: Male Indicator
- Figure 3.3:Figure 3.3: RD Plots for Predetermined Covariates—LSO Application, 1.8 in the report: Born in North America Indicator
- Code snippet 6 (Figure 3.4) - 1.9 in the report: Using rdplot on the outcome - Figure 1.9: RD Plot for nextGPA—LSO Data
- Code snippet 7: Using rdrobust on the outcome
- Code snippet 8: Using rdrobust and showing the objects it returns
- Code snippet 9: Using rdrobust with clustered standard errors
- Code snippet 10: Using rdrobust on the collapsed data
- Code snippet 11: Binomial test with rdwinselect
- Binomial Test by Hand
- Code snippet 13: Using rdrandinf on hsgrade_pct
- Code snippet 14: Using rdwinselect with covariates and windows with equal number of mass points to each side of the cutoff
- Code snippet 15: Using rdrandinf on the Outcome
- Reference
A re-analysis of regression discontinuity (RD) design on a discrete running variable for academic probation placements.
Reference: A Practical Introduction to Regression Discontinuity Designs Volume II Authors: Matias D. Cattaneo, Nicolas Idrobo and Rocio Titiunik
Package: https://sites.google.com/site/rdpackages/
Workstation Setup
FOREIGN: install.packages(‘foreign’); GGPLOT2: install.packages(‘ggplot2’); GRID: install.packages(‘grid’); LPDENSITY: install.packages(‘lpdensity’); RDDENSITY: install.packages(‘rddensity’); RDLOCRAND: install.packages(‘rdlocrand’); RDROBUST: install.packages(‘rdrobust’); TEACHINGDEMOS: install.packages(‘TeachingDemos’); GGTHEMES:install.package(‘ggthemes’) \(~\)
Workstation Setup
\(~\)
library(foreign)
library(ggplot2)
library(lpdensity)
library(rddensity)
library(rdrobust)
library(rdlocrand)
library(TeachingDemos)
library(ggthemes)Get the dataset from desktop and re-define the main variables
\(~\)
library(readr)
data <- read_csv("CIT_2019_Cambridge_education.csv")## Parsed with column specification:
## cols(
## X = col_double(),
## nextGPA = col_double(),
## left_school = col_double(),
## nextGPA_nonorm = col_double(),
## T = col_double(),
## hsgrade_pct = col_double(),
## totcredits_year1 = col_double(),
## age_at_entry = col_double(),
## male = col_double(),
## bpl_north_america = col_double(),
## english = col_double(),
## loc_campus1 = col_double(),
## loc_campus2 = col_double(),
## loc_campus3 = col_double(),
## clustervar = col_double()
## )nextGPA = data$nextGPA
left_school = data$left_school
nextGPA_nonorm = data$nextGPA_nonorm
X = data$X
T = data$T
T_X = T*XDescriptive statistics for the main data (2010)
\(~\)
installed.packages("skimr")## Package LibPath Version Priority Depends Imports LinkingTo Suggests
## Enhances License License_is_FOSS License_restricts_use OS_type Archs
## MD5sum NeedsCompilation Builtlibrary(skimr)
skim(data)| Name | data |
| Number of rows | 44362 |
| Number of columns | 15 |
| _______________________ | |
| Column type frequency: | |
| numeric | 15 |
| ________________________ | |
| Group variables | None |
Variable type: numeric
| skim_variable | n_missing | complete_rate | mean | sd | p0 | p25 | p50 | p75 | p100 | hist |
|---|---|---|---|---|---|---|---|---|---|---|
| X | 0 | 1.00 | -0.91 | 0.90 | -2.8 | -1.58 | -0.98 | -0.32 | 1.6 | ▃▇▇▃▁ |
| nextGPA | 3780 | 0.91 | 1.05 | 0.92 | -1.6 | 0.50 | 1.17 | 1.73 | 2.8 | ▁▂▆▇▃ |
| left_school | 0 | 1.00 | 0.05 | 0.22 | 0.0 | 0.00 | 0.00 | 0.00 | 1.0 | ▇▁▁▁▁ |
| nextGPA_nonorm | 3780 | 0.91 | 2.57 | 0.91 | 0.0 | 2.00 | 2.70 | 3.26 | 4.3 | ▁▃▆▇▃ |
| T | 0 | 1.00 | 0.16 | 0.37 | 0.0 | 0.00 | 0.00 | 0.00 | 1.0 | ▇▁▁▁▂ |
| hsgrade_pct | 0 | 1.00 | 50.17 | 28.86 | 1.0 | 25.00 | 50.00 | 75.00 | 100.0 | ▇▇▇▇▇ |
| totcredits_year1 | 0 | 1.00 | 4.57 | 0.51 | 3.0 | 4.00 | 5.00 | 5.00 | 6.5 | ▁▃▇▁▁ |
| age_at_entry | 0 | 1.00 | 18.67 | 0.74 | 17.0 | 18.00 | 19.00 | 19.00 | 21.0 | ▁▅▇▁▁ |
| male | 0 | 1.00 | 0.38 | 0.49 | 0.0 | 0.00 | 0.00 | 1.00 | 1.0 | ▇▁▁▁▅ |
| bpl_north_america | 0 | 1.00 | 0.87 | 0.34 | 0.0 | 1.00 | 1.00 | 1.00 | 1.0 | ▁▁▁▁▇ |
| english | 0 | 1.00 | 0.71 | 0.45 | 0.0 | 0.00 | 1.00 | 1.00 | 1.0 | ▃▁▁▁▇ |
| loc_campus1 | 0 | 1.00 | 0.58 | 0.49 | 0.0 | 0.00 | 1.00 | 1.00 | 1.0 | ▆▁▁▁▇ |
| loc_campus2 | 0 | 1.00 | 0.17 | 0.38 | 0.0 | 0.00 | 0.00 | 0.00 | 1.0 | ▇▁▁▁▂ |
| loc_campus3 | 0 | 1.00 | 0.24 | 0.43 | 0.0 | 0.00 | 0.00 | 0.00 | 1.0 | ▇▁▁▁▂ |
| clustervar | 0 | 1.00 | 0.91 | 0.90 | -1.6 | 0.32 | 0.98 | 1.58 | 2.8 | ▁▃▇▇▃ |
| *** | ||||||||||
| ### A Practical Intr | oduction to | Regression Disco | ntinuity | Designs | , Volum | e II, Se | ction 3: | RD Desi | gns with | Discrete Scores, Figure 3.1 - 1.1 in Report: Histogram |
| \(~\) |
tempdata = as.data.frame(X); colnames(tempdata) = c("v1");
p = ggplot(data=tempdata, aes(tempdata$v1))+
geom_histogram(breaks=seq(-2.8, 0, by = 0.1), col="black", fill="blue", alpha = 1)+
geom_histogram(breaks=seq(0, 1.6, by = 0.1), col="black", fill="red", alpha = 1)+
labs(x="Score", y="Number of Observations", title = "Histogram of Transformed Running Variable—LSO Data")+geom_vline(xintercept=0, color="black")+
theme_economist()
p## Warning: Use of `tempdata$v1` is discouraged. Use `v1` instead.
## Warning: Use of `tempdata$v1` is discouraged. Use `v1` instead.
# ggsave("./outputs/Vol-2-R_histogram.png", plot = p, width = 10, height = 5, units = "in")Code snippet 1: Counting the number of observations with X different from missing
\(~\)
txtStart("./outputs/Vol-2-R_lso_countX.txt", commands=TRUE, results=TRUE, append=FALSE, visible.only=TRUE)## Output being copied to text file,
## use txtStop to endlength(X)## [1] 44362txtStop()Code snippet 2: Counting the unique values of X
\(~\)
txtStart("./outputs/Vol-2-R_lso_uniqueX.txt", commands=TRUE, results=TRUE, append=FALSE, visible.only=TRUE)## Output being copied to text file,
## use txtStop to endlength(unique(X))## [1] 430txtStop()Code snippet 3: Using rddensity
\(~\)
txtStart("./outputs/Vol-2-R_lso_falsification_rddensity.txt", commands=TRUE, results=TRUE, append=FALSE, visible.only=TRUE)## Output being copied to text file,
## use txtStop to endout = rddensity(X)
summary(out)##
## RD Manipulation Test using local polynomial density estimation.
##
## Number of obs = 44362
## Model = unrestricted
## Kernel = triangular
## BW method = comb
## VCE method = jackknife
##
## Cutoff c = 0 Left of c Right of c
## Number of obs 37211 7151
## Eff. Number of obs 10083 4137
## Order est. (p) 2 2
## Order bias (q) 3 3
## BW est. (h) 0.706 0.556
##
## Method T P > |T|
## Robust -0.4544 0.6496txtStop()Code snippet 4: Using rdrobust on hsgrade_pct
\(~\)
txtStart("./outputs/Vol-2-R_lso_falsification_rdrobust_hsgrade_pct.txt", commands=TRUE, results=TRUE, append=FALSE, visible.only=TRUE)## Output being copied to text file,
## use txtStop to endout = rdrobust(data$hsgrade_pct, X)## [1] "Mass points detected in the running variable."summary(out)## Call: rdrobust
##
## Number of Obs. 44362
## BW type mserd
## Kernel Triangular
## VCE method NN
##
## Number of Obs. 37211 7151
## Eff. Number of Obs. 6934 3972
## Order est. (p) 1 1
## Order bias (q) 2 2
## BW est. (h) 0.518 0.518
## BW bias (b) 0.801 0.801
## rho (h/b) 0.646 0.646
## Unique Obs. 274 156
##
## =============================================================================
## Method Coef. Std. Err. z P>|z| [ 95% C.I. ]
## =============================================================================
## Conventional 1.140 0.961 1.186 0.235 [-0.743 , 3.023]
## Robust - - 1.140 0.254 [-0.943 , 3.566]
## =============================================================================txtStop()Code snippet 5: (Figure 3.2)-Figure 2 in report: Using rdplot on hsgrade_pct
\(~\)
txtStart("./outputs/Vol-2-R_lso_falsification_rdplot_hsgrade_pct-COMMAND-ONLY.txt",
commands=TRUE, results=FALSE, append=FALSE, visible.only=TRUE)## Output being copied to text file,
## use txtStop to endrdplot(data$hsgrade_pct, X, x.label = "Score", y.label = "Frequency Count", title="Regression Discontinuity Plot for High School Grade Percentile - LSO Data",col.dots ="blue", col.lines ="red")
txtStop()
# png("./outputs/Vol-2-R_lso_falsification_rdplot_hsgrade_pct.png")
rdplot(data$hsgrade_pct, X, x.label = "Score", y.label = "Frequency Count", title="Regression Discontinuity Plot for 'hsgrade_pct' - LSO Data")
dev.off()## null device
## 1Table 3.3 in A Practical Introduction to Regression Discontinuity Designs, Volume II
\(~\)
summary(rdrobust(data$hsgrade_pct, X))## [1] "Mass points detected in the running variable."
## Call: rdrobust
##
## Number of Obs. 44362
## BW type mserd
## Kernel Triangular
## VCE method NN
##
## Number of Obs. 37211 7151
## Eff. Number of Obs. 6934 3972
## Order est. (p) 1 1
## Order bias (q) 2 2
## BW est. (h) 0.518 0.518
## BW bias (b) 0.801 0.801
## rho (h/b) 0.646 0.646
## Unique Obs. 274 156
##
## =============================================================================
## Method Coef. Std. Err. z P>|z| [ 95% C.I. ]
## =============================================================================
## Conventional 1.140 0.961 1.186 0.235 [-0.743 , 3.023]
## Robust - - 1.140 0.254 [-0.943 , 3.566]
## =============================================================================summary(rdrobust(data$totcredits_year1, X))## [1] "Mass points detected in the running variable."
## Call: rdrobust
##
## Number of Obs. 44362
## BW type mserd
## Kernel Triangular
## VCE method NN
##
## Number of Obs. 37211 7151
## Eff. Number of Obs. 3995 2799
## Order est. (p) 1 1
## Order bias (q) 2 2
## BW est. (h) 0.322 0.322
## BW bias (b) 0.609 0.609
## rho (h/b) 0.529 0.529
## Unique Obs. 274 156
##
## =============================================================================
## Method Coef. Std. Err. z P>|z| [ 95% C.I. ]
## =============================================================================
## Conventional 0.077 0.028 2.722 0.006 [0.022 , 0.133]
## Robust - - 2.768 0.006 [0.026 , 0.151]
## =============================================================================summary(rdrobust(data$age_at_entry, X))## [1] "Mass points detected in the running variable."
## Call: rdrobust
##
## Number of Obs. 44362
## BW type mserd
## Kernel Triangular
## VCE method NN
##
## Number of Obs. 37211 7151
## Eff. Number of Obs. 6444 3798
## Order est. (p) 1 1
## Order bias (q) 2 2
## BW est. (h) 0.492 0.492
## BW bias (b) 0.755 0.755
## rho (h/b) 0.652 0.652
## Unique Obs. 274 156
##
## =============================================================================
## Method Coef. Std. Err. z P>|z| [ 95% C.I. ]
## =============================================================================
## Conventional 0.017 0.033 0.511 0.609 [-0.048 , 0.081]
## Robust - - 0.528 0.598 [-0.057 , 0.098]
## =============================================================================summary(rdrobust(data$male, X))## [1] "Mass points detected in the running variable."
## Call: rdrobust
##
## Number of Obs. 44362
## BW type mserd
## Kernel Triangular
## VCE method NN
##
## Number of Obs. 37211 7151
## Eff. Number of Obs. 6825 3932
## Order est. (p) 1 1
## Order bias (q) 2 2
## BW est. (h) 0.508 0.508
## BW bias (b) 0.828 0.828
## rho (h/b) 0.614 0.614
## Unique Obs. 274 156
##
## =============================================================================
## Method Coef. Std. Err. z P>|z| [ 95% C.I. ]
## =============================================================================
## Conventional -0.012 0.021 -0.561 0.575 [-0.053 , 0.029]
## Robust - - -0.687 0.492 [-0.066 , 0.032]
## =============================================================================summary(rdrobust(data$bpl_north_america, X))## [1] "Mass points detected in the running variable."
## Call: rdrobust
##
## Number of Obs. 44362
## BW type mserd
## Kernel Triangular
## VCE method NN
##
## Number of Obs. 37211 7151
## Eff. Number of Obs. 7171 4068
## Order est. (p) 1 1
## Order bias (q) 2 2
## BW est. (h) 0.536 0.536
## BW bias (b) 0.858 0.858
## rho (h/b) 0.624 0.624
## Unique Obs. 274 156
##
## =============================================================================
## Method Coef. Std. Err. z P>|z| [ 95% C.I. ]
## =============================================================================
## Conventional 0.015 0.014 1.077 0.282 [-0.012 , 0.043]
## Robust - - 1.091 0.275 [-0.015 , 0.051]
## =============================================================================summary(rdrobust(data$english, X))## [1] "Mass points detected in the running variable."
## Call: rdrobust
##
## Number of Obs. 44362
## BW type mserd
## Kernel Triangular
## VCE method NN
##
## Number of Obs. 37211 7151
## Eff. Number of Obs. 7588 4206
## Order est. (p) 1 1
## Order bias (q) 2 2
## BW est. (h) 0.568 0.568
## BW bias (b) 0.920 0.920
## rho (h/b) 0.618 0.618
## Unique Obs. 274 156
##
## =============================================================================
## Method Coef. Std. Err. z P>|z| [ 95% C.I. ]
## =============================================================================
## Conventional -0.035 0.019 -1.899 0.058 [-0.072 , 0.001]
## Robust - - -1.809 0.070 [-0.083 , 0.003]
## =============================================================================summary(rdrobust(data$loc_campus1, X))## [1] "Mass points detected in the running variable."
## Call: rdrobust
##
## Number of Obs. 44362
## BW type mserd
## Kernel Triangular
## VCE method NN
##
## Number of Obs. 37211 7151
## Eff. Number of Obs. 3917 2655
## Order est. (p) 1 1
## Order bias (q) 2 2
## BW est. (h) 0.313 0.313
## BW bias (b) 0.534 0.534
## rho (h/b) 0.586 0.586
## Unique Obs. 274 156
##
## =============================================================================
## Method Coef. Std. Err. z P>|z| [ 95% C.I. ]
## =============================================================================
## Conventional -0.016 0.027 -0.596 0.551 [-0.069 , 0.037]
## Robust - - -0.854 0.393 [-0.089 , 0.035]
## =============================================================================summary(rdrobust(data$loc_campus2, X))## [1] "Mass points detected in the running variable."
## Call: rdrobust
##
## Number of Obs. 44362
## BW type mserd
## Kernel Triangular
## VCE method NN
##
## Number of Obs. 37211 7151
## Eff. Number of Obs. 6934 3972
## Order est. (p) 1 1
## Order bias (q) 2 2
## BW est. (h) 0.517 0.517
## BW bias (b) 0.795 0.795
## rho (h/b) 0.650 0.650
## Unique Obs. 274 156
##
## =============================================================================
## Method Coef. Std. Err. z P>|z| [ 95% C.I. ]
## =============================================================================
## Conventional -0.016 0.017 -0.899 0.369 [-0.050 , 0.018]
## Robust - - -0.863 0.388 [-0.059 , 0.023]
## =============================================================================summary(rdrobust(data$loc_campus3, X))## [1] "Mass points detected in the running variable."
## Call: rdrobust
##
## Number of Obs. 44362
## BW type mserd
## Kernel Triangular
## VCE method NN
##
## Number of Obs. 37211 7151
## Eff. Number of Obs. 3623 2519
## Order est. (p) 1 1
## Order bias (q) 2 2
## BW est. (h) 0.292 0.292
## BW bias (b) 0.523 0.523
## rho (h/b) 0.559 0.559
## Unique Obs. 274 156
##
## =============================================================================
## Method Coef. Std. Err. z P>|z| [ 95% C.I. ]
## =============================================================================
## Conventional 0.032 0.027 1.203 0.229 [-0.020 , 0.085]
## Robust - - 1.427 0.154 [-0.016 , 0.104]
## =============================================================================Figure 3.3:Figure 3.3: RD Plots for Predetermined Covariates—LSO Application, 1.3 in the report: High School Grade Percentile
\(~\)
# png("./outputs/Vol-2-R_lso_falsification_rdplot_hsgrade_pct.png")
rdplot(data$hsgrade_pct, X, x.label = "Score", y.label = "Frequency Count", title="High School Grade Percentile")
dev.off()## null device
## 1Figure 3.3:Figure 3.3: RD Plots for Predetermined Covariates—LSO Application, 1.4 in the report: Total Credits in First Year
\(~\)
# png("./outputs/Vol-2-R_lso_falsification_rdplot_totcredits_year1.png")
rdplot(data$totcredits_year1, X, x.label = "Score", y.label = "Frequency Count", title="Total Credits in First Year")
dev.off()## null device
## 1Figure 3.3:Figure 3.3: RD Plots for Predetermined Covariates—LSO Application, 1.5 in the
\(~\)
# png("./outputs/Vol-2-R_lso_falsification_rdplot_age_at_entry.png")
rdplot(data$age_at_entry, X, x.label = "Score", y.label = "Frequency Count", title="Age at Entry")
dev.off()## null device
## 1Figure 3.3:Figure 3.3: RD Plots for Predetermined Covariates—LSO Application, 1.6 in the report: English First Language Indicator
\(~\)
# png("./outputs/Vol-2-R_lso_falsification_rdplot_english.png")
rdplot(data$english, X, x.label = "Score", y.label = "Frequency Count", title="English First Language Indicator")
dev.off()## null device
## 1Figure 3.3:Figure 3.3: RD Plots for Predetermined Covariates—LSO Application, 1.7 in the report: Male Indicator
\(~\)
# png("./outputs/Vol-2-R_lso_falsification_rdplot_male.png")
rdplot(data$male, X, x.label = "Score", y.label = "Frequency Count", title="Male Indicator")
dev.off()## null device
## 1Figure 3.3:Figure 3.3: RD Plots for Predetermined Covariates—LSO Application, 1.8 in the report: Born in North America Indicator
\(~\)
# png("./outputs/Vol-2-R_lso_falsification_rdplot_bpl_north_america.png")
rdplot(data$bpl_north_america, X, x.label = "Score", y.label = "Frequency Count", title="Born in North America Indicator")
dev.off()## null device
## 1Code snippet 6 (Figure 3.4) - 1.9 in the report: Using rdplot on the outcome - Figure 1.9: RD Plot for nextGPA—LSO Data
\(~\)
txtStart("./outputs/Vol-2-R_lso3_rdplot_esmv.txt", commands=TRUE, results=TRUE, append=FALSE, visible.only=TRUE)## Output being copied to text file,
## use txtStop to endout = rdplot(nextGPA_nonorm, X, binselect = 'esmv')
summary(out)## Call: rdplot
##
## Number of Obs. 40582
## Kernel Uniform
##
## Number of Obs. 34854 5728
## Eff. Number of Obs. 34854 5728
## Order poly. fit (p) 4 4
## BW poly. fit (h) 2.800 1.600
## Number of bins scale 1 1
##
## Bins Selected 690 362
## Average Bin Length 0.004 0.004
## Median Bin Length 0.004 0.004
##
## IMSE-optimal bins 44 13
## Mimicking Variance bins 690 362
##
## Relative to IMSE-optimal:
## Implied scale 15.682 27.846
## WIMSE variance weight 0.000 0.000
## WIMSE bias weight 1.000 1.000txtStop()
# pdf("./outputs/Vol-2-R_lso3_rdplot_esmv.pdf")
rdplot(nextGPA_nonorm, X, binselect = 'esmv', x.label = 'Score', y.label = 'Outcome', title = 'Regression Discontinuity Plot for nextGPA—LSO Data')
dev.off()## null device
## 1Code snippet 7: Using rdrobust on the outcome
\(~\)
txtStart("./outputs/Vol-2-R_lso3_rdrobust_triangular_mserd_p1_rhofree_regterm1.txt",
commands=TRUE, results=TRUE, append=FALSE, visible.only=TRUE)## Output being copied to text file,
## use txtStop to endout = rdrobust(nextGPA_nonorm, X, kernel = 'triangular', p = 1, bwselect = 'mserd')## [1] "Mass points detected in the running variable."summary(out)## Call: rdrobust
##
## Number of Obs. 40582
## BW type mserd
## Kernel Triangular
## VCE method NN
##
## Number of Obs. 34854 5728
## Eff. Number of Obs. 5612 3164
## Order est. (p) 1 1
## Order bias (q) 2 2
## BW est. (h) 0.470 0.470
## BW bias (b) 0.746 0.746
## rho (h/b) 0.630 0.630
## Unique Obs. 274 155
##
## =============================================================================
## Method Coef. Std. Err. z P>|z| [ 95% C.I. ]
## =============================================================================
## Conventional 0.224 0.038 5.852 0.000 [0.149 , 0.299]
## Robust - - 4.726 0.000 [0.126 , 0.304]
## =============================================================================txtStop()Code snippet 8: Using rdrobust and showing the objects it returns
\(~\)
txtStart("./outputs/Vol-2-R_lso3_rdrobust_triangular_mserd_p1_rhofree_regterm1_namescoefsout_all.txt",
commands=TRUE, results=TRUE, append=FALSE, visible.only=TRUE)## Output being copied to text file,
## use txtStop to endrdout = rdrobust(nextGPA_nonorm, X, kernel = 'triangular', p = 1, bwselect = 'mserd')## [1] "Mass points detected in the running variable."print(names(rdout))## [1] "Estimate" "bws" "coef" "se" "z"
## [6] "pv" "ci" "beta_p_l" "beta_p_r" "V_cl_l"
## [11] "V_cl_r" "V_rb_l" "V_rb_r" "N" "N_h"
## [16] "N_b" "M" "tau_cl" "tau_bc" "c"
## [21] "p" "q" "bias" "kernel" "all"
## [26] "vce" "bwselect" "level" "masspoints" "call"print(rdout$beta_p_r)## [,1]
## [1,] 2.0681763
## [2,] -0.6804732print(rdout$beta_p_l)## [,1]
## [1,] 1.8444877
## [2,] -0.6853278txtStop()Code snippet 9: Using rdrobust with clustered standard errors
\(~\)
txtStart("./outputs/Vol-2-R_lso3_rdrobust_triangular_mserd_p1_rhofree_regterm1_cluster.txt",
commands=TRUE, results=TRUE, append=FALSE, visible.only=TRUE)## Output being copied to text file,
## use txtStop to endclustervar = X
out = rdrobust(nextGPA_nonorm, X, kernel = 'triangular', p = 1, bwselect = 'mserd', vce = 'hc0', cluster = clustervar)## [1] "Mass points detected in the running variable."summary(out)## Call: rdrobust
##
## Number of Obs. 40582
## BW type mserd
## Kernel Triangular
## VCE method HC0
##
## Number of Obs. 34854 5728
## Eff. Number of Obs. 5008 3016
## Order est. (p) 1 1
## Order bias (q) 2 2
## BW est. (h) 0.428 0.428
## BW bias (b) 0.696 0.696
## rho (h/b) 0.616 0.616
## Unique Obs. 274 155
##
## =============================================================================
## Method Coef. Std. Err. z P>|z| [ 95% C.I. ]
## =============================================================================
## Conventional 0.221 0.032 6.991 0.000 [0.159 , 0.283]
## Robust - - 5.768 0.000 [0.140 , 0.284]
## =============================================================================txtStop()Code snippet 10: Using rdrobust on the collapsed data
\(~\)
txtStart("./outputs/Vol-2-R_lso3_rdrobust_collapsed.txt", commands=TRUE, results=TRUE, append=FALSE, visible.only=TRUE)## Output being copied to text file,
## use txtStop to enddata2 = data.frame(nextGPA_nonorm, X)
dim(data2)## [1] 44362 2collapsed = aggregate(nextGPA_nonorm ~ X, data = data2, mean)
dim(collapsed)## [1] 429 2out = rdrobust(collapsed$nextGPA_nonorm, collapsed$X)
summary(out)## Call: rdrobust
##
## Number of Obs. 429
## BW type mserd
## Kernel Triangular
## VCE method NN
##
## Number of Obs. 274 155
## Eff. Number of Obs. 51 50
## Order est. (p) 1 1
## Order bias (q) 2 2
## BW est. (h) 0.508 0.508
## BW bias (b) 0.805 0.805
## rho (h/b) 0.630 0.630
## Unique Obs. 274 155
##
## =============================================================================
## Method Coef. Std. Err. z P>|z| [ 95% C.I. ]
## =============================================================================
## Conventional 0.246 0.033 7.422 0.000 [0.181 , 0.310]
## Robust - - 6.083 0.000 [0.163 , 0.319]
## =============================================================================txtStop()Code snippet 11: Binomial test with rdwinselect
\(~\)
txtStart("./outputs/Vol-2-R_lso_falsification_binomial.txt", commands=TRUE, results=TRUE, append=FALSE, visible.only=TRUE)## Output being copied to text file,
## use txtStop to endout = rdwinselect(X, wmin = 0.01, nwindows = 1, cutoff = 5.00000000000e-06)## Mass points detected in running variable
##
##
## Window selection for RD under local randomization
##
## Number of obs = 44362
## Order of poly = 0
## Kernel type = uniform
## Reps = 1000
## Testing method = rdrandinf
## Balance test = diffmeans
##
## Cutoff c = 0.000 Left of c Right of c
## Number of obs 37211 7151
## 1st percentile 298 0
## 5th percentile 1817 269
## 10th percentile 3829 663
## 20th percentile 7588 1344
##
## ================================================================================
## Window length / 2 p-value Var. name Bin.test Obs<c Obs>=c
## ================================================================================
## 0.010 NA NA 0.000 228 77
## ================================================================================
## Note: no covariates specified.
## Need to specify covariates to find recommended length.txtStop()Binomial Test by Hand
\(~\)
txtStart("./outputs/Vol-2-R_lso_falsification_binomial_byhand.txt", commands=TRUE, results=TRUE, append=FALSE, visible.only=TRUE)## Output being copied to text file,
## use txtStop to endbinom.test(77, 305, 1/2)##
## Exact binomial test
##
## data: 77 and 305
## number of successes = 77, number of trials = 305, p-value < 2.2e-16
## alternative hypothesis: true probability of success is not equal to 0.5
## 95 percent confidence interval:
## 0.2046807 0.3051175
## sample estimates:
## probability of success
## 0.252459txtStop()Code snippet 13: Using rdrandinf on hsgrade_pct
\(~\)
txtStart("./outputs/Vol-2-R_lso_falsification_rdrandinf_hsgrade_pct.txt", commands=TRUE, results=TRUE, append=FALSE, visible.only=TRUE)## Output being copied to text file,
## use txtStop to endout = rdrandinf(data$hsgrade_pct, X, wl = -0.005, wr = 0.01, seed = 50)##
## Selected window = [-0.005;0.01]
##
## Running randomization-based test...
##
## Randomization-based test complete.
##
##
## Number of obs = 44362
## Order of poly = 0
## Kernel type = uniform
## Reps = 1000
## Window = set by user
## H0: tau = 0
## Randomization = fixed margins
##
## Cutoff c = 0.000 Left of c Right of c
## Number of obs 37211 7151
## Eff. number of obs 228 77
## Mean of outcome 29.118 32.675
## S.d. of outcome 21.737 21.625
## Window -0.005 0.010
##
## ================================================================================
## Finite sample Large sample
## ------------------ -----------------------------
## Statistic T P>|T| P>|T| Power vs d = 10.868
## ================================================================================
## Diff. in means 3.557 0.223 0.213 0.968
## ================================================================================txtStop()Code snippet 14: Using rdwinselect with covariates and windows with equal number of mass points to each side of the cutoff
\(~\)
txtStart("./outputs/Vol-2-R_lso_rdwinselect_consecutive_windows.txt", commands=TRUE, results=TRUE, append=FALSE, visible.only=TRUE)## Output being copied to text file,
## use txtStop to endZ = cbind(data$hsgrade_pct, data$totcredits_year1, data$age_at_entry, data$male,
data$bpl_north_america, data$english, data$loc_campus1, data$loc_campus2,
data$loc_campus3)
colnames(Z) = c("hsgrade_pct", "totcredits_year1", "age_at_entry", "male",
"bpl_north_america", "english", "loc_campus1", "loc_campus2",
"loc_campus3")
out = rdwinselect(X, Z, p = 1, seed = 50, wmin = 0.01, wstep = 0.01, cutoff = 5.00000000000e-06)## Mass points detected in running variable
##
##
## Window selection for RD under local randomization
##
## Number of obs = 44362
## Order of poly = 1
## Kernel type = uniform
## Reps = 1000
## Testing method = rdrandinf
## Balance test = diffmeans
##
## Cutoff c = 0.000 Left of c Right of c
## Number of obs 37211 7151
## 1st percentile 298 0
## 5th percentile 1817 269
## 10th percentile 3829 663
## 20th percentile 7588 1344
##
## ================================================================================
## Window length / 2 p-value Var. name Bin.test Obs<c Obs>=c
## ================================================================================
## 0.010 0.005 loc_campus3 0.000 228 77
## 0.020 0.000 totcredits_year 0.000 298 214
## 0.030 0.000 hsgrade_pct 0.000 374 269
## 0.040 0.000 loc_campus1 0.000 494 375
## 0.050 0.000 totcredits_year 0.000 636 418
## 0.060 0.000 totcredits_year 0.000 714 497
## 0.070 0.000 hsgrade_pct 0.000 807 663
## 0.080 0.000 totcredits_year 0.000 877 727
## 0.090 0.000 totcredits_year 0.000 1049 815
## 0.100 0.000 totcredits_year 0.001 1131 973
## ================================================================================
## Smallest window does not pass covariate test.
## Decrease smallest window or reduce level.txtStop()Code snippet 15: Using rdrandinf on the Outcome
\(~\)
txtStart("./outputs/Vol-2-R_lso3_rdrandinf_adhocsmall_p0.txt", commands=TRUE, results=TRUE, append=FALSE, visible.only=TRUE)## Output being copied to text file,
## use txtStop to endout = rdrandinf(nextGPA_nonorm, X, wl = -0.005, wr = 0.01, seed = 50)##
## Selected window = [-0.005;0.01]
##
## Running randomization-based test...
##
## Randomization-based test complete.
##
##
## Number of obs = 40582
## Order of poly = 0
## Kernel type = uniform
## Reps = 1000
## Window = set by user
## H0: tau = 0
## Randomization = fixed margins
##
## Cutoff c = 0.000 Left of c Right of c
## Number of obs 34854 5728
## Eff. number of obs 208 67
## Mean of outcome 1.830 2.063
## S.d. of outcome 0.868 0.846
## Window -0.005 0.010
##
## ================================================================================
## Finite sample Large sample
## ------------------ -----------------------------
## Statistic T P>|T| P>|T| Power vs d = 0.434
## ================================================================================
## Diff. in means 0.234 0.058 0.051 0.952
## ================================================================================txtStop()Reference
Greg Snow (2016). TeachingDemos: Demonstrations for Teaching and Learning. R package version 2.10. https://CRAN.R-project.org/package=TeachingDemos
H. Wickham. ggplot2: Elegant Graphics for Data Analysis. Springer-Verlag New York, 2016.
Jeffrey B. Arnold (2019). ggthemes: Extra Themes, Scales and Geoms for ‘ggplot2’. R package version 4.2.0. https://CRAN.R-project.org/package=ggthemes
Matias D. Cattaneo, Michael Jansson and Xinwei Ma (2019). lpdensity: Local Polynomial Density Estimation and Inference. R package version 1.0. https://CRAN.R-project.org/package=lpdensity
Matias D. Cattaneo, Michael Jansson and Xinwei Ma (2019). rddensity: Manipulation Testing Based on Density Discontinuity. R package version 1.0. https://CRAN.R-project.org/package=rddensity
Matias D. Cattaneo, Rocio Titiunik and Gonzalo Vazquez-Bare (2019). rdlocrand: Local Randomization Methods for RD Designs. R package version 0.5. https://CRAN.R-project.org/package=rdlocrand
R Core Team (2020). foreign: Read Data Stored by ‘Minitab’, ‘S’, ‘SAS’, ‘SPSS’, ‘Stata’, ‘Systat’, ‘Weka’, ‘dBase’, …. R package version 0.8-75. https://CRAN.R-project.org/package=foreign
Sebastian Calonico, Matias D. Cattaneo, Max H. Farrell and Rocio Titiunik (2020). rdrobust: Robust Data-Driven Statistical Inference in Regression-Discontinuity Designs. R package version 0.99.6. https://CRAN.R-project.org/package=rdrobust