Factors of successful protection from pressure on business

car::vif(logit_1)

## GVIF Df GVIF^(1/(2*Df))
## macro_okved_code_group 2.557973 8 1.060458
## spark_stock_ticket 1.414953 1 1.189518
## category_by_size_missing 2.121544 3 1.133553
## administrative_position 2.036637 1 1.427108
## administrative_connections 1.568957 1 1.252580
## in_political_party 1.805759 1 1.343785
## in_association_or_sro 1.247275 1 1.116815
## case_publications 1.327666 1 1.152244
## criminal_prosecution 2.745054 1 1.656821
## capture 1.789498 1 1.337721
## corruption 1.241003 1 1.114003
## barriers 2.342071 1 1.530383
## have_court_case 2.395586 1 1.547768
## is_guilty 2.315013 1 1.521517
## cop_stage 1.478668 2 1.102726
## to_ombudsman 1.447562 1 1.203147

anova(logit_1, test="Chisq")

## Analysis of Deviance Table
##
## Model: binomial, link: logit
##
## Response: target_light_clear
##
## Terms added sequentially (first to last)
##
##
## Df Deviance Resid. Df Resid. Dev Pr(>Chi)
## NULL 323 448.85
## macro_okved_code_group 8 16.7573 315 432.09 0.03274 *
## spark_stock_ticket 1 0.1476 314 431.95 0.70081
## category_by_size_missing 3 9.7226 311 422.22 0.02108 *
## administrative_position 1 2.5703 310 419.65 0.10889
## administrative_connections 1 1.4585 309 418.19 0.22717
## in_political_party 1 0.7229 308 417.47 0.39519
## in_association_or_sro 1 4.6740 307 412.80 0.03062 *
## case_publications 1 4.9270 306 407.87 0.02644 *
## criminal_prosecution 1 0.0211 305 407.85 0.88452
## capture 1 5.7522 304 402.10 0.01647 *
## corruption 1 0.1625 303 401.93 0.68683
## barriers 1 0.1503 302 401.78 0.69827
## have_court_case 1 0.9809 301 400.80 0.32199
## is_guilty 1 0.1772 300 400.63 0.67381
## cop_stage 2 7.0978 298 393.53 0.02876 *
## to_ombudsman 1 2.5657 297 390.96 0.10920
## ---
## Signif. codes: 0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1

And hand removal of features:

light_clear_vars <-c(
c(
# "federal_districts",
# "largest_fed_districts",
#"macro_okved_code",
"macro_okved_code_group",
# "spark_web_site",
"spark_stock_ticket",
"administrative_position",
"administrative_connections",
#"in_political_party",
"in_association_or_sro",
"case_publications",
#"criminal_prosecution",
"capture", "corruption", "barriers",
#"have_court_case",
#"is_guilty",
#"max_bac_stage",
"cop_stage",
"to_ombudsman",
#"age_till_application_date",

#missing data
#"category_by_size_missing",
"category_by_size_melse",
#"category_by_size_2_cat",
#"auth_capital_group",

"target_light_clear" )
)


light_clear_data <-dataset[light_clear_vars]
light_clear_data <-light_clear_data[!is.na(light_clear_data$target_light_clear),]
light_clear_data <-light_clear_data[!is.na(light_clear_data$category_by_size_melse),]

light_clear_data$macro_okved_code_group <-factor(light_clear_data$macro_okved_code_group)
light_clear_data$cop_stage <-factor(light_clear_data$cop_stage)
logit_1<-glm(target_light_clear~., family = binomial,data = light_clear_data)
summary(logit_1)

##
## Call:
## glm(formula = target_light_clear ~ ., family = binomial, data = light_clear_data)
##
## Deviance Residuals:
## Min 1Q Median 3Q Max
## -1.912 -1.078 0.455 1.025 1.885
##
## Coefficients:
## Estimate Std. Error z value
## (Intercept) 1.08786 0.57154 1.903
## macro_okved_code_groupFinancial_insurance -0.42433 0.69318 -0.612
## macro_okved_code_groupmanufacturing 0.53058 0.43358 1.224
## macro_okved_code_groupother_categories 0.70314 0.45339 1.551
## macro_okved_code_groupreal_estate 1.66517 0.54751 3.041
## macro_okved_code_grouprural 0.96884 0.61205 1.583
## macro_okved_code_groupScience 1.58326 0.53004 2.987
## macro_okved_code_groupTrading 0.46256 0.43139 1.072
## macro_okved_code_groupTransportation 0.35589 0.64183 0.554
## spark_stock_ticket -1.20365 0.72018 -1.671
## administrative_position 0.89808 0.44792 2.005
## administrative_connections -0.44952 0.31667 -1.420
## in_association_or_sro 0.58687 0.28070 2.091
## case_publications -0.73970 0.38566 -1.918
## capture -0.81110 0.29780 -2.724
## corruption 0.21337 0.46218 0.462
## barriers -0.08009 0.41428 -0.193
## cop_stageInformation_collection -0.05519 0.38442 -0.144
## cop_stageResolution -0.86601 0.32208 -2.689
## to_ombudsman 0.65540 0.40944 1.601
## category_by_size_melseMicro -0.79328 0.28470 -2.786
## Pr(>|z|)
## (Intercept) 0.05699 .
## macro_okved_code_groupFinancial_insurance 0.54043
## macro_okved_code_groupmanufacturing 0.22106
## macro_okved_code_groupother_categories 0.12094
## macro_okved_code_groupreal_estate 0.00236 **
## macro_okved_code_grouprural 0.11344
## macro_okved_code_groupScience 0.00282 **
## macro_okved_code_groupTrading 0.28360
## macro_okved_code_groupTransportation 0.57924
## spark_stock_ticket 0.09466 .
## administrative_position 0.04497 *
## administrative_connections 0.15575
## in_association_or_sro 0.03655 *
## case_publications 0.05511 .
## capture 0.00646 **
## corruption 0.64432
## barriers 0.84671
## cop_stageInformation_collection 0.88583
## cop_stageResolution 0.00717 **
## to_ombudsman 0.10944
## category_by_size_melseMicro 0.00533 **
## ---
## Signif. codes: 0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1
##
## (Dispersion parameter for binomial family taken to be 1)
##
## Null deviance: 448.85 on 323 degrees of freedom
## Residual deviance: 393.46 on 303 degrees of freedom
## AIC: 435.46
##
## Number of Fisher Scoring iterations: 4

car::vif(logit_1)

## GVIF Df GVIF^(1/(2*Df))
## macro_okved_code_group 1.853034 8 1.039304
## spark_stock_ticket 1.156994 1 1.075636
## administrative_position 1.415450 1 1.189727
## administrative_connections 1.544208 1 1.242662
## in_association_or_sro 1.222597 1 1.105711
## case_publications 1.299073 1 1.139769
## capture 1.233162 1 1.110478
## corruption 1.077379 1 1.037969
## barriers 1.289705 1 1.135652
## cop_stage 1.379900 2 1.083832
## to_ombudsman 1.384966 1 1.176846
## category_by_size_melse 1.254458 1 1.120026

anova(logit_1, test="Chisq")

## Analysis of Deviance Table
##
## Model: binomial, link: logit
##
## Response: target_light_clear
##
## Terms added sequentially (first to last)
##
##
## Df Deviance Resid. Df Resid. Dev Pr(>Chi)
## NULL 323 448.85
## macro_okved_code_group 8 16.7573 315 432.09 0.032738 *
## spark_stock_ticket 1 0.1476 314 431.95 0.700807
## administrative_position 1 4.3378 313 427.61 0.037275 *
## administrative_connections 1 1.1043 312 426.50 0.293329
## in_association_or_sro 1 6.0639 311 420.44 0.013797 *
## case_publications 1 4.6309 310 415.81 0.031401 *
## capture 1 3.5487 309 412.26 0.059594 .
## corruption 1 0.4253 308 411.83 0.514315
## barriers 1 0.0047 307 411.83 0.945223
## cop_stage 2 8.1451 305 403.69 0.017034 *
## to_ombudsman 1 2.2527 304 401.43 0.133381
## category_by_size_melse 1 7.9714 303 393.46 0.004752 **
## ---
## Signif. codes: 0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1

By AIC maximization

light_clear_vars <-c(
c(
# "spark_web_site",
"spark_stock_ticket",
#"category_by_size_missing",
"category_by_size_melse",
#"category_by_size_2_cat",
#"administrative_position",
"administrative_connections",
"in_political_party",
"in_association_or_sro",
"case_publications",
"criminal_prosecution",
"capture",
"corruption",
"barriers",
"have_court_case",
"is_guilty",
#"reviewed_by_bac",
#"supported_by_bac_public_council",
#"max_bac_stage",
"cop_stage",
# "reaction_not_passed_by_applicant",
# "reaction_consultation",
# "reaction_target_letters_control",
# "reaction_not_passed_by_bac",
"to_ombudsman",
"macro_okved_code_group",
"target_light_clear")
)

light_clear_data <-dataset[light_clear_vars]
light_clear_data <-light_clear_data[!is.na(light_clear_data$target_light_clear),]
light_clear_data <-light_clear_data[!is.na(light_clear_data$category_by_size_melse),]

light_clear_data$macro_okved_code_group <-factor(light_clear_data$macro_okved_code_group)
light_clear_data$category_by_size <-factor(light_clear_data$category_by_size)
light_clear_data$cop_stage <-factor(light_clear_data$cop_stage)


logit_1<-glm(target_light_clear~., family = binomial,data = light_clear_data)
summary(logit_1)

##
## Call:
## glm(formula = target_light_clear ~ ., family = binomial, data = light_clear_data)
##
## Deviance Residuals:
## Min 1Q Median 3Q Max
## -1.9522 -1.0958 0.5036 1.0075 1.9987
##
## Coefficients: (1 not defined because of singularities)
## Estimate Std. Error z value
## (Intercept) 1.22600 0.71851 1.706
## spark_stock_ticket -1.08050 0.70714 -1.528
## category_by_size_melseMicro -0.88720 0.28619 -3.100
## administrative_connections -0.20977 0.29382 -0.714
## in_political_party 0.21547 0.39022 0.552
## in_association_or_sro 0.54852 0.27785 1.974
## case_publications -0.75929 0.38918 -1.951
## criminal_prosecution -0.11450 0.43443 -0.264
## capture -0.89510 0.35595 -2.515
## corruption 0.22136 0.49373 0.448
## barriers -0.18604 0.54835 -0.339
## have_court_case 0.44432 0.36743 1.209
## is_guilty -0.26840 0.37600 -0.714
## cop_stageInformation_collection -0.08523 0.39262 -0.217
## cop_stageResolution -0.85803 0.32460 -2.643
## to_ombudsman 0.65342 0.41341 1.581
## macro_okved_code_groupFinancial_insurance -0.44827 0.68757 -0.652
## macro_okved_code_groupmanufacturing 0.43664 0.44073 0.991
## macro_okved_code_groupother_categories 0.57604 0.45207 1.274
## macro_okved_code_groupreal_estate 1.60905 0.55049 2.923
## macro_okved_code_grouprural 0.88682 0.61787 1.435
## macro_okved_code_groupScience 1.52577 0.53037 2.877
## macro_okved_code_groupTrading 0.44702 0.43759 1.022
## macro_okved_code_groupTransportation 0.28016 0.64599 0.434
## category_by_sizeMicro NA NA NA
## Pr(>|z|)
## (Intercept) 0.08795 .
## spark_stock_ticket 0.12651
## category_by_size_melseMicro 0.00193 **
## administrative_connections 0.47527
## in_political_party 0.58082
## in_association_or_sro 0.04836 *
## case_publications 0.05106 .
## criminal_prosecution 0.79212
## capture 0.01191 *
## corruption 0.65390
## barriers 0.73441
## have_court_case 0.22655
## is_guilty 0.47534
## cop_stageInformation_collection 0.82815
## cop_stageResolution 0.00821 **
## to_ombudsman 0.11398
## macro_okved_code_groupFinancial_insurance 0.51442
## macro_okved_code_groupmanufacturing 0.32182
## macro_okved_code_groupother_categories 0.20258
## macro_okved_code_groupreal_estate 0.00347 **
## macro_okved_code_grouprural 0.15120
## macro_okved_code_groupScience 0.00402 **
## macro_okved_code_groupTrading 0.30700
## macro_okved_code_groupTransportation 0.66452
## category_by_sizeMicro NA
## ---
## Signif. codes: 0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1
##
## (Dispersion parameter for binomial family taken to be 1)
##
## Null deviance: 448.85 on 323 degrees of freedom
## Residual deviance: 395.71 on 300 degrees of freedom
## AIC: 443.71
##
## Number of Fisher Scoring iterations: 4

logit_2<-stepAIC(logit_1)

## Start: AIC=443.71
## target_light_clear ~ spark_stock_ticket + category_by_size_melse +
## administrative_connections + in_political_party + in_association_or_sro +
## case_publications + criminal_prosecution + capture + corruption +
## barriers + have_court_case + is_guilty + cop_stage + to_ombudsman +
## macro_okved_code_group + category_by_size
##
##
## Step: AIC=443.71
## target_light_clear ~ spark_stock_ticket + category_by_size_melse +
## administrative_connections + in_political_party + in_association_or_sro +
## case_publications + criminal_prosecution + capture + corruption +
## barriers + have_court_case + is_guilty + cop_stage + to_ombudsman +
## macro_okved_code_group
##
## Df Deviance AIC
## - criminal_prosecution 1 395.78 441.78
## - barriers 1 395.83 441.83
## - corruption 1 395.91 441.91
## - in_political_party 1 396.02 442.02
## - administrative_connections 1 396.22 442.22
## - is_guilty 1 396.22 442.22
## - have_court_case 1 397.19 443.19
## <none> 395.71 443.71
## - spark_stock_ticket 1 398.07 444.07
## - to_ombudsman 1 398.26 444.26
## - case_publications 1 399.61 445.61
## - in_association_or_sro 1 399.65 445.65
## - macro_okved_code_group 8 414.10 446.10
## - capture 1 402.32 448.32
## - cop_stage 2 405.06 449.06
## - category_by_size_melse 1 405.67 451.67
##
## Step: AIC=441.78
## target_light_clear ~ spark_stock_ticket + category_by_size_melse +
## administrative_connections + in_political_party + in_association_or_sro +
## case_publications + capture + corruption + barriers + have_court_case +
## is_guilty + cop_stage + to_ombudsman + macro_okved_code_group
##
## Df Deviance AIC
## - barriers 1 395.83 439.83
## - in_political_party 1 396.08 440.08
## - corruption 1 396.11 440.11
## - administrative_connections 1 396.29 440.29
## - is_guilty 1 396.31 440.31
## - have_court_case 1 397.30 441.30
## <none> 395.78 441.78
## - spark_stock_ticket 1 398.13 442.13
## - to_ombudsman 1 398.59 442.59
## - in_association_or_sro 1 399.68 443.68
## - case_publications 1 399.86 443.86
## - macro_okved_code_group 8 415.02 445.02
## - cop_stage 2 405.11 447.11
## - capture 1 404.01 448.01
## - category_by_size_melse 1 405.75 449.75
##
## Step: AIC=439.83
## target_light_clear ~ spark_stock_ticket + category_by_size_melse +
## administrative_connections + in_political_party + in_association_or_sro +
## case_publications + capture + corruption + have_court_case +
## is_guilty + cop_stage + to_ombudsman + macro_okved_code_group
##
## Df Deviance AIC
## - in_political_party 1 396.15 438.15
## - corruption 1 396.18 438.18
## - administrative_connections 1 396.36 438.36
## - is_guilty 1 396.39 438.39
## - have_court_case 1 397.43 439.43
## <none> 395.83 439.83
## - spark_stock_ticket 1 398.15 440.15
## - to_ombudsman 1 398.69 440.69
## - in_association_or_sro 1 399.77 441.77
## - case_publications 1 399.86 441.86
## - macro_okved_code_group 8 415.02 443.02
## - cop_stage 2 405.16 445.16
## - capture 1 404.16 446.16
## - category_by_size_melse 1 405.78 447.78
##
## Step: AIC=438.15
## target_light_clear ~ spark_stock_ticket + category_by_size_melse +
## administrative_connections + in_association_or_sro + case_publications +
## capture + corruption + have_court_case + is_guilty + cop_stage +
## to_ombudsman + macro_okved_code_group
##
## Df Deviance AIC
## - administrative_connections 1 396.47 436.47
## - corruption 1 396.54 436.54
## - is_guilty 1 396.68 436.68
## - have_court_case 1 397.70 437.70
## <none> 396.15 438.15
## - spark_stock_ticket 1 398.46 438.46
## - to_ombudsman 1 399.12 439.12
## - in_association_or_sro 1 400.08 440.08
## - case_publications 1 400.09 440.09
## - macro_okved_code_group 8 415.47 441.47
## - cop_stage 2 405.28 443.28
## - capture 1 404.47 444.47
## - category_by_size_melse 1 406.72 446.72
##
## Step: AIC=436.47
## target_light_clear ~ spark_stock_ticket + category_by_size_melse +
## in_association_or_sro + case_publications + capture + corruption +
## have_court_case + is_guilty + cop_stage + to_ombudsman +
## macro_okved_code_group
##
## Df Deviance AIC
## - corruption 1 396.81 434.81
## - is_guilty 1 396.94 434.94
## - have_court_case 1 397.95 435.95
## <none> 396.47 436.47
## - spark_stock_ticket 1 398.65 436.65
## - to_ombudsman 1 399.24 437.24
## - in_association_or_sro 1 400.20 438.20
## - case_publications 1 401.32 439.32
## - macro_okved_code_group 8 415.98 439.98
## - cop_stage 2 405.35 441.35
## - capture 1 404.55 442.55
## - category_by_size_melse 1 406.81 444.81
##
## Step: AIC=434.81
## target_light_clear ~ spark_stock_ticket + category_by_size_melse +
## in_association_or_sro + case_publications + capture + have_court_case +
## is_guilty + cop_stage + to_ombudsman + macro_okved_code_group
##
## Df Deviance AIC
## - is_guilty 1 397.25 433.25
## - have_court_case 1 398.19 434.19
## <none> 396.81 434.81
## - spark_stock_ticket 1 398.87 434.87
## - to_ombudsman 1 399.70 435.70
## - in_association_or_sro 1 400.47 436.47
## - case_publications 1 402.13 438.13
## - macro_okved_code_group 8 416.33 438.33
## - cop_stage 2 405.74 439.74
## - capture 1 404.67 440.67
## - category_by_size_melse 1 407.34 443.34
##
## Step: AIC=433.25
## target_light_clear ~ spark_stock_ticket + category_by_size_melse +
## in_association_or_sro + case_publications + capture + have_court_case +
## cop_stage + to_ombudsman + macro_okved_code_group
##
## Df Deviance AIC
## - have_court_case 1 398.26 432.26
## <none> 397.25 433.25
## - spark_stock_ticket 1 399.46 433.46
## - to_ombudsman 1 400.11 434.11
## - in_association_or_sro 1 400.85 434.85
## - case_publications 1 402.38 436.38
## - macro_okved_code_group 8 417.12 437.12
## - cop_stage 2 406.14 438.14
## - capture 1 405.50 439.50
## - category_by_size_melse 1 407.58 441.58
##
## Step: AIC=432.26
## target_light_clear ~ spark_stock_ticket + category_by_size_melse +
## in_association_or_sro + case_publications + capture + cop_stage +
## to_ombudsman + macro_okved_code_group
##
## Df Deviance AIC
## <none> 398.26 432.26
## - spark_stock_ticket 1 400.43 432.43
## - to_ombudsman 1 400.94 432.94
## - in_association_or_sro 1 401.79 433.79
## - case_publications 1 403.31 435.31
## - macro_okved_code_group 8 418.05 436.05
## - cop_stage 2 407.30 437.30
## - capture 1 406.25 438.25
## - category_by_size_melse 1 408.07 440.07

summary(logit_2)

##
## Call:
## glm(formula = target_light_clear ~ spark_stock_ticket + category_by_size_melse +
## in_association_or_sro + case_publications + capture + cop_stage +
## to_ombudsman + macro_okved_code_group, family = binomial,
## data = light_clear_data)
##
## Deviance Residuals:
## Min 1Q Median 3Q Max
## -1.895 -1.079 0.516 1.039 1.906
##
## Coefficients:
## Estimate Std. Error z value
## (Intercept) 1.19219 0.54942 2.170
## spark_stock_ticket -1.03508 0.70717 -1.464
## category_by_size_melseMicro -0.86030 0.27934 -3.080
## in_association_or_sro 0.51215 0.27397 1.869
## case_publications -0.81046 0.36657 -2.211
## capture -0.79577 0.28593 -2.783
## cop_stageInformation_collection -0.03539 0.37857 -0.093
## cop_stageResolution -0.81405 0.31659 -2.571
## to_ombudsman 0.61921 0.38153 1.623
## macro_okved_code_groupFinancial_insurance -0.39772 0.67519 -0.589
## macro_okved_code_groupmanufacturing 0.44885 0.42521 1.056
## macro_okved_code_groupother_categories 0.56455 0.44288 1.275
## macro_okved_code_groupreal_estate 1.65214 0.54465 3.033
## macro_okved_code_grouprural 0.97348 0.61176 1.591
## macro_okved_code_groupScience 1.53430 0.52313 2.933
## macro_okved_code_groupTrading 0.41920 0.42661 0.983
## macro_okved_code_groupTransportation 0.27912 0.63916 0.437
## Pr(>|z|)
## (Intercept) 0.03001 *
## spark_stock_ticket 0.14328
## category_by_size_melseMicro 0.00207 **
## in_association_or_sro 0.06157 .
## case_publications 0.02704 *
## capture 0.00538 **
## cop_stageInformation_collection 0.92553
## cop_stageResolution 0.01013 *
## to_ombudsman 0.10459
## macro_okved_code_groupFinancial_insurance 0.55583
## macro_okved_code_groupmanufacturing 0.29116
## macro_okved_code_groupother_categories 0.20240
## macro_okved_code_groupreal_estate 0.00242 **
## macro_okved_code_grouprural 0.11154
## macro_okved_code_groupScience 0.00336 **
## macro_okved_code_groupTrading 0.32579
## macro_okved_code_groupTransportation 0.66233
## ---
## Signif. codes: 0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1
##
## (Dispersion parameter for binomial family taken to be 1)
##
## Null deviance: 448.85 on 323 degrees of freedom
## Residual deviance: 398.26 on 307 degrees of freedom
## AIC: 432.26
##
## Number of Fisher Scoring iterations: 4

exp(logit_2$coefficients)

## (Intercept)
## 3.2942900
## spark_stock_ticket
## 0.3551982
## category_by_size_melseMicro
## 0.4230331
## in_association_or_sro
## 1.6688717
## case_publications
## 0.4446514
## capture
## 0.4512358
## cop_stageInformation_collection
## 0.9652325
## cop_stageResolution
## 0.4430602
## to_ombudsman
## 1.8574687
## macro_okved_code_groupFinancial_insurance
## 0.6718499
## macro_okved_code_groupmanufacturing
## 1.5665129
## macro_okved_code_groupother_categories
## 1.7586538
## macro_okved_code_groupreal_estate
## 5.2181086
## macro_okved_code_grouprural
## 2.6471417
## macro_okved_code_groupScience
## 4.6380801
## macro_okved_code_groupTrading
## 1.5207508
## macro_okved_code_groupTransportation
## 1.3219715

car::vif(logit_2)

## GVIF Df GVIF^(1/(2*Df))
## spark_stock_ticket 1.129585 1 1.062819
## category_by_size_melse 1.226221 1 1.107349
## in_association_or_sro 1.183847 1 1.088048
## case_publications 1.182098 1 1.087243
## capture 1.145953 1 1.070492
## cop_stage 1.342656 2 1.076443
## to_ombudsman 1.220107 1 1.104585
## macro_okved_code_group 1.651925 8 1.031869

anova(logit_2, test="Chisq")

## Analysis of Deviance Table
##
## Model: binomial, link: logit
##
## Response: target_light_clear
##
## Terms added sequentially (first to last)
##
##
## Df Deviance Resid. Df Resid. Dev Pr(>Chi)
## NULL 323 448.85
## spark_stock_ticket 1 0.1690 322 448.68 0.680998
## category_by_size_melse 1 7.0519 321 441.63 0.007918 **
## in_association_or_sro 1 3.3556 320 438.27 0.066979 .
## case_publications 1 4.2246 319 434.05 0.039842 *
## capture 1 5.1490 318 428.90 0.023259 *
## cop_stage 2 5.9033 316 423.00 0.052252 .
## to_ombudsman 1 4.9497 315 418.05 0.026095 *
## macro_okved_code_group 8 19.7868 307 398.26 0.011173 *
## ---
## Signif. codes: 0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1

TARGET 3 - TARGET LIGHT EXTENDED Baseline

light_extended_vars <-c(
c(
#"federal_districts",
#"largest_fed_districts",
#"macro_okved_code",
"macro_okved_code_group",
#"spark_web_site",
"spark_stock_ticket",
"category_by_size_missing",
#"category_by_size_melse",
# "category_by_size_2_cat",
"administrative_position",
"administrative_connections",
"in_political_party",
"in_association_or_sro",
"case_publications",
"criminal_prosecution",
"capture", "corruption", "barriers",
"have_court_case",
"is_guilty",
"reviewed_by_bac",
#"supported_by_bac_public_council",
# "max_bac_stage",
"cop_stage",
#"reaction_not_passed_by_applicant",
#"reaction_consultation",
#"reaction_target_letters_control",
#"reaction_not_passed_by_bac",
"to_ombudsman",
"target_light_extended")
)

light_extended_data <-dataset[light_extended_vars]
light_extended_data =light_extended_data[!is.na(light_extended_data$target_light_extended),]
light_extended_data =light_extended_data[!is.na(light_extended_data$category_by_size_missing),]

light_extended_data$macro_okved_code_group <-factor(light_extended_data$macro_okved_code_group)
light_extended_data$category_by_size_missing <-factor(light_extended_data$category_by_size_missing)
light_extended_data$cop_stage <-factor(light_extended_data$cop_stage)

logit_1<-glm(target_light_extended~., family = binomial,data = light_extended_data)
summary(logit_1)

##
## Call:
## glm(formula = target_light_extended ~ ., family = binomial, data = light_extended_data)
##
## Deviance Residuals:
## Min 1Q Median 3Q Max
## -2.0204 -1.0737 0.5114 0.9313 1.9580
##
## Coefficients:
## Estimate Std. Error z value
## (Intercept) 1.57462 0.85906 1.833
## macro_okved_code_groupFinancial_insurance -0.76958 0.71166 -1.081
## macro_okved_code_groupmanufacturing 0.41639 0.43846 0.950
## macro_okved_code_groupother_categories 0.53999 0.44913 1.202
## macro_okved_code_groupreal_estate 1.77777 0.53030 3.352
## macro_okved_code_grouprural 0.81475 0.59453 1.370
## macro_okved_code_groupScience 1.44886 0.51960 2.788
## macro_okved_code_groupTrading 0.40197 0.42315 0.950
## macro_okved_code_groupTransportation 0.19112 0.64108 0.298
## spark_stock_ticket -1.36013 0.79645 -1.708
## category_by_size_missingMedium 0.06309 0.66335 0.095
## category_by_size_missingMicro -1.01687 0.51089 -1.990
## category_by_size_missingSmall -0.12544 0.55397 -0.226
## administrative_position 0.81112 0.51771 1.567
## administrative_connections -0.26735 0.30032 -0.890
## in_political_party -0.15237 0.43370 -0.351
## in_association_or_sro 0.65803 0.27460 2.396
## case_publications -1.03627 0.35410 -2.926
## criminal_prosecution -0.10433 0.43183 -0.242
## capture -0.80989 0.34829 -2.325
## corruption 0.20013 0.48506 0.413
## barriers -0.14932 0.53783 -0.278
## have_court_case 0.51118 0.36328 1.407
## is_guilty -0.24348 0.36998 -0.658
## reviewed_by_bac -0.05802 0.53693 -0.108
## cop_stageInformation_collection -0.07223 0.54355 -0.133
## cop_stageResolution -0.78140 0.49569 -1.576
## to_ombudsman 0.64249 0.40398 1.590
## Pr(>|z|)
## (Intercept) 0.066810 .
## macro_okved_code_groupFinancial_insurance 0.279526
## macro_okved_code_groupmanufacturing 0.342284
## macro_okved_code_groupother_categories 0.229244
## macro_okved_code_groupreal_estate 0.000801 ***
## macro_okved_code_grouprural 0.170558
## macro_okved_code_groupScience 0.005297 **
## macro_okved_code_groupTrading 0.342137
## macro_okved_code_groupTransportation 0.765606
## spark_stock_ticket 0.087685 .
## category_by_size_missingMedium 0.924231
## category_by_size_missingMicro 0.046547 *
## category_by_size_missingSmall 0.820865
## administrative_position 0.117172
## administrative_connections 0.373353
## in_political_party 0.725340
## in_association_or_sro 0.016561 *
## case_publications 0.003428 **
## criminal_prosecution 0.809098
## capture 0.020056 *
## corruption 0.679905
## barriers 0.781301
## have_court_case 0.159389
## is_guilty 0.510478
## reviewed_by_bac 0.913956
## cop_stageInformation_collection 0.894278
## cop_stageResolution 0.114932
## to_ombudsman 0.111745
## ---
## Signif. codes: 0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1
##
## (Dispersion parameter for binomial family taken to be 1)
##
## Null deviance: 488.55 on 355 degrees of freedom
## Residual deviance: 419.08 on 328 degrees of freedom
## AIC: 475.08
##
## Number of Fisher Scoring iterations: 4

car::vif(logit_1)

## GVIF Df GVIF^(1/(2*Df))
## macro_okved_code_group 2.537750 8 1.059932
## spark_stock_ticket 1.406349 1 1.185896
## category_by_size_missing 2.058000 3 1.127823
## administrative_position 1.887402 1 1.373827
## administrative_connections 1.492303 1 1.221599
## in_political_party 1.684771 1 1.297987
## in_association_or_sro 1.234705 1 1.111173
## case_publications 1.236103 1 1.111802
## criminal_prosecution 2.808749 1 1.675932
## capture 1.805700 1 1.343763
## corruption 1.212152 1 1.100978
## barriers 2.311503 1 1.520363
## have_court_case 2.372303 1 1.540228
## is_guilty 2.316439 1 1.521985
## reviewed_by_bac 3.414270 1 1.847774
## cop_stage 3.742819 2 1.390912
## to_ombudsman 1.445193 1 1.202162