-
Notifications
You must be signed in to change notification settings - Fork 0
/
Copy path4.R
429 lines (332 loc) · 17.3 KB
/
4.R
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
89
90
91
92
93
94
95
96
97
98
99
100
101
102
103
104
105
106
107
108
109
110
111
112
113
114
115
116
117
118
119
120
121
122
123
124
125
126
127
128
129
130
131
132
133
134
135
136
137
138
139
140
141
142
143
144
145
146
147
148
149
150
151
152
153
154
155
156
157
158
159
160
161
162
163
164
165
166
167
168
169
170
171
172
173
174
175
176
177
178
179
180
181
182
183
184
185
186
187
188
189
190
191
192
193
194
195
196
197
198
199
200
201
202
203
204
205
206
207
208
209
210
211
212
213
214
215
216
217
218
219
220
221
222
223
224
225
226
227
228
229
230
231
232
233
234
235
236
237
238
239
240
241
242
243
244
245
246
247
248
249
250
251
252
253
254
255
256
257
258
259
260
261
262
263
264
265
266
267
268
269
270
271
272
273
274
275
276
277
278
279
280
281
282
283
284
285
286
287
288
289
290
291
292
293
294
295
296
297
298
299
300
301
302
303
304
305
306
307
308
309
310
311
312
313
314
315
316
317
318
319
320
321
322
323
324
325
326
327
328
329
330
331
332
333
334
335
336
337
338
339
340
341
342
343
344
345
346
347
348
349
350
351
352
353
354
355
356
357
358
359
360
361
362
363
364
365
366
367
368
369
370
371
372
373
374
375
376
377
378
379
380
381
382
383
384
385
386
387
388
389
390
391
392
393
394
395
396
397
398
399
400
401
402
403
404
405
406
407
408
409
410
411
412
413
414
415
416
417
418
419
420
421
422
423
424
425
426
427
428
429
#!/usr/bin/env Rscript
# clear workspace
rm(list=ls())
# Function to load packages
loadPkg=function(toLoad){
for(lib in toLoad){
if(! lib %in% installed.packages()[,1])
{ install.packages(lib, repos='http://cran.rstudio.com/') }
suppressMessages( library(lib, character.only=TRUE) ) }
}
# load libraries
packs=c('haven', 'tidyverse', 'RColorBrewer', 'stringr', 'hrbrthemes',
'wesanderson') #
loadPkg(packs)
# ----------------------------------------------------------------------------------------------------------------------
# victimization models
#-----------------------------------------------------------------------------------------------------------------------
# load the data
vic = read_dta('Data/victimization.dta')
voted = vic[complete.cases(vic$votedref), ]
# hold the results
results = list()
# set up processing
var = c('gender',
'age_group',
'adjincome',
'employed',
'language_tri_dum3',
'language_tri_dum1',
'ideology',
'educ_3',
'cat_origins',
'refvictim',
'refvictim_close',
'(Intercept)',
'N',
'll'
)
get_se = function(mod){
se = sqrt(diag(summary(mod)$cov.unscaled)*summary(mod)$dispersion)
tib = tibble(se, var=names(se))
return(tib)
}
create_coef_stack = function(mod){
coefs = mod$coefficients
N = length(mod$residuals)
ll = ll=round(logLik(mod))
combined = c(coefs, N, ll)
names(combined) = c(names(coefs), c('N', 'll'))
tib = tibble(combined, var=names(combined))
return(tib)
}
create_table = function(mod1, mod2, subset){
dv = all.vars(mod1$formula)[1]
coefs1 = create_coef_stack(mod1)
names(coefs1) = c(dv, 'var')
coefs2 = create_coef_stack(mod2)
names(coefs2) = c(paste0(dv, '_1'), 'var')
se1 = get_se(mod1)
names(se1) = c('SE_1', 'var')
se2 = get_se(mod2)
names(se2) = c('SE_2', 'var')
tt = tibble('var'=var)
tt = left_join(x=tt, y=coefs1, by='var')
tt = left_join(x=tt, y=coefs2, by='var')
tt = left_join(x=tt, y=se1, by='var')
tt = left_join(x=tt, y=se2, by='var')
tt$subset = subset
tt$var = dplyr::recode(tt$var, '(Intercept)'='_cons')
return(tt)
}
# ----------------------------------------------------------------------------
# regressions for people who VOTED in the 1-0 referendum
# ----------------------------------------------------------------------------
# -----------------------------
# DV: they would vote YES in a potential legal/agreed referendum
# -----------------------------
m1_form = as.formula(voteYES_agreedref ~ gender + age_group + adjincome +
employed + language_tri_dum3 + language_tri_dum1 +
ideology + educ_3 + cat_origins + refvictim)
M1 <- glm(m1_form, data=voted[voted$votedref==1, ], family="binomial")
m2_form = as.formula(voteYES_agreedref ~ gender + age_group + adjincome +
employed + language_tri_dum3 + language_tri_dum1 +
ideology + educ_3 + cat_origins + refvictim_close)
M2 <- glm(m2_form, data=voted[voted$votedref==1, ], family="binomial")
results[[1]] = create_table(M1, M2, subset="voted (N = 1,423)")
# logit voteYES_agreedref gender age_group adjincome employed language_tri_dum3 language_tri_dum1 ideology educ_3 cat_origins refvictim if votedref==1, rob
# est store M1
# logit voteYES_agreedref gender age_group adjincome employed language_tri_dum3 language_tri_dum1 ideology educ_3 cat_origins refvictim_close if votedref==1, rob
# est store M2
# esttab M1 M2 using Table1.tex, b(a2) se(a2) scalars(ll) star(+ 0.10 * 0.05 ** 0.01) nogaps replace
# -----------------------------
# DV: participation in protests
# -----------------------------
m3_form = as.formula(partprotest ~ gender + age_group + adjincome +
employed + language_tri_dum3 + language_tri_dum1 +
ideology + educ_3 + cat_origins + refvictim)
M3 <- glm(m3_form, data=voted[voted$votedref==1, ], family="binomial")
m4_form = as.formula(partprotest ~ gender + age_group + adjincome +
employed + language_tri_dum3 + language_tri_dum1 +
ideology + educ_3 + cat_origins + refvictim_close)
M4 <- glm(m4_form, data=voted[voted$votedref==1, ], family="binomial")
results[[2]] = create_table(M3, M4, subset="voted (N = 1,423)")
# logit partprotest gender age_group adjincome employed language_tri_dum3 language_tri_dum1 ideology educ_3 cat_origins refvictim if votedref==1, rob
# est store M1
# logit partprotest gender age_group adjincome employed language_tri_dum3 language_tri_dum1 ideology educ_3 cat_origins refvictim_close if votedref==1, rob
# est store M2
# esttab M1 M2 using Table2.tex, b(a2) se(a2) scalars(ll) star(+ 0.10 * 0.05 ** 0.01) nogaps replace
# ----------------------------------------------------------------------------
# regressions for people who VOTED and voted YES in the 1-0 referendum
# ----------------------------------------------------------------------------
# -----------------------------
# DV: they would vote YES in a potential legal/agreed referendum
# -----------------------------
m5_form = as.formula(voteYES_agreedref ~ gender + age_group + adjincome +
employed + language_tri_dum3 + language_tri_dum1 +
ideology + educ_3 + cat_origins + refvictim)
M5 <- glm(m5_form, data=voted[voted$votedyes==1, ], family="binomial")
m6_form = as.formula(voteYES_agreedref ~ gender + age_group + adjincome +
employed + language_tri_dum3 + language_tri_dum1 +
ideology + educ_3 + cat_origins + refvictim_close)
M6 <- glm(m6_form, data=voted[voted$votedyes==1, ], family="binomial")
results[[3]] = create_table(M5, M6, subset="voted yes (N = 1,221)")
# logit voteYES_agreedref gender age_group adjincome employed language_tri_dum3 language_tri_dum1 ideology educ_3 cat_origins refvictim if votedyes==1, rob
# est store M1
# logit voteYES_agreedref gender age_group adjincome employed language_tri_dum3 language_tri_dum1 ideology educ_3 cat_origins refvictim_close if votedyes==1, rob
# est store M2
# esttab M1 M2 using Table3.tex, b(a2) se(a2) scalars(ll) star(+ 0.10 * 0.05 ** 0.01) nogaps replace
# -----------------------------
# DV: participation in protests
# -----------------------------
m7_form = as.formula(partprotest ~ gender + age_group + adjincome +
employed + language_tri_dum3 + language_tri_dum1 +
ideology + educ_3 + cat_origins + refvictim)
M7 <- glm(m7_form, data=voted[voted$votedyes==1, ], family="binomial")
m8_form = as.formula(partprotest ~ gender + age_group + adjincome +
employed + language_tri_dum3 + language_tri_dum1 +
ideology + educ_3 + cat_origins + refvictim_close)
M8 <- glm(m8_form, data=voted[voted$votedyes==1, ], family="binomial")
results[[4]] = create_table(M7, M8, subset="voted yes (N = 1,221)")
# logit partprotest gender age_group adjincome employed language_tri_dum3 language_tri_dum1 ideology educ_3 cat_origins refvictim if votedyes==1, rob
# est store M1
# logit partprotest gender age_group adjincome employed language_tri_dum3 language_tri_dum1 ideology educ_3 cat_origins refvictim_close if votedyes==1, rob
# est store M2
# esttab M1 M2 using Table4.tex, b(a2) se(a2) scalars(ll) star(+ 0.10 * 0.05 ** 0.01) nogaps replace
# ----------------------------------------------------------------------------
# regressions for people who WOULD HAVE VOTED in the 1-0 referendum:
# this includes people who voted AND people who did not vote because of logistics
# or because the police impeded them to vote (thus, who were willing to vote but they could not)
# ----------------------------------------------------------------------------
voted$reasons_novote = replace_na(voted$reasons_novote, -999)
voted$wvotedref = voted$votedref
voted[voted$reasons_novote==4, 'wvotedref'] = 1
voted[voted$reasons_novote==5, 'wvotedref'] = 1
# /*gen wvotedref = votedref
# replace wvotedref = 1 if reasons_novote == 4
# replace wvotedref = 1 if reasons_novote == 5
# replace votedyes =. if votedref==.
# */
# -----------------------------
# DV: they would vote YES in a potential legal/agreed referendum
# -----------------------------
m9_form = as.formula(voteYES_agreedref ~ gender + age_group + adjincome +
employed + language_tri_dum3 + language_tri_dum1 +
ideology + educ_3 + cat_origins + refvictim)
M9 <- glm(m9_form, data=voted[voted$wvotedref==1, ], family="binomial")
m10_form = as.formula(voteYES_agreedref ~ gender + age_group + adjincome +
employed + language_tri_dum3 + language_tri_dum1 +
ideology + educ_3 + cat_origins + refvictim_close)
M10 <- glm(m10_form, data=voted[voted$wvotedref==1, ], family="binomial")
results[[5]] = create_table(M9, M10, subset="would have voted (N = 1,574)")
# logit voteYES_agreedref gender age_group adjincome employed language_tri_dum3 language_tri_dum1 ideology educ_3 cat_origins refvictim if wvotedref==1, rob
# est store M1
# logit voteYES_agreedref gender age_group adjincome employed language_tri_dum3 language_tri_dum1 ideology educ_3 cat_origins refvictim_close if wvotedref==1, rob
# est store M2
# esttab M1 M2 using Table5.tex, b(a2) se(a2) scalars(ll) star(+ 0.10 * 0.05 ** 0.01) nogaps replace
# -----------------------------
# DV: participation in protests
# -----------------------------
m11_form = as.formula(partprotest ~ gender + age_group + adjincome +
employed + language_tri_dum3 + language_tri_dum1 +
ideology + educ_3 + cat_origins + refvictim)
M11 <- glm(m11_form, data=voted[voted$wvotedref==1, ], family="binomial")
m12_form = as.formula(partprotest ~ gender + age_group + adjincome +
employed + language_tri_dum3 + language_tri_dum1 +
ideology + educ_3 + cat_origins + refvictim_close)
M12 <- glm(m12_form, data=voted[voted$wvotedref==1, ], family="binomial")
results[[6]] = create_table(M11, M12, subset="would have voted (N = 1,574)")
# logit partprotest gender age_group adjincome employed language_tri_dum3 language_tri_dum1 ideology educ_3 cat_origins refvictim if wvotedref==1, rob
# est store M1
# logit partprotest gender age_group adjincome employed language_tri_dum3 language_tri_dum1 ideology educ_3 cat_origins refvictim_close if wvotedref==1, rob
# est store M2
# esttab M1 M2 using Table6.tex, b(a2) se(a2) scalars(ll) star(+ 0.10 * 0.05 ** 0.01) nogaps replace
# ----------------------------------------------------------------------------
# regressions for people who did NOT VOTE in the 1-0 referendum
# ----------------------------------------------------------------------------
# -----------------------------
# DV: they would vote YES in a potential legal/agreed referendum
# -----------------------------
m13_form = as.formula(voteYES_agreedref ~ gender + age_group + adjincome +
employed + language_tri_dum3 + language_tri_dum1 +
ideology + educ_3 + cat_origins + refvictim)
M13 <- glm(m13_form, data=voted[voted$votedref==0, ], family="binomial")
m14_form = as.formula(voteYES_agreedref ~ gender + age_group + adjincome +
employed + language_tri_dum3 + language_tri_dum1 +
ideology + educ_3 + cat_origins + refvictim_close)
M14 <- glm(m14_form, data=voted[voted$votedref==0, ], family="binomial")
results[[7]] = create_table(M13, M14, subset="did not vote (N = 995)")
# logit voteYES_agreedref gender age_group adjincome employed language_tri_dum3 language_tri_dum1 ideology educ_3 cat_origins refvictim if votedref==0, rob
# est store M1
# logit voteYES_agreedref gender age_group adjincome employed language_tri_dum3 language_tri_dum1 ideology educ_3 cat_origins refvictim_close if votedref==0, rob
# est store M2
# esttab M1 M2 using Table7.tex, b(a2) se(a2) scalars(ll) star(+ 0.10 * 0.05 ** 0.01) nogaps replace
# -----------------------------
# DV: participation in protests
# -----------------------------
m15_form = as.formula(partprotest ~ gender + age_group + adjincome +
employed + language_tri_dum3 + language_tri_dum1 +
ideology + educ_3 + cat_origins + refvictim)
M15 <- glm(m15_form, data=voted[voted$votedref==0, ], family="binomial")
m16_form = as.formula(partprotest ~ gender + age_group + adjincome +
employed + language_tri_dum3 + language_tri_dum1 +
ideology + educ_3 + cat_origins + refvictim_close)
M16 <- glm(m16_form, data=voted[voted$votedref==0, ], family="binomial")
results[[8]] = create_table(M15, M16, subset="did not vote (N = 995)")
# logit partprotest gender age_group adjincome employed language_tri_dum3 language_tri_dum1 ideology educ_3 cat_origins refvictim if votedref==0, rob
# est store M1
# logit partprotest gender age_group adjincome employed language_tri_dum3 language_tri_dum1 ideology educ_3 cat_origins refvictim_close if votedref==0, rob
# est store M2
# esttab M1 M2 using Table8.tex, b(a2) se(a2) scalars(ll) star(+ 0.10 * 0.05 ** 0.01) nogaps replace
#-----------------------------------------------------------------------------------------------------------------------
# plot
#-----------------------------------------------------------------------------------------------------------------------
# tidy results df
tidy_list = function(frame)
{
# victim table
vict_df = frame[,c(1,2,4, 6)]
colnames(vict_df) <- c('var', 'est', 'se', 'subset')
vict_df$outcome = colnames(frame)[2]
vict_df$type = 'Victim (Any)'
vict_df = na.omit(vict_df)
# close victim table
close_vict_df = frame[,c(1,3,5, 6)]
colnames(close_vict_df) <- c('var', 'est', 'se', 'subset')
close_vict_df$outcome = colnames(frame)[2]
close_vict_df$type = 'Victim (Close)'
close_vict_df = na.omit(close_vict_df)
# combine tidy output
pDat = rbind(vict_df, close_vict_df)
return(pDat)
}
# tidy results
tidy_results = map(results, tidy_list)
# combine lists into one tidy df
tidy_frame = do.call(rbind, tidy_results)
# make 95\% CI
tidy_frame$lo95 = tidy_frame$est - 1.96*tidy_frame$se
tidy_frame$hi95 = tidy_frame$est + 1.96*tidy_frame$se
# labels dictionary
labels_dict = data.frame(var = c('gender',
'age_group',
'adjincome',
'employed',
'language_tri_dum3',
'language_tri_dum1',
'ideology',
'educ_3',
'cat_origins',
'refvictim',
'refvictim_close',
'_cons',
'N',
'll'
),
label = c('Female',
'Age',
'HH Income',
'Employed',
'Bilingual',
'Catalan Speaker',
'Ideology',
'Education',
'Catalan Family',
'Victim (any)',
'Victim (close)',
'Constant',
'N',
'LL')
)
outcomes_dict = data.frame(outcome = c('voteYES_agreedref',
'indyaxis',
'partprotest',
'repressionfear_bi'
),
outcome_lab = c('DV: Would Vote Yes\n in Legal Ref.',
'DV: Support Level\n for Independence',
'DV: Participated in Protest',
'DV: Fear Repression'
)
)
# get rid of intercept and diag stats
pDat =
tidy_frame %>%
filter(!var %in% c('N', 'll', '_cons')) %>%
left_join(labels_dict) %>%
left_join(outcomes_dict)
# highlight victim vars
pDat$hilite = ifelse(str_detect(pDat$label, 'Victim'), T, F)
# reorder outcomes
pDat$outcome_lab = forcats::fct_rev(pDat$outcome_lab)
# fix victims plot
pDat = filter(pDat, hilite == TRUE, outcome != 'indyaxis')
pDat = filter(pDat, hilite == TRUE, outcome != 'repressionfear_bi')
ggplot(pDat) +
geom_hline(yintercept = 0, colour = gray(1/2), lty = 2) +
geom_pointrange(aes(x = type, y = est,
shape = subset,
ymin = lo95,
ymax = hi95),
position = position_dodge(width = .5),
size = .8) +
facet_grid(~outcome_lab) +
coord_flip() +
theme_ipsum_ps(grid = 'X') +
labs(x = 'Victimization Type',
y = 'Coefficient Estimate',
shape = 'Sample Subset:') +
theme(legend.position = 'bottom', axis.title.y = element_text(hjust = .5)) +
scale_color_manual(values = wes_palettes$Darjeeling1) +
guides(shape = guide_legend(nrow = 2, byrow = T))
ggsave('Figures/victim_coefplot.pdf', device = cairo_pdf)