An informal analysis conducted on 19 Feb 2024 of whether demographics vary by call attempt using Bihar V18 round.

library(tidyverse)
library(ggplot2)
library(lubridate)
library(janitor)
library(readr)
library(table1)
library(gtsummary)
library(nnet)

setwd(dirname(rstudioapi::getActiveDocumentContext()$path))
options(scipen = 999)

d1 <- read_csv("Nomination Survey V18_WIDE.csv")

d1 %>% tabyl(call_status)

 call_status     n    percent
           1  3001 0.13406299
           2  1188 0.05307125
           3  2393 0.10690194
           4 15004 0.67027027
           5   799 0.03569354

1 = picked up 2 = asked to call back later 3 = invalid/ wrong number 4 = rang, no answer 5 = DND, do not disturb

d1 %>% group_by(sno) %>% summarise(count= n())
d1 %>% group_by(sno, call_status) %>% summarise(count= n())
`summarise()` has grouped output by 'sno'. You can override using the `.groups` argument.
d1 <- d1 %>% group_by(sno) %>% mutate(count= n()) %>% ungroup()

d1 <- d1 %>% group_by(sno) %>% mutate(attempts = sum(call_status == 4, na.rm = TRUE)+ 1) %>% ungroup()

d1 %>% tabyl(attempts, count)
 attempts    1    2    3   4    5    6   7  8 9
        1 1176  794 1263 240   40    0   0  0 0
        2   23 1168  603 356  250   36   0  0 0
        3    0    4 1011 424  290   36   0  0 0
        4    0    0    3 916  955  192  14  8 0
        5    0    0    0   0 2480  426  21  0 0
        6    0    0    0   0 7400 1230 126 24 0
        7    0    0    0   0    0  666  98 16 0
        8    0    0    0   0    0    0  70  0 9
        9    0    0    0   0    0    0   0  8 9
d2 <- d1 %>% filter(consent == 1)
#glimpse(d2)
1
[1] 1
table1(~ age + factor(gender) + factor(caste) + factor(education) + factor(occupation) + factor(is_vhnd_RI) + factor(nominate) | factor(count), data = d1)
1
(N=1199)		 2
(N=1966)		 3
(N=2880)		 4
(N=1936)		 5
(N=11415)		 6
(N=2586)		 7
(N=329)		 8
(N=56)		 9
(N=18)		 Overall
(N=22385)
age
Mean (SD) 31.1 (9.67) 31.6 (10.1) 30.8 (9.72) 31.7 (9.63) 30.2 (8.33) 31.9 (10.1) 27.6 (5.77) 35.0 (NA) NA (NA) 31.2 (9.70)
Median [Min, Max] 28.0 [18.0, 73.0] 29.0 [18.0, 75.0] 28.0 [18.0, 70.0] 30.0 [18.0, 63.0] 28.0 [18.0, 69.0] 30.0 [18.0, 70.0] 25.0 [19.0, 35.0] 35.0 [35.0, 35.0] NA [NA, NA] 29.0 [18.0, 75.0]
Missing 291 (24.3%) 1199 (61.0%) 2380 (82.6%) 1624 (83.9%) 11221 (98.3%) 2523 (97.6%) 320 (97.3%) 55 (98.2%) 18 (100%) 19631 (87.7%)
factor(gender)
1 624 (52.0%) 545 (27.7%) 361 (12.5%) 208 (10.7%) 122 (1.1%) 49 (1.9%) 6 (1.8%) 1 (1.8%) 0 (0%) 1916 (8.6%)
2 280 (23.4%) 219 (11.1%) 135 (4.7%) 101 (5.2%) 71 (0.6%) 14 (0.5%) 3 (0.9%) 0 (0%) 0 (0%) 823 (3.7%)
Missing 295 (24.6%) 1202 (61.1%) 2384 (82.8%) 1627 (84.0%) 11222 (98.3%) 2523 (97.6%) 320 (97.3%) 55 (98.2%) 18 (100%) 19646 (87.8%)
factor(caste)
-9999 4 (0.3%) 1 (0.1%) 2 (0.1%) 0 (0%) 0 (0%) 1 (0.0%) 0 (0%) 0 (0%) 0 (0%) 8 (0.0%)
-99 0 (0%) 5 (0.3%) 2 (0.1%) 0 (0%) 0 (0%) 0 (0%) 0 (0%) 0 (0%) 0 (0%) 7 (0.0%)
1 199 (16.6%) 152 (7.7%) 114 (4.0%) 54 (2.8%) 42 (0.4%) 19 (0.7%) 1 (0.3%) 0 (0%) 0 (0%) 581 (2.6%)
2 568 (47.4%) 484 (24.6%) 306 (10.6%) 206 (10.6%) 113 (1.0%) 36 (1.4%) 8 (2.4%) 0 (0%) 0 (0%) 1721 (7.7%)
3 133 (11.1%) 122 (6.2%) 72 (2.5%) 50 (2.6%) 38 (0.3%) 7 (0.3%) 0 (0%) 1 (1.8%) 0 (0%) 423 (1.9%)
Missing 295 (24.6%) 1202 (61.1%) 2384 (82.8%) 1626 (84.0%) 11222 (98.3%) 2523 (97.6%) 320 (97.3%) 55 (98.2%) 18 (100%) 19645 (87.8%)
factor(education)
-99 0 (0%) 1 (0.1%) 1 (0.0%) 0 (0%) 0 (0%) 0 (0%) 0 (0%) 0 (0%) 0 (0%) 2 (0.0%)
0 83 (6.9%) 58 (3.0%) 42 (1.5%) 20 (1.0%) 17 (0.1%) 8 (0.3%) 0 (0%) 0 (0%) 0 (0%) 228 (1.0%)
1 129 (10.8%) 109 (5.5%) 62 (2.2%) 38 (2.0%) 29 (0.3%) 3 (0.1%) 0 (0%) 0 (0%) 0 (0%) 370 (1.7%)
2 349 (29.1%) 286 (14.5%) 201 (7.0%) 121 (6.3%) 74 (0.6%) 25 (1.0%) 4 (1.2%) 0 (0%) 0 (0%) 1060 (4.7%)
3 342 (28.5%) 308 (15.7%) 190 (6.6%) 131 (6.8%) 73 (0.6%) 27 (1.0%) 5 (1.5%) 1 (1.8%) 0 (0%) 1077 (4.8%)
Missing 296 (24.7%) 1204 (61.2%) 2384 (82.8%) 1626 (84.0%) 11222 (98.3%) 2523 (97.6%) 320 (97.3%) 55 (98.2%) 18 (100%) 19648 (87.8%)
factor(occupation)
-9999 2 (0.2%) 0 (0%) 1 (0.0%) 0 (0%) 0 (0%) 0 (0%) 0 (0%) 0 (0%) 0 (0%) 3 (0.0%)
-99 0 (0%) 2 (0.1%) 0 (0%) 0 (0%) 0 (0%) 0 (0%) 0 (0%) 0 (0%) 0 (0%) 2 (0.0%)
1 125 (10.4%) 88 (4.5%) 61 (2.1%) 32 (1.7%) 32 (0.3%) 10 (0.4%) 3 (0.9%) 1 (1.8%) 0 (0%) 352 (1.6%)
2 145 (12.1%) 116 (5.9%) 72 (2.5%) 48 (2.5%) 38 (0.3%) 12 (0.5%) 0 (0%) 0 (0%) 0 (0%) 431 (1.9%)
3 24 (2.0%) 16 (0.8%) 11 (0.4%) 9 (0.5%) 3 (0.0%) 3 (0.1%) 0 (0%) 0 (0%) 0 (0%) 66 (0.3%)
4 15 (1.3%) 11 (0.6%) 9 (0.3%) 2 (0.1%) 0 (0%) 0 (0%) 1 (0.3%) 0 (0%) 0 (0%) 38 (0.2%)
5 97 (8.1%) 75 (3.8%) 57 (2.0%) 29 (1.5%) 20 (0.2%) 5 (0.2%) 0 (0%) 0 (0%) 0 (0%) 283 (1.3%)
6 175 (14.6%) 140 (7.1%) 72 (2.5%) 47 (2.4%) 37 (0.3%) 10 (0.4%) 0 (0%) 0 (0%) 0 (0%) 481 (2.1%)
7 312 (26.0%) 295 (15.0%) 192 (6.7%) 117 (6.0%) 55 (0.5%) 22 (0.9%) 5 (1.5%) 0 (0%) 0 (0%) 998 (4.5%)
8 3 (0.3%) 0 (0%) 0 (0%) 0 (0%) 0 (0%) 0 (0%) 0 (0%) 0 (0%) 0 (0%) 3 (0.0%)
9 5 (0.4%) 19 (1.0%) 21 (0.7%) 25 (1.3%) 7 (0.1%) 1 (0.0%) 0 (0%) 0 (0%) 0 (0%) 78 (0.3%)
Missing 296 (24.7%) 1204 (61.2%) 2384 (82.8%) 1627 (84.0%) 11223 (98.3%) 2523 (97.6%) 320 (97.3%) 55 (98.2%) 18 (100%) 19650 (87.8%)
factor(is_vhnd_RI)
-9999 262 (21.9%) 210 (10.7%) 127 (4.4%) 73 (3.8%) 42 (0.4%) 27 (1.0%) 2 (0.6%) 0 (0%) 0 (0%) 743 (3.3%)
0 351 (29.3%) 319 (16.2%) 230 (8.0%) 139 (7.2%) 90 (0.8%) 19 (0.7%) 4 (1.2%) 1 (1.8%) 0 (0%) 1153 (5.2%)
1 291 (24.3%) 233 (11.9%) 140 (4.9%) 98 (5.1%) 62 (0.5%) 16 (0.6%) 3 (0.9%) 0 (0%) 0 (0%) 843 (3.8%)
Missing 295 (24.6%) 1204 (61.2%) 2383 (82.7%) 1626 (84.0%) 11221 (98.3%) 2524 (97.6%) 320 (97.3%) 55 (98.2%) 18 (100%) 19646 (87.8%)
factor(nominate)
-9999 3 (0.3%) 1 (0.1%) 2 (0.1%) 2 (0.1%) 0 (0%) 1 (0.0%) 0 (0%) 0 (0%) 0 (0%) 9 (0.0%)
0 435 (36.3%) 375 (19.1%) 242 (8.4%) 171 (8.8%) 97 (0.8%) 26 (1.0%) 5 (1.5%) 0 (0%) 0 (0%) 1351 (6.0%)
1 470 (39.2%) 391 (19.9%) 256 (8.9%) 139 (7.2%) 97 (0.8%) 36 (1.4%) 4 (1.2%) 1 (1.8%) 0 (0%) 1394 (6.2%)
Missing 291 (24.3%) 1199 (61.0%) 2380 (82.6%) 1624 (83.9%) 11221 (98.3%) 2523 (97.6%) 320 (97.3%) 55 (98.2%) 18 (100%) 19631 (87.7%) 
NA
d3 <- d1 %>% filter(consent == 1 & count <7) %>% 
  select(age, gender, caste, education, occupation, is_vhnd_RI, nominate, count, attempts) 

d3 <- d3 %>% mutate_at(vars(age, gender, caste, education, occupation, is_vhnd_RI, nominate), ~ replace(., . < 0, NA)) 

d3 %>%
  tbl_summary(by = count, 
              type = list(occupation ~ "categorical"),
              missing = "no", 
              statistic = list(
                all_categorical() ~ "{p}%", 
                all_continuous() ~ "{mean}, ({sd})"
              )) %>% 
  add_p()
There was an error in 'add_p()/add_difference()' for variable 'occupation', p-value omitted:
Error in stats::fisher.test(c(1, 2, 6, 7, 7, 5, 2, 7, 7, 7, 7, 7, 2, 5, : FEXACT error 5.
The hash table key cannot be computed because the largest key
is larger than the largest representable int.
The algorithm cannot proceed.
Reduce the workspace, consider using 'simulate.p.value=TRUE' or another algorithm.
There was an error in 'add_p()/add_difference()' for variable 'attempts', p-value omitted:
Error in stats::fisher.test(c(1, 1, 1, 1, 2, 1, 1, 1, 1, 1, 1, 1, 1, 1, : FEXACT error 5.
The hash table key cannot be computed because the largest key
is larger than the largest representable int.
The algorithm cannot proceed.
Reduce the workspace, consider using 'simulate.p.value=TRUE' or another algorithm.
Characteristic 1, N = 9091 2, N = 7741 3, N = 5051 4, N = 3141 5, N = 1941 6, N = 631 p-value2
age 31, (10) 32, (10) 31, (10) 32, (10) 30, (8) 32, (10) 0.3
gender 0.076
    1 69% 71% 73% 67% 63% 78%
    2 31% 29% 27% 33% 37% 22%
caste 0.3
    1 22% 20% 23% 17% 22% 31%
    2 63% 64% 62% 66% 59% 58%
    3 15% 16% 15% 16% 20% 11%
education 0.7
    0 9.2% 7.6% 8.5% 6.5% 8.8% 13%
    1 14% 14% 13% 12% 15% 4.8%
    2 39% 38% 41% 39% 38% 40%
    3 38% 40% 38% 42% 38% 43%
occupation
    1 14% 12% 12% 10% 17% 16%
    2 16% 15% 15% 16% 20% 19%
    3 2.7% 2.1% 2.2% 2.9% 1.6% 4.8%
    4 1.7% 1.4% 1.8% 0.6% 0% 0%
    5 11% 9.9% 12% 9.4% 10% 7.9%
    6 19% 18% 15% 15% 19% 16%
    7 35% 39% 39% 38% 29% 35%
    8 0.3% 0% 0% 0% 0% 0%
    9 0.6% 2.5% 4.2% 8.1% 3.6% 1.6%
is_vhnd_RI 45% 42% 38% 41% 41% 46% 0.3
nominate 52% 51% 51% 45% 50% 58% 0.3
attempts
    1 100% 41% 12% 1.3% 0% 0%
    2 0% 59% 34% 13% 1.5% 4.8%
    3 0% 0% 54% 29% 12% 7.9%
    4 0% 0% 0% 57% 36% 25%
    5 0% 0% 0% 0% 50% 27%
    6 0% 0% 0% 0% 0% 35%
1 Mean, (SD); %
2 Kruskal-Wallis rank sum test; Pearson’s Chi-squared test 
d3  %>%  
  tbl_summary(by = attempts, 
              type = list(occupation ~ "categorical"),
              missing = "no", 
              statistic = list(
                all_categorical() ~ "{p}%", 
                all_continuous() ~ "{mean}, ({sd})"
              )) %>% 
  add_p()
There was an error in 'add_p()/add_difference()' for variable 'caste', p-value omitted:
Error in stats::fisher.test(c(1, 2, 2, 2, 2, 3, 2, 2, 1, 2, 2, 2, 2, 2, : FEXACT error 6.  LDKEY=605 is too small for this problem,
  (ii := key2[itp=997] = 399847912, ldstp=18150)
Try increasing the size of the workspace and possibly 'mult'
There was an error in 'add_p()/add_difference()' for variable 'education', p-value omitted:
Error in stats::fisher.test(c(0, 2, 1, 2, 2, 3, 2, 3, 0, 1, 1, 2, 0, 2, : FEXACT error 501.
The hash table key cannot be computed because the largest key
is larger than the largest representable int.
The algorithm cannot proceed.
Reduce the workspace, consider using 'simulate.p.value=TRUE' or another algorithm.
There was an error in 'add_p()/add_difference()' for variable 'occupation', p-value omitted:
Error in stats::fisher.test(c(1, 2, 6, 7, 7, 5, 2, 7, 7, 7, 7, 7, 2, 5, : FEXACT error 5.
The hash table key cannot be computed because the largest key
is larger than the largest representable int.
The algorithm cannot proceed.
Reduce the workspace, consider using 'simulate.p.value=TRUE' or another algorithm.
Warning for variable 'is_vhnd_RI':
simpleWarning in stats::chisq.test(x = c(NA, 0, 0, 0, 0, 0, 0, 0, 1, 0, 0, NA, : Chi-squared approximation may be incorrect
There was an error in 'add_p()/add_difference()' for variable 'count', p-value omitted:
Error in stats::fisher.test(c(1L, 1L, 1L, 1L, 2L, 1L, 1L, 2L, 1L, 1L, : FEXACT error 5.
The hash table key cannot be computed because the largest key
is larger than the largest representable int.
The algorithm cannot proceed.
Reduce the workspace, consider using 'simulate.p.value=TRUE' or another algorithm.
Characteristic 1, N = 1,2941 2, N = 6711 3, N = 3941 4, N = 2641 5, N = 1141 6, N = 221 p-value2
age 32, (10) 31, (10) 31, (9) 30, (9) 31, (9) 30, (10) 0.4
gender 0.3
    1 72% 70% 68% 68% 63% 77%
    2 28% 30% 32% 32% 37% 23%
caste
    1 21% 21% 22% 20% 25% 38%
    2 63% 65% 61% 67% 54% 48%
    3 16% 14% 17% 13% 20% 14%
education
    0 8.5% 8.5% 7.4% 7.7% 11% 9.1%
    1 13% 17% 11% 13% 14% 0%
    2 39% 37% 44% 37% 35% 45%
    3 40% 38% 37% 43% 40% 45%
occupation
    1 13% 13% 13% 14% 11% 18%
    2 16% 13% 17% 16% 24% 23%
    3 2.5% 3.0% 1.5% 1.9% 2.6% 0%
    4 1.6% 1.2% 1.8% 0.4% 0% 0%
    5 11% 9.7% 10% 11% 6.1% 4.5%
    6 19% 17% 16% 14% 24% 23%
    7 37% 40% 35% 35% 27% 32%
    8 0.2% 0% 0% 0% 0% 0%
    9 0.5% 3.5% 6.2% 7.0% 5.3% 0%
is_vhnd_RI 45% 40% 39% 43% 36% 64% 0.2
nominate 53% 50% 47% 46% 50% 57% 0.10
count
    1 70% 0% 0% 0% 0% 0%
    2 25% 68% 0% 0% 0% 0%
    3 4.6% 26% 69% 0% 0% 0%
    4 0.3% 6.0% 23% 67% 0% 0%
    5 0% 0.4% 6.1% 27% 85% 0%
    6 0% 0.4% 1.3% 6.1% 15% 100%
1 Mean, (SD); %
2 Kruskal-Wallis rank sum test; Pearson’s Chi-squared test 
d3 %>% ggplot(aes(x=age, fill = factor(count))) + geom_density(alpha = 0.3) + theme_bw()

d3 %>% ggplot(aes(x=age, fill = factor(attempts))) + geom_density(alpha = 0.3) + theme_bw()



d3 %>% ggplot(aes(x= factor(count), fill= factor(occupation))) + geom_bar(position= "fill") + theme_bw() + coord_flip()


d3 %>% ggplot(aes(x= factor(count), fill= factor(caste))) + geom_bar(position= "fill") + theme_bw() + coord_flip()


d3 %>% ggplot(aes(x= factor(count), fill= factor(gender))) + geom_bar(position= "fill") + theme_bw() + coord_flip()



d3 %>% ggplot(aes(x= factor(attempts), fill= factor(occupation))) + geom_bar(position= "fill") + theme_bw() + coord_flip()


d3 %>% ggplot(aes(x= factor(attempts), fill= factor(caste))) + geom_bar(position= "fill") + theme_bw() + coord_flip()


d3 %>% ggplot(aes(x= factor(attempts), fill= factor(gender))) + geom_bar(position= "fill") + theme_bw() + coord_flip()




d3 %>% ggplot(aes(x= factor(attempts), fill= factor(nominate))) + geom_bar(position= "fill") + theme_bw() + coord_flip()




#geom_text(stat = "count", aes(label = paste0(round(100 * ..count../sum(..count..)), "%")),
            #position = position_fill(vjust = 0.5), color = "white")



 glm( nominate ~  attempts, family = binomial(), data = d3) %>% tbl_regression(exponentiate = TRUE)
Characteristic OR1 95% CI1 p-value
attempts 0.93 0.87, 0.99 0.022
1 OR = Odds Ratio, CI = Confidence Interval 
 glm( nominate ~  attempts, data = d3) %>% tbl_regression(exponentiate = TRUE)
Characteristic exp(Beta) 95% CI1 p-value
attempts 0.98 0.97, 1.00 0.022
1 CI = Confidence Interval 
 
 glm( is_vhnd_RI ~  attempts, family = binomial(), data = d3) %>% tbl_regression(exponentiate = TRUE)
Characteristic OR1 95% CI1 p-value
attempts 0.95 0.89, 1.03 0.2
1 OR = Odds Ratio, CI = Confidence Interval 
 


 multinom(caste ~ attempts, data = d3) %>% tbl_regression(exponentiate =TRUE )
# weights:  9 (4 variable)
initial  value 2982.732364 
iter  10 value 2469.258613
final  value 2469.258453 
converged
ℹ Multinomial models have a different underlying structure than the models
gtsummary was designed for. Other gtsummary functions designed to work with
tbl_regression objects may yield unexpected results.
Characteristic OR1 95% CI1 p-value
2
attempts 0.96 0.89, 1.04 0.3
3
attempts 0.98 0.88, 1.08 0.7
1 OR = Odds Ratio, CI = Confidence Interval 
 multinom(education ~ attempts, data = d3) %>% tbl_regression(exponentiate =TRUE )
# weights:  12 (6 variable)
initial  value 3777.652134 
iter  10 value 3305.330625
final  value 3305.330523 
converged
ℹ Multinomial models have a different underlying structure than the models
gtsummary was designed for. Other gtsummary functions designed to work with
tbl_regression objects may yield unexpected results.
Characteristic OR1 95% CI1 p-value
1
attempts 0.97 0.85, 1.11 0.7
2
attempts 1.01 0.90, 1.13 0.9
3
attempts 1.01 0.90, 1.14 0.8
1 OR = Odds Ratio, CI = Confidence Interval 
 multinom(occupation ~ attempts, data = d3) %>% tbl_regression(exponentiate =TRUE )
# weights:  27 (16 variable)
initial  value 5976.450850 
iter  10 value 4850.548500
iter  20 value 4667.603492
iter  30 value 4661.792595
iter  40 value 4661.286125
iter  50 value 4661.259834
iter  60 value 4661.161012
final  value 4661.160377 
converged
ℹ Multinomial models have a different underlying structure than the models
gtsummary was designed for. Other gtsummary functions designed to work with
tbl_regression objects may yield unexpected results.
Characteristic OR1 95% CI1 p-value
2
attempts 1.04 0.92, 1.16 0.5
3
attempts 0.89 0.71, 1.12 0.3
4
attempts 0.74 0.53, 1.03 0.075
5
attempts 0.91 0.80, 1.04 0.2
6
attempts 0.96 0.86, 1.07 0.5
7
attempts 0.95 0.86, 1.04 0.3
8
attempts 0.00 0.00, 0.00 <0.001
9
attempts 1.56 1.31, 1.85 <0.001
1 OR = Odds Ratio, CI = Confidence Interval 
  multinom(gender ~ attempts, data = d3) %>% tbl_regression(exponentiate =TRUE )
# weights:  3 (2 variable)
initial  value 1891.598656 
final  value 1666.193203 
converged
ℹ Multinomial models have a different underlying structure than the models
gtsummary was designed for. Other gtsummary functions designed to work with
tbl_regression objects may yield unexpected results.
Characteristic OR1 95% CI1 p-value
1
attempts 1.07 1.00, 1.14 0.047
1 OR = Odds Ratio, CI = Confidence Interval 
NA
NA

1 “Not in work force/no occupation” 2 “Agricultural / Non-agricultural labor” 3 “Government service/ Elected Official” 4 “Private Doctor, Lawyer, Accountant” 5 “Own business” 6 “Services/household and domestic” 7 “Manual - skilled and unskilled” 8 “Other” 9 “Homemaker” Looks like for occupation. An increase in odds for someone to be a homemaker. And to be “other”

1 “Male” 2 “Female” 888 “other” -99 “Refused to answer”.

Slightly more likely to be women the more call attempts.

Slightly less likely to give nominations based on more call attempts.

RUF survey

RUF_Survey_2021_WIDE (2)

b1 <- read_csv("RUF_Survey_2021_WIDE (2).csv")
Warning: One or more parsing issues, call `problems()` on your data frame for details, e.g.:
  dat <- vroom(...)
  problems(dat)Rows: 2248 Columns: 181── Column specification ──────────────────────────────────────────────────────────────
Delimiter: ","
chr  (45): SubmissionDate, starttime, endtime, deviceid, childuid_biharhealthregis...
dbl (129): subscriberid, simid, devicephonenum, duration, sno, contact_number, cal...
lgl   (7): sms_vaccine, days_since_sms, response_exotel, not_verify_respondent_oth...
ℹ Use `spec()` to retrieve the full column specification for this data.
ℹ Specify the column types or set `show_col_types = FALSE` to quiet this message.
b1 %>% tabyl(sno, call_status)

b1 %>% tabyl(call_status)

label define call_status 1 "Picked Up" 2 "Ringing but no answer" 3 "Cut after a few beeps" 4 "Number unreachable/switched off (Audio Voice Recording)" 5 "Invalid number (AVR)" 6 "Phone not in use (AVR)" 7 "Wrong number (AVR)" 8 "No incoming call facility on this number (AVR)" 9 "Respondent asked to call later" 10 "Do not disturb (DND)" 888 "other [specify]"
b1 %>% group_by(sno) %>% summarise(count= n())
b1 %>% group_by(sno, call_status) %>% summarise(count= n())
`summarise()` has grouped output by 'sno'. You can override using the `.groups` argument.
b1 <- b1 %>% group_by(sno) %>% mutate(count= n()) %>% ungroup()


b1 <- b1 %>% group_by(sno) %>% mutate(attempts = sum(call_status == 4, na.rm = TRUE)+ 1) %>% ungroup()

b1 %>% tabyl(attempts, count)
 attempts   1   2   3   4  5  6 7
        1 354 296 180 480 50  6 7
        2  15  64  87 140 25 12 0
        3   0   0  57  84 15  0 0
        4   0   0  30  84 25  6 0
        5   0   0   0 196  5  0 0
        6   0   0   0   0  0 12 0
        7   0   0   0   0  0 18 0
b2 <- b1 %>% filter(consent == 1)



b3 <- b2 %>% filter(count <7) %>% 
  select(is_mother, get_mother, relmother, same_hh, yourphone, literacy_resp, sms_rec, count, attempts) 

b3 <- b3 %>% mutate_at(vars(is_mother, get_mother, relmother, same_hh, yourphone, literacy_resp, sms_rec,  count, attempts), ~ replace(., . < 0, NA)) 


b3 %>%
  tbl_summary(by = count, 
              #type = list(occupation ~ "categorical"),
              missing = "no", 
              statistic = list(
                all_categorical() ~ "{p}%", 
                all_continuous() ~ "{mean}, ({sd})"
              )) %>% 
  add_p()
There was an error in 'add_p()/add_difference()' for variable 'yourphone', p-value omitted:
Error in stats::fisher.test(c(NA, NA, NA, NA, NA, NA, NA, NA, NA, NA, : 'x' and 'y' must have at least 2 levels
There was an error in 'add_p()/add_difference()' for variable 'attempts', p-value omitted:
Error in stats::fisher.test(c(1, 1, 1, 1, 1, 1, 2, 1, 1, 1, 1, 1, 1, 1, : FEXACT error 7(location). LDSTP=18630 is too small for this problem,
  (pastp=20.355, ipn_0:=ipoin[itp=161]=18629, stp[ipn_0]=22.3897).
Increase workspace or consider using 'simulate.p.value=TRUE'
Characteristic 1, N = 1241 2, N = 501 3, N = 291 4, N = 181 5, N = 61 6, N = 11 p-value2
is_mother 59% 62% 66% 83% 50% 0% 0.3
get_mother 98% 100% 100% 100% 67% 100% 0.2
relmother >0.9
    4 100% NA% NA% NA% 0% NA%
    5 0% NA% NA% NA% 100% NA%
same_hh 100% NA% NA% NA% 0% NA% >0.9
yourphone 0% NA% NA% NA% 0% NA%
literacy_resp 0.8
    0 25% 30% 34% 33% 17% 100%
    1 73% 68% 66% 67% 83% 0%
    2 2.4% 2.0% 0% 0% 0% 0%
sms_rec 0.6
    0 17% 26% 33% 29% 40% NA%
    1 63% 52% 58% 43% 40% NA%
    2 19% 22% 8.3% 14% 20% NA%
    888 1.6% 0% 0% 14% 0% NA%
attempts
    1 100% 80% 59% 78% 83% 0%
    2 0% 20% 28% 11% 0% 100%
    3 0% 0% 14% 5.6% 0% 0%
    4 0% 0% 0% 5.6% 17% 0%
1 %
2 Fisher’s exact test 
d3  %>%  
  tbl_summary(by = attempts, 
              type = list(occupation ~ "categorical"),
              missing = "no", 
              statistic = list(
                all_categorical() ~ "{p}%", 
                all_continuous() ~ "{mean}, ({sd})"
              )) %>% 
  add_p()
There was an error in 'add_p()/add_difference()' for variable 'caste', p-value omitted:
Error in stats::fisher.test(c(1, 2, 2, 2, 2, 3, 2, 2, 1, 2, 2, 2, 2, 2, : FEXACT error 6.  LDKEY=605 is too small for this problem,
  (ii := key2[itp=997] = 399847912, ldstp=18150)
Try increasing the size of the workspace and possibly 'mult'
There was an error in 'add_p()/add_difference()' for variable 'education', p-value omitted:
Error in stats::fisher.test(c(0, 2, 1, 2, 2, 3, 2, 3, 0, 1, 1, 2, 0, 2, : FEXACT error 501.
The hash table key cannot be computed because the largest key
is larger than the largest representable int.
The algorithm cannot proceed.
Reduce the workspace, consider using 'simulate.p.value=TRUE' or another algorithm.
There was an error in 'add_p()/add_difference()' for variable 'occupation', p-value omitted:
Error in stats::fisher.test(c(1, 2, 6, 7, 7, 5, 2, 7, 7, 7, 7, 7, 2, 5, : FEXACT error 5.
The hash table key cannot be computed because the largest key
is larger than the largest representable int.
The algorithm cannot proceed.
Reduce the workspace, consider using 'simulate.p.value=TRUE' or another algorithm.
Warning for variable 'is_vhnd_RI':
simpleWarning in stats::chisq.test(x = c(NA, 0, 0, 0, 0, 0, 0, 0, 1, 0, 0, NA, : Chi-squared approximation may be incorrect
There was an error in 'add_p()/add_difference()' for variable 'count', p-value omitted:
Error in stats::fisher.test(c(1L, 1L, 1L, 1L, 2L, 1L, 1L, 2L, 1L, 1L, : FEXACT error 5.
The hash table key cannot be computed because the largest key
is larger than the largest representable int.
The algorithm cannot proceed.
Reduce the workspace, consider using 'simulate.p.value=TRUE' or another algorithm.
Characteristic 1, N = 1,2941 2, N = 6711 3, N = 3941 4, N = 2641 5, N = 1141 6, N = 221 p-value2
age 32, (10) 31, (10) 31, (9) 30, (9) 31, (9) 30, (10) 0.4
gender 0.3
    1 72% 70% 68% 68% 63% 77%
    2 28% 30% 32% 32% 37% 23%
caste
    1 21% 21% 22% 20% 25% 38%
    2 63% 65% 61% 67% 54% 48%
    3 16% 14% 17% 13% 20% 14%
education
    0 8.5% 8.5% 7.4% 7.7% 11% 9.1%
    1 13% 17% 11% 13% 14% 0%
    2 39% 37% 44% 37% 35% 45%
    3 40% 38% 37% 43% 40% 45%
occupation
    1 13% 13% 13% 14% 11% 18%
    2 16% 13% 17% 16% 24% 23%
    3 2.5% 3.0% 1.5% 1.9% 2.6% 0%
    4 1.6% 1.2% 1.8% 0.4% 0% 0%
    5 11% 9.7% 10% 11% 6.1% 4.5%
    6 19% 17% 16% 14% 24% 23%
    7 37% 40% 35% 35% 27% 32%
    8 0.2% 0% 0% 0% 0% 0%
    9 0.5% 3.5% 6.2% 7.0% 5.3% 0%
is_vhnd_RI 45% 40% 39% 43% 36% 64% 0.2
nominate 53% 50% 47% 46% 50% 57% 0.10
count
    1 70% 0% 0% 0% 0% 0%
    2 25% 68% 0% 0% 0% 0%
    3 4.6% 26% 69% 0% 0% 0%
    4 0.3% 6.0% 23% 67% 0% 0%
    5 0% 0.4% 6.1% 27% 85% 0%
    6 0% 0.4% 1.3% 6.1% 15% 100%
1 Mean, (SD); %
2 Kruskal-Wallis rank sum test; Pearson’s Chi-squared test 
d3 %>% ggplot(aes(x=age, fill = factor(count))) + geom_density(alpha = 0.3) + theme_bw()

d3 %>% ggplot(aes(x=age, fill = factor(attempts))) + geom_density(alpha = 0.3) + theme_bw()



d3 %>% ggplot(aes(x= factor(count), fill= factor(occupation))) + geom_bar(position= "fill") + theme_bw() + coord_flip()

MAha RUF

m1 <- read_csv("Maha Child RUF Survey Round 2_WIDE.csv")
Rows: 2052 Columns: 117── Column specification ──────────────────────────────────────────────────────────────
Delimiter: ","
chr (35): SubmissionDate, starttime, endtime, deviceid, contact_name, mother_name,...
dbl (78): subscriberid, simid, devicephonenum, duration, sno, child_rchid, contact...
lgl  (4): sms_vaccine, days_since_sms, village_enrollee, response_exotel
ℹ Use `spec()` to retrieve the full column specification for this data.
ℹ Specify the column types or set `show_col_types = FALSE` to quiet this message.

m1 %>% tabyl(call_status)
 call_status   n    percent
           1 821 0.40009747
           2  35 0.01705653
           3 173 0.08430799
           4 902 0.43957115
           5 121 0.05896686

1 “Picked Up” 2 “Respondent asked to call later” 3 “Invalid number / Wrong number” 4 “Number unreachable/switched off/ Ringing” 5 “Do not disturb (DND)”

m1 %>% group_by(sno) %>% summarise(count= n())
m1 %>% group_by(sno, call_status) %>% summarise(count= n())
`summarise()` has grouped output by 'sno'. You can override using the `.groups` argument.
m1 <- m1 %>% group_by(sno) %>% mutate(count= n()) %>% ungroup()


m1 <- m1 %>% group_by(sno) %>% mutate(attempts = sum(call_status == 4, na.rm = TRUE)+ 1) %>% ungroup()

m1 %>% tabyl(attempts, count)
 attempts   1   2  3   4   5  6
        1 897  90  9   0   0  0
        2 173 158 18   4   0  0
        3   0 130 90   4   0  0
        4   0   0 42  92   0  0
        5   0   0  0 120  20  0
        6   0   0  0   0 175  6
        7   0   0  0   0   0 24
m2 <- m1 %>% filter(consent == 1)



m3 <- m2 %>% filter(count <5) %>% 
  select(is_mother,   whosephone, child_sex, verify_dob, vaccinated, age, education, caste, isuseful, counterfact, literacy_resp, sms_rec, count, attempts) 

m3 <- m3 %>% mutate_at(vars(is_mother,    whosephone, child_sex, verify_dob, vaccinated, age, education, caste, isuseful, counterfact, literacy_resp, sms_rec, count, attempts), ~ replace(., . < 0, NA)) %>% 
  mutate_at(vars(is_mother, whosephone, child_sex, verify_dob, vaccinated, age, education, caste, isuseful, counterfact, literacy_resp, sms_rec, count, attempts), ~ replace(., . > 100, NA)) 


m3 %>%
  tbl_summary(by = count, 
              type = list(whosephone ~ "categorical"),
              missing = "no", 
              statistic = list(
                all_categorical() ~ "{p}%", 
                all_continuous() ~ "{mean}, ({sd})"
              )) %>% 
  add_p()
There was an error in 'add_p()/add_difference()' for variable 'whosephone', p-value omitted:
Error in stats::fisher.test(c(1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 0, 1, 1, 0, : FEXACT[f3xact()] error: hash key 3e+09 > INT_MAX, kyy=549, it[i (= nco = 5)]= -2114443664.
Rather set 'simulate.p.value=TRUE'

There was an error in 'add_p()/add_difference()' for variable 'education', p-value omitted:
Error in stats::fisher.test(c(1, 3, 2, 3, 2, 3, 2, 2, 3, 1, 2, 2, 2, 2, : FEXACT error 6.  LDKEY=618 is too small for this problem,
  (ii := key2[itp=672] = 57907270, ldstp=18540)
Try increasing the size of the workspace and possibly 'mult'
There was an error in 'add_p()/add_difference()' for variable 'attempts', p-value omitted:
Error in stats::fisher.test(c(1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, : FEXACT error 6 (f5xact).  LDKEY=618 is too small for this problem: kval=10739405.
Try increasing the size of the workspace.
Characteristic 1, N = 5511 2, N = 951 3, N = 271 4, N = 181 p-value2
is_mother 32% 41% 44% 56% 0.035
whosephone
    0 47% 36% 50% 39%
    1 48% 54% 38% 61%
    2 0.5% 3.2% 0% 0%
    3 2.4% 1.1% 0% 0%
    4 1.1% 2.1% 7.7% 0%
    5 0.4% 1.1% 0% 0%
    6 0.2% 0% 0% 0%
    7 0.4% 1.1% 0% 0%
    8 0.2% 1.1% 3.8% 0%
    9 0% 1.1% 0% 0%
child_sex 0.8
    1 52% 56% 54% 44%
    2 48% 44% 46% 56%
verify_dob 84% 82% 89% 78% 0.7
vaccinated 93% 96% 96% 100% 0.8
age 27.7, (6.1) 27.5, (6.6) 26.6, (6.2) 26.5, (4.6) 0.4
education
    0 1.8% 2.3% 4.0% 0%
    1 8.6% 9.3% 4.0% 11%
    2 38% 43% 36% 22%
    3 52% 45% 56% 67%
caste 0.15
    1 29% 33% 39% 44%
    2 24% 25% 39% 11%
    3 47% 42% 22% 44%
isuseful 0.027
    1 56% 54% 38% 50%
    2 42% 41% 25% 50%
    3 1.1% 2.7% 25% 0%
    4 1.1% 2.7% 13% 0%
counterfact 0.2
    1 53% 70% 63% 50%
    2 34% 16% 25% 30%
    3 12% 14% 0% 20%
    4 1.1% 0% 13% 0%
literacy_resp 60% 60% 50% 50% >0.9
sms_rec 48% 35% 26% 56% 0.014
attempts
    1 100% 47% 15% 0%
    2 0% 53% 26% 5.6%
    3 0% 0% 59% 11%
    4 0% 0% 0% 83%
1 %; Mean, (SD)
2 Pearson’s Chi-squared test; Fisher’s exact test; Kruskal-Wallis rank sum test 
    label define sms_rec 1 "Yes" 0 "No"
    
caste 1 "Scheduled Caste/Scheduled Tribe /Nomadic Tribe(SC/ST/NT)" 2 "Backward Caste (OBC/BC/MBC)" 3 "Forward Caste (FC/OC)" 

isuseful 1 "Very helpful" 2 "helpful" 3 "not helpful" 4 "Not helpful at all"
label values isuseful isuseful

counterfact 1 "Yes definitely" 2 "Yes probably" 3 "No probably not" 4 "No definitely not"
m3 %>% ggplot(aes(x=age, fill = factor(count))) + geom_density(alpha = 0.3) + theme_bw()


m3 %>% ggplot(aes(x= factor(count), fill= factor(isuseful))) + geom_bar(position= "fill") + theme_bw() + coord_flip()


m3 %>% ggplot(aes(x= factor(count), fill= factor(sms_rec))) + geom_bar(position= "fill") + theme_bw() + coord_flip()


m3 %>% ggplot(aes(x= factor(count), fill= factor(education))) + geom_bar(position= "fill") + theme_bw() + coord_flip()


m3 %>% ggplot(aes(x= factor(count), fill= factor(is_mother))) + geom_bar(position= "fill") + theme_bw() + coord_flip()


m3 %>% ggplot(aes(x= factor(count), fill= factor(whosephone))) + geom_bar(position= "fill") + theme_bw() + coord_flip()


 glm( is_mother ~  attempts, family = binomial(), data = m3) %>% tbl_regression(exponentiate = TRUE)
Characteristic OR1 95% CI1 p-value
attempts 1.32 1.02, 1.71 0.034
1 OR = Odds Ratio, CI = Confidence Interval 
 glm( is_mother ~  attempts, data = m3) %>% tbl_regression()
Characteristic Beta 95% CI1 p-value
attempts 0.07 0.01, 0.13 0.031
1 CI = Confidence Interval 
NA
NA
 glm( counterfact ~  attempts,  data = m3) %>% tbl_regression()
Characteristic Beta 95% CI1 p-value
attempts -0.01 -0.15, 0.12 0.8
1 CI = Confidence Interval 
glm( isuseful ~  attempts,  data = m3) %>% tbl_regression()
Characteristic Beta 95% CI1 p-value
attempts 0.08 -0.03, 0.19 0.14
1 CI = Confidence Interval 
NA
NA
NA
 glm( sms_rec ~  attempts, family = binomial(), data = m3) %>% tbl_regression(exponentiate = TRUE)
Characteristic OR1 95% CI1 p-value
attempts 0.93 0.72, 1.20 0.6
1 OR = Odds Ratio, CI = Confidence Interval 

 multinom(caste ~ attempts, data = m3) %>% tbl_regression(exponentiate =TRUE )
# weights:  9 (4 variable)
initial  value 700.914640 
final  value 676.945299 
converged
ℹ Multinomial models have a different underlying structure than the models
gtsummary was designed for. Other gtsummary functions designed to work with
tbl_regression objects may yield unexpected results.
Characteristic OR1 95% CI1 p-value
2
attempts 0.75 0.52, 1.06 0.10
3
attempts 0.72 0.54, 0.97 0.033
1 OR = Odds Ratio, CI = Confidence Interval

BIhar, resources used



d1 %>% group_by(consent, count) %>% summarise(time = sum(duration))
`summarise()` has grouped output by 'consent'. You can override using the `.groups` argument.
d1 %>% group_by(sno) %>% mutate(count= n()) %>% ungroup()
NA
Sum of time Column Labels
Row Labels 1 2 3 4 5 6 7 8 9 Grand Total
Group1 2% 4% 5% 4% 23% 6% 1% 0% 0% 45%
1 20% 15% 10% 6% 3% 1% 0% 0% 0% 55%
Grand Total 22% 19% 15% 10% 27% 7% 1% 0% 0% 100%
Sum of time Column Labels
Row Labels 1 2 3 4 5 6 7 8 9 Grand Total
Group1 7% 19% 36% 41% 87% 83% 84% 84% 100% 45%
1 93% 81% 64% 59% 13% 17% 16% 16% 0% 55%
Grand Total 100% 100% 100% 100% 100% 100% 100% 100% 100% 100%
Sum of time Column Labels
Row Labels 1 2 3 4 5 6 7 8 9 Grand Total
Group1 4% 8% 12% 9% 53% 13% 2% 0% 0% 100%
1 37% 27% 17% 10% 6% 2% 0% 0% 0% 100%
Grand Total 22% 19% 15% 10% 27% 7% 1% 0% 0% 100% 

m1 <- m1 %>% group_by(sno) %>% mutate(attempts = sum(call_status == 4, na.rm = TRUE)+ 1) %>% ungroup()

m1 %>% tabyl(attempts, count)
 attempts   1   2  3   4   5  6
        1 897  90  9   0   0  0
        2 173 158 18   4   0  0
        3   0 130 90   4   0  0
        4   0   0 42  92   0  0
        5   0   0  0 120  20  0
        6   0   0  0   0 175  6
        7   0   0  0   0   0 24

  
d1 <- d1 %>% mutate(starttime = mdy_hms(starttime))

d1 <- d1 %>% group_by(sno) %>% mutate(minstart = min(starttime)) %>% ungroup()

d1 %>% mutate(firstcall = minstart==starttime)  %>% filter(firstcall == TRUE) %>% tabyl(call_status)
 call_status    n    percent
           1 1026 0.16041276
           2  419 0.06550969
           3  706 0.11038149
           4 3506 0.54815510
           5  739 0.11554096

Time of day


d1$starttime %>% hour() %>% tabyl()
  .    n      percent
  7    9 0.0004020549
  8  136 0.0060754970
  9 3462 0.1546571365
 10 2835 0.1266473085
 11 2587 0.1155684610
 12 3361 0.1501451865
 13  730 0.0326111235
 14 2824 0.1261559080
 15 2628 0.1174000447
 16 3360 0.1501005137
 17  235 0.0104981014
 18  104 0.0046459683
 19   79 0.0035291490
 20   23 0.0010274738
 21   12 0.0005360733
d1 <- d1 %>% mutate(
  hour = hour(starttime),
  minute = minute(starttime)
)

d1 <- d1 %>% mutate(
  time_number = round(hour +  minute/60,2))

tabyl(d1$minute)
 d1$minute   n    percent
         0 236 0.01054277
         1 266 0.01188296
         2 230 0.01027474
         3 264 0.01179361
         4 240 0.01072147
         5 250 0.01116819
         6 296 0.01322314
         7 325 0.01451865
         8 340 0.01518874
         9 357 0.01594818
        10 378 0.01688631
        11 406 0.01813715
        12 390 0.01742238
        13 404 0.01804780
        14 393 0.01755640
        15 417 0.01862855
        16 417 0.01862855
        17 426 0.01903060
        18 412 0.01840518
        19 414 0.01849453
        20 429 0.01916462
        21 429 0.01916462
        22 417 0.01862855
        23 408 0.01822649
        24 430 0.01920929
        25 432 0.01929864
        26 422 0.01885191
        27 403 0.01800313
        28 426 0.01903060
        29 410 0.01831584
        30 389 0.01737771
        31 413 0.01844985
        32 409 0.01827116
        33 368 0.01643958
        34 413 0.01844985
        35 388 0.01733304
        36 400 0.01786911
        37 395 0.01764574
        38 419 0.01871789
        39 421 0.01880724
        40 362 0.01617154
        41 405 0.01809247
        42 426 0.01903060
        43 422 0.01885191
        44 410 0.01831584
        45 402 0.01795845
        46 379 0.01693098
        47 407 0.01818182
        48 412 0.01840518
        49 399 0.01782444
        50 373 0.01666294
        51 359 0.01603753
        52 374 0.01670762
        53 362 0.01617154
        54 347 0.01550145
        55 325 0.01451865
        56 316 0.01411660
        57 280 0.01250838
        58 282 0.01259772
        59 261 0.01165959
library(ggridges)
Warning: package ‘ggridges’ was built under R version 4.2.2
d1 %>% filter(hour > 7 & hour <18) %>% ggplot(aes(x= time_number, y= fo_name, fill=fo_name)) + geom_density_ridges(alpha = 0.3, adjust= 0.5) + theme_bw()
Warning: Ignoring unknown parameters: `adjust`

d1 %>% filter(hour > 7 & hour <18) %>% ggplot(aes(x= time_number, fill= fo_name)) + geom_density(alpha = 0.3, adjust= 0.3) + theme_bw() +facet_wrap(vars(fo_name), ncol=3) + theme(legend.position = "none")


d4 <- d1 %>% filter(consent == 1) %>% filter(hour > 7 & hour <20) %>% select(hour, age, caste, gender, occupation, education ) %>% mutate_at(vars(hour, age, caste, gender, occupation, education), ~ replace(., . < 0, NA)) 





d4  %>% tabyl(hour)
 hour   n     percent
    8  27 0.009789703
    9 439 0.159173314
   10 390 0.141406817
   11 339 0.122915156
   12 362 0.131254532
   13  97 0.035170413
   14 381 0.138143582
   15 346 0.125453227
   16 318 0.115300943
   17  25 0.009064540
   18  15 0.005438724
   19  19 0.006889050
d4 %>%  ggplot(aes(x=age, fill = factor(hour))) + geom_density(alpha = 0.3) + theme_bw()


d4  %>%  ggplot(aes(x= factor(hour), fill= factor(gender))) + geom_bar(position= "fill") + theme_bw() + coord_flip()


d4  %>% ggplot(aes(x= factor(hour), fill= factor(caste))) + geom_bar(position= "fill") + theme_bw() + coord_flip()


d4  %>%  ggplot(aes(x= factor(hour), fill= factor(occupation))) + geom_bar(position= "fill") + theme_bw() + coord_flip()


d4 %>%  ggplot(aes(x= factor(hour), fill= factor(education))) + geom_bar(position= "fill") + theme_bw() + coord_flip()

NA
NA
NA
NA
NA

fisher_sim_p <- function(data, variable, by, ...) {
  result <- list()
  result$p <- stats::fisher.test(x = data %>% pull({{variable}}), y = data %>% pull({{by}}), simulate.p.value = T)$p.value
  result$test <- "Fisher's test with simulated p-value"
  result
}

d4  %>% 
    tbl_summary(by = hour, 
             # type = list(whosephone ~ "categorical"),
              missing = "no", 
              statistic = list(
                all_categorical() ~ "{p}%", 
                all_continuous() ~ "{mean}, ({sd})"
              )) %>% 
  add_p(test = list(
      c("caste", "education", "gender", "occupation") ~ "fisher_sim_p"   # can also use helper functions instead of character vector of columns - e.g., all_categorical() ~ ...
    )
  )
Characteristic 8, N = 271 9, N = 4391 10, N = 3901 11, N = 3391 12, N = 3621 13, N = 971 14, N = 3811 15, N = 3461 16, N = 3181 17, N = 251 18, N = 151 19, N = 191 p-value2
age 30, (9) 32, (10) 32, (10) 31, (10) 31, (10) 31, (9) 31, (9) 31, (10) 30, (9) 30, (7) 33, (10) 31, (9) >0.9
caste 0.5
    1 15% 25% 18% 23% 18% 27% 22% 20% 23% 13% 27% 26%
    2 73% 61% 68% 59% 66% 63% 61% 65% 62% 63% 53% 63%
    3 12% 14% 14% 18% 17% 10% 17% 15% 16% 25% 20% 11%
gender 0.14
    1 81% 70% 68% 69% 69% 80% 71% 65% 71% 76% 80% 89%
    2 19% 30% 32% 31% 31% 20% 29% 35% 29% 24% 20% 11%
occupation 0.068
    1 11% 15% 13% 14% 11% 9.4% 13% 12% 14% 8.0% 13% 11%
    2 3.7% 17% 17% 18% 16% 16% 12% 15% 17% 12% 20% 26%
    3 7.4% 5.0% 1.6% 0.9% 2.0% 1.0% 2.1% 2.9% 1.6% 8.0% 0% 0%
    4 0% 0.9% 1.0% 1.2% 1.1% 2.1% 3.2% 0.9% 1.6% 0% 0% 0%
    5 19% 7.3% 10% 12% 11% 7.3% 12% 10% 9.4% 8.0% 20% 5.3%
    6 11% 15% 16% 21% 17% 9.4% 19% 21% 18% 16% 0% 11%
    7 44% 36% 38% 32% 37% 51% 35% 35% 35% 48% 47% 47%
    8 0% 0% 0.5% 0.3% 0% 0% 0% 0% 0% 0% 0% 0%
    9 3.7% 3.4% 3.4% 0.6% 3.6% 4.2% 2.7% 2.9% 3.2% 0% 0% 0%
education 0.4
    0 11% 9.4% 6.5% 9.0% 8.7% 7.3% 7.5% 11% 6.4% 8.0% 13% 5.3%
    1 11% 12% 15% 14% 15% 16% 11% 15% 13% 0% 20% 16%
    2 26% 40% 42% 36% 35% 45% 38% 39% 38% 56% 33% 47%
    3 52% 38% 37% 41% 42% 32% 43% 35% 43% 36% 33% 32%
1 Mean, (SD); %
2 Kruskal-Wallis rank sum test; Fisher’s test with simulated p-value 
NA
NA

caste 1 “Scheduled Caste/Scheduled Tribe /Nomadic Tribe(SC/ST/NT)” 2 “Backward Caste (OBC/BC/MBC)” 3 “Forward Caste (FC/OC)”

1 “Not in work force/no occupation” 2 “Agricultural / Non-agricultural labor” 3 “Government service/ Elected Official” 4 “Private Doctor, Lawyer, Accountant” 5 “Own business” 6 “Services/household and domestic” 7 “Manual - skilled and unskilled” 8 “Other” 9 “Homemaker” Looks like for occupation. An increase in odds for someone to be a homemaker. And to be “other”

1 “Male” 2 “Female” 888 “other” -99 “Refused to answer”.

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