Civil Cases of New Hampshire Businesses
Belknap, Merrimack and Hillsborough County Superior Court
Daniel Lee
November 8, 2019
Import data
- The data includes 667 civil cases in Belknap County, BCDD in Concord, and Hillsborough North.
- A few cases have more than one case type. The data was manipulated so that one row has only one case type. As a result, a case can show up more than once when it has more than once case type.
library(tidyverse)
# Import data
DIR <- "C:/Users/sclee1/OneDrive/Documents/R/legalAnalytics/data/"
data <- read.csv(paste0(DIR,"stateCases.csv"), skip = 1) # 1st row is the description of variables
# A few cases have more than one case type. Manipulate it so that one row has only one case type.
# As a result, a case can show up more than once when it has more than once case type.
data <-
data %>%
# row number makes it easire to remove Cam's data
mutate(case_id = row_number()) %>%
separate(CaseType, c("s1","s2","s3")) %>%
gather(type,CaseType, s1:s3) %>%
filter(!is.na(CaseType)) %>%
select(-type)
# Import CaseType description and merge
CaseType_description <- read.csv(paste0(DIR,"CaseType_description.csv")) %>%
# delete explanations within ()
mutate(CaseType_des = str_remove_all(CaseType_des, regex("\\(.*")),
CaseType = as.factor(CaseType))
data_merged <-
data %>%
left_join(CaseType_description)
# Sturecture of data
str(data_merged)
## 'data.frame': 698 obs. of 41 variables:
## $ DoNo : Factor w/ 667 levels "211-2015-CV-00001",..: 605 62 357 122 295 356 120 27 28 41 ...
## $ FileDate : Factor w/ 344 levels "01/03/2017 •",..: 199 231 236 342 319 125 306 135 135 170 ...
## $ PNames : Factor w/ 556 levels "1466 Bodwell Road, LLC",..: 397 20 173 185 420 16 17 410 221 458 ...
## $ DNames : Factor w/ 655 levels "10 Commerce Park North Condominium Association; Albireo Properties, LLC; Andover - 10 Commerce Park North, LLC;"| __truncated__,..: 480 378 474 94 583 196 33 524 92 407 ...
## $ ProSe : Factor w/ 5 levels "Both","Defendant",..: 3 2 3 5 3 3 2 2 2 2 ...
## $ PLaw : Factor w/ 389 levels "9","A. Gerard O'Neil, Jr.",..: 88 120 378 16 172 227 102 25 312 3 ...
## $ PFirm : Factor w/ 262 levels "9","Abrahamsen Ratchford, P.C.",..: 211 262 149 5 171 36 262 67 192 221 ...
## $ EqReq : Factor w/ 5 levels "0","9","No","Unknown",..: 3 5 5 5 5 5 5 5 5 5 ...
## $ MonReq : Factor w/ 184 levels "$0 ","$0.00 ",..: 111 79 87 111 60 110 47 45 111 83 ...
## $ LoC : int 7 0 11 3 0 2 2 0 0 0 ...
## $ DLaw : Factor w/ 339 levels "0","9","Abbott, Jr., Winslow Kirk; Bello, James A",..: 93 2 28 314 26 240 2 2 2 2 ...
## $ DFirm : Factor w/ 189 levels "","0","9","Andrew Eills Law Offices, PLLC; Sheehan, Phinney, Bass + Green, P.A.",..: 46 3 128 130 31 6 3 3 3 3 ...
## $ X3d : Factor w/ 44 levels "","0","Alitz, Jeffrey L; Lasna, Meredith M.",..: 2 2 2 2 2 2 2 2 2 2 ...
## $ X3dLaw : Factor w/ 17 levels "0","9","David W. Johnston",..: 1 1 1 1 1 1 1 1 1 1 ...
## $ X3dFirm : Factor w/ 14 levels "0","9","Devine, Millimet & Branch, P.A.",..: 1 1 1 1 1 1 1 1 1 1 ...
## $ CtrCl : int 1 0 1 0 0 0 0 0 0 0 ...
## $ CrsCl : int 0 0 0 2 0 0 0 0 0 0 ...
## $ Rec : Factor w/ 6 levels "Ignatius","MacLeod; Borenstein",..: NA NA NA NA NA NA NA NA NA NA ...
## $ Rem : Factor w/ 7 levels "HRC","Made the appeal and then withdrew, so NA (thought you'd like to know… the guy sued NH and won!)",..: NA NA NA NA NA NA NA NA NA NA ...
## $ Judge : Factor w/ 49 levels "","9","Abramson",..: 22 37 14 37 37 23 37 37 37 37 ...
## $ OthMot : Factor w/ 325 levels "","Assented Motion to Continue",..: 166 84 306 312 90 207 86 82 325 164 ...
## $ MTime : Factor w/ 6 levels "","0","1","1; 2",..: 6 2 6 2 2 2 2 2 6 2 ...
## $ MTA : Factor w/ 9 levels "","0","1","2",..: 2 2 4 3 2 2 2 2 2 2 ...
## $ MTD : Factor w/ 14 levels "0","1","1; 3",..: 1 1 1 1 1 1 1 1 1 1 ...
## $ DSJ : Factor w/ 7 levels "0","1","10","2",..: 1 1 1 1 1 1 1 1 1 1 ...
## $ PSJ : Factor w/ 6 levels "0","1","1;3",..: 1 1 2 1 1 1 1 1 1 1 ...
## $ DMIL : Factor w/ 10 levels "0","1","1; 10",..: 1 1 1 1 1 1 1 1 1 1 ...
## $ PMIL : int 0 0 0 0 0 0 0 0 0 0 ...
## $ Stp : int 0 0 0 1 1 0 0 0 0 0 ...
## $ Arb : Factor w/ 3 levels "N/A","No","Yes": 2 NA 2 2 NA 2 2 NA NA NA ...
## $ TrialType : Factor w/ 6 levels "0","1","2","8",..: 2 5 5 5 5 5 5 5 5 5 ...
## $ DISP : Factor w/ 16 levels "0","1","10","2",..: 9 15 4 7 6 6 11 6 6 15 ...
## $ JMent : Factor w/ 7 levels "1","1; NA","2",..: 1 1 1 2 NA NA 1 NA NA 1 ...
## $ NOJ : Factor w/ 7 levels "0","1","1; 0",..: 2 2 2 2 1 1 2 1 1 2 ...
## $ Award : Factor w/ 87 levels "$1,700,000 ",..: 5 47 54 50 24 70 16 71 71 51 ...
## $ Eq : Factor w/ 6 levels "$34,000 ","0",..: 2 6 2 2 6 2 2 6 6 6 ...
## $ TermDate : Factor w/ 448 levels "1/10/2017","1/10/2018",..: 302 89 59 49 271 252 210 444 253 164 ...
## $ Enforce : Factor w/ 10 levels "0","1","1; 2",..: 7 4 3 3 3 9 8 8 8 8 ...
## $ case_id : int 1 2 3 4 5 6 7 8 9 10 ...
## $ CaseType : chr "4" "2" "2" "2" ...
## $ CaseType_des: chr "Contract - Other" "Contract - Collection" "Contract - Collection" "Contract - Collection" ...
# Basic descriptive statistics of data
summary(data_merged)
## DoNo FileDate
## 212-2013-CV-00012: 3 2/24/2016 : 14
## 216-2011-CV-00105: 3 2/1/2016 : 12
## 217-2010-CV-05006: 3 10/31/2016: 9
## 217-2011-CV-00270: 3 5/2/2016 : 8
## 217-2011-CV-00730: 3 9/12/2016 : 8
## 217-2012-CV-00658: 3 1/25/2016 : 7
## (Other) :680 (Other) :640
## PNames
## Discover Bank : 20
## American Express Centurion Bank : 16
## Barclays Bank Delaware : 8
## Paugus Bay Plaza Condominium Association: 8
## Barcklay Bank Delaware : 7
## Concord Hospital : 6
## (Other) :633
## DNames
## Town of Gilford : 6
## AEO Associates, LLC; KMO Associates, LLC; KMO Associates, LP; Aaron Olson : 3
## Arthur Cleaves; Jacob Emerson; Tom Gerrish; Joseph Guerra; Mark Jadkowski; Majella Global Technologies Asia Pacific PTY LTD : 3
## Ballentine & Company, Inc.; Ballentine Partners, LLC; Roy Ballentine; Andrew McMorrow; Gregory Peterson; Kyle Schaffer; Claudia Shilo: 3
## BK Systems, Inc; Ernest Houle; Karlton Klardie : 3
## Bryan Gould; Thomas Janosz; N.H. Board of Registration of Funeral Directors and Embalmers; Susan Simonds : 3
## (Other) :677
## ProSe PLaw
## Both : 32 9 : 45
## Defendant:216 Robert L. O'Brien : 24
## Neither :386 O'Brien, Robert L : 19
## Plaintiff: 24 N/A : 12
## Some DF : 40 Arnold Rosenblatt : 10
## Michael J. Fontaine: 10
## (Other) :578
## PFirm EqReq
## 9 :164 0 : 2
## Schlee & Stillman, LLC : 24 9 : 14
## Zwicker & Associates, P.C. : 20 No : 94
## Welts, White, & Fontaine, P.C. : 16 Unknown: 7
## Cook, Little, Rosenblatt, & Manson, PLLC : 15 Yes :580
## McLane, Graf, Raulerson & Middleton, P.A.: 15 NA's : 1
## (Other) :444
## MonReq LoC DLaw
## 1 :383 Min. : 0.000 9 :188
## 0 :122 1st Qu.: 0.000 N/A : 83
## 9 : 11 Median : 0.000 Burt, Gary Michael: 5
## 25,000.00: 2 Mean : 3.378 Joseph D. Becher : 5
## $0 : 1 3rd Qu.: 4.000 Daniel J. Orroth : 4
## $0.00 : 1 Max. :56.000 Edward D. Philpot : 4
## (Other) :178 (Other) :409
## DFirm
## 9 :299
## N/A : 83
## Getman, Schulthess, Steere, & Poulin, P.A.: 11
## Civil Bureau : 8
## Mitchell Municipal Group, P.A. : 8
## Primmer, Piper, Eggleston, & Cramer, P.C. : 8
## (Other) :281
## X3d X3dLaw
## 0 :640 0 :667
## Andrew Howe; Martina Howe: 4 N/A : 8
## : 3 Ethan G. Wood : 4
## Cross Insurance Agency : 3 Donald A. Kennedy: 3
## Delaney O'Hara, Diana, : 3 Leigh S. Willey : 3
## Donald A. Kennedy : 3 Jamie N. Hage : 2
## (Other) : 42 (Other) : 11
## X3dFirm CtrCl CrsCl
## 0 :666 Min. :0.0000 Min. :0.00000
## N/A : 8 1st Qu.:0.0000 1st Qu.:0.00000
## 9 : 6 Median :0.0000 Median :0.00000
## Patrick Wood Law Office, PLLC : 4 Mean :0.1074 Mean :0.02436
## Devine, Millimet & Branch, P.A.: 3 3rd Qu.:0.0000 3rd Qu.:0.00000
## Hage Hodes, P.A. : 2 Max. :2.0000 Max. :2.00000
## (Other) : 9
## Rec Rem Judge
## Ignatius : 4 N/A :185 O'Neill :260
## MacLeod; Borenstein: 1 Removed : 22 Brown : 75
## N/a : 1 Transfer : 4 Abramson : 66
## N/A :192 Transferred: 2 McNamara : 58
## Na : 1 HRC : 1 Kissinger: 36
## O'Neill : 5 (Other) : 2 (Other) :151
## NA's :494 NA's :482 NA's : 52
## OthMot MTime MTA
## None :205 : 2 0 :615
## Motion for Entry of Final Judgment: 44 0 :486 1 : 66
## Motion for Voluntary Nonsuit : 16 1 : 30 2 : 10
## Assented to Motion to Continue : 14 1; 2: 5 3 : 2
## Motion to Strike : 12 1;2 : 6 : 1
## Motion to Amend Complaint : 9 2 :169 2;4 : 1
## (Other) :398 (Other): 3
## MTD DSJ PSJ DMIL PMIL
## 0 :601 0 :653 0 :667 0 :679 Min. : 0.000
## 1 : 32 1 : 9 1 : 17 10 : 5 1st Qu.: 0.000
## 3 : 20 10 : 21 1;3: 1 9 : 5 Median : 0.000
## 10 : 11 2 : 2 10 : 5 1 : 3 Mean : 0.149
## 2 : 9 2; 4: 1 3 : 7 1; 10 : 1 3rd Qu.: 0.000
## 6 : 9 3 : 11 6 : 1 1; 3 : 1 Max. :10.000
## (Other): 16 4 : 1 (Other): 4
## Stp Arb TrialType DISP JMent
## Min. :0.0000 N/A :193 0 : 17 3 :226 1 :142
## 1st Qu.:0.0000 No :110 1 : 89 9 :121 2 :102
## Median :0.0000 Yes : 58 2 : 54 8 : 83 x : 89
## Mean :0.5057 NA's:337 8 : 1 6 : 78 N/A : 23
## 3rd Qu.:1.0000 9 :535 1 : 73 9 : 21
## Max. :2.0000 N/A: 2 (Other):116 (Other): 3
## NA's : 1 NA's :318
## NOJ Award Eq TermDate
## 0 :362 0 :250 $34,000 : 1 Open : 19
## 9 :231 1 :216 0 :191 4/3/2017 : 6
## 1 : 89 9 :146 1 : 4 3/30/2015: 5
## 3 : 9 $143,552 : 3 2 : 2 4/13/2017: 5
## 2 : 3 $1,700,000 : 1 3 : 6 6/27/2016: 5
## (Other): 2 $1,700.74 : 1 9 :494 (Other) :657
## NA's : 2 (Other) : 81 NA's : 1
## Enforce case_id CaseType CaseType_des
## 0 :646 Min. : 1.0 Length:698 Length:698
## 1 : 19 1st Qu.:144.2 Class :character Class :character
## 3 : 16 Median :318.5 Mode :character Mode :character
## 1;3 : 4 Mean :321.9
## 2 : 4 3rd Qu.:492.8
## 1; 2 : 3 Max. :667.0
## (Other): 6Clean
# Create a variable, def_cat (plt_cat), whether a business is involved in defendant (or plaintiff)
bus_symbols <- "\\sL\\.?L\\.?C\\.?|\\sP\\.?C\\.?|\\sP\\.?A\\.?\\s|\\sInc|\\sL\\.?P\\.?|\\sL\\.?L\\.?P\\.?|\\sP\\.?L\\.?P\\.?|\\sP\\.?L\\.?L\\.?P\\.?|\\sP\\.?L\\.?L\\.?C\\.?|\\sd\\.?b\\.?a\\.?|\\sd\\.?b\\.?a\\.?|\\sd/b/a|\\sp\\.?b\\.?c\\.?|\\sg\\.?p\\.?|\\scorp|\\slimited|\\sltd\\.?|\\sco(\\.|$)|\\scom(,\\s|\\s|$)|\\scomp\\s?|\\scompany\\s?|\\scompanies\\.?|\\spartnership|\\ssole\\sproprietorship|\\sassociation|bank"
gov_symbols <- "state|city|town"
data_processed <-
data_merged %>%
# separate docket number into county, year, type and case
separate(DoNo, c("county","year","type","caseN"), remove = FALSE) %>%
# create length of suit
mutate(FileDate = as.Date(FileDate, "%m/%d/%Y"),
TermDate = as.Date(TermDate, "%m/%d/%Y"),
lengthOfsuit = as.numeric(difftime(time1 = TermDate,
time2 = FileDate,
units = "days"))) %>%
# create def_cat, whether a business is involved in defendant
mutate(
def_cat = "Neither",
def_cat = case_when(
str_detect(DNames, regex(bus_symbols, ignore_case = TRUE)) ~ "Bus",
str_detect(DNames, regex(gov_symbols, ignore_case = TRUE)) ~ "Gov",
TRUE ~ def_cat),
plt_cat = "neither",
plt_cat = case_when(
str_detect(PNames, regex(bus_symbols, ignore_case = TRUE)) ~ "Bus",
str_detect(PNames, regex(gov_symbols, ignore_case = TRUE)) ~ "Gov",
TRUE ~ plt_cat),
business = "neither",
business = case_when(
def_cat == "Bus" & plt_cat == "Bus" ~ "both",
def_cat == "Bus" | plt_cat == "Bus" ~ "only_one",
TRUE ~ business),
government = "neither",
government = case_when(
def_cat == "Gov" & plt_cat == "Gov" ~ "both",
def_cat == "Gov" | plt_cat == "Gov" ~ "only_one",
TRUE ~ government)
) %>%
# label disposition
mutate(
DISP_des = case_when(
DISP == 1 ~ "dismissed",
DISP == 2 ~ "summary judgement",
DISP == 3 ~ "settled",
DISP == 4 ~ "jury verdict",
DISP == 5 ~ "bench trial",
DISP == 6 ~ "default judgment",
DISP == 7 ~ "voluntary nonsuit",
DISP == 8 ~ "other resolution",
DISP == 9 ~ "unknown resolution",
DISP == 10 ~ "pending"
)
) %>%
# label trial type
mutate(
TrialType_des = case_when(
TrialType == "1" ~ "jury trial",
TrialType == "2" ~ "bench trial",
TrialType == "9" ~ "no trial",
TrialType == "0" ~ "still open",
TRUE ~ as.character(TrialType)
)
) %>%
# label judgment
mutate(
JMent_des = case_when(
JMent == 1 ~ "Pltf",
JMent == 2 ~ "Def",
JMent == 3 ~ "Both parties",
JMent == 9 ~ "still open",
TRUE ~ "Settlement"
)
) %>%
# plt_winning will be a response var to predict the odds of the plaintiff winning
mutate(plt_winning = NA_character_,
plt_winning = case_when(
JMent_des == "Pltf" ~ "1",
JMent_des == "Def" ~ "0",
TRUE ~ plt_winning
)) %>%
# remove $ and , and white space from MonReq
mutate(MonReq = str_squish(MonReq),
MonReq = str_remove_all(MonReq, "\\$|\\,"),
MonReq = as.numeric(MonReq)) %>%
# remove $ and , and white space from Award
mutate(Award = str_squish(Award),
Award = str_remove_all(Award, "\\$|\\,"),
Award = as.numeric(Award)) %>%
# county (origin of case) is taken from docket id but Jurisdiction (where it's tried)
# is identified by Chantalle
mutate(
Jurisdiction = NA_character_,
Jurisdiction = case_when(
case_id <= 425 ~ "Belknap",
case_id >= 426 & case_id <= 618 ~ "Merrimack",
case_id >= 619 ~ "Hills North",
TRUE ~ Jurisdiction)
) %>%
# Chantalle request changes in case type after finding errors in students work
mutate(CaseType_des = case_when(
CaseType %in% c("1","2","4","24") ~ "Contract Claims",
CaseType %in% c("7","8","9","11","12","13","37","40") ~ "Negligence Claims other than Medical Malpractice",
CaseType %in% c("14","15","16","17") ~ "Real Property Cases",
TRUE ~ CaseType_des
)) %>%
mutate(CaseType_des = case_when(
CaseType_des == "Declaratory Judgement Unknown" ~ "Declaratory Judgment",
CaseType_des == "Negligence Claims other than Medical Malpractice" ~ "Negligence other than Medical Malpractice",
TRUE ~ CaseType_des
)) %>%
# simplify ProSe
mutate(
ProSe_rev = NA_character_,
ProSe_rev = case_when(
ProSe == "Defendant" | ProSe == "Plaintiff" ~ "only_one",
ProSe == "Both" ~ "both",
ProSe == "Neither" ~ "neither",
TRUE ~ ProSe_rev
))
glimpse(data_processed)
## Observations: 698
## Variables: 56
## $ DoNo <fct> 217-2011-CV-00264, 211-2015-CV-00092, 216-2012-C...
## $ county <chr> "217", "211", "216", "211", "211", "216", "211",...
## $ year <chr> "2011", "2015", "2012", "2015", "2016", "2012", ...
## $ type <chr> "CV", "CV", "CV", "CV", "CV", "CV", "CV", "CV", ...
## $ caseN <chr> "00264", "00092", "00521", "00203", "00224", "00...
## $ FileDate <date> 2011-04-27, 2015-05-06, 2012-06-13, 2015-09-08,...
## $ PNames <fct> "Olde Province Commons, LLC", "American Express ...
## $ DNames <fct> "Nitro Fireworks Co., Inc; Charles Shivery", "Ke...
## $ ProSe <fct> Neither, Defendant, Neither, Some DF, Neither, N...
## $ PLaw <fct> "Donald C. Crandlemire", "Frederick B. Seitz, IV...
## $ PFirm <fct> "Shaheen & Gordon, P.A.", "Zwicker & Associates,...
## $ EqReq <fct> No, Yes, Yes, Yes, Yes, Yes, Yes, Yes, Yes, Yes,...
## $ MonReq <dbl> 1.00, 4768.06, 485154.04, 1.00, 27112.01, 0.00, ...
## $ LoC <int> 7, 0, 11, 3, 0, 2, 2, 0, 0, 0, 0, 0, 0, 0, 0, 0,...
## $ DLaw <fct> Edward D. Philpot, 9, Brian D. Duffy, Timothy K....
## $ DFirm <fct> "Edward D. Philpot, Jr., PLLC", "9", "Nixon Peab...
## $ X3d <fct> 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, ...
## $ X3dLaw <fct> 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, ...
## $ X3dFirm <fct> 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, ...
## $ CtrCl <int> 1, 0, 1, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, ...
## $ CrsCl <int> 0, 0, 0, 2, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, ...
## $ Rec <fct> NA, NA, NA, NA, NA, NA, NA, NA, NA, NA, NA, NA, ...
## $ Rem <fct> NA, NA, NA, NA, NA, NA, NA, NA, NA, NA, NA, NA, ...
## $ Judge <fct> "McNamara", "O'Neill", "Garfunkel; McNamara", "O...
## $ OthMot <fct> Motion to Amend Brief Statement of Defenses; Mot...
## $ MTime <fct> 2, 0, 2, 0, 0, 0, 0, 0, 2, 0, 0, 0, 0, 0, 0, 0, ...
## $ MTA <fct> 0, 0, 2, 1, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 1, 0, ...
## $ MTD <fct> 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, ...
## $ DSJ <fct> 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, ...
## $ PSJ <fct> 0, 0, 1, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, ...
## $ DMIL <fct> 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, ...
## $ PMIL <int> 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, ...
## $ Stp <int> 0, 0, 0, 1, 1, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, ...
## $ Arb <fct> No, NA, No, No, NA, No, No, NA, NA, NA, NA, NA, ...
## $ TrialType <fct> 1, 9, 9, 9, 9, 9, 9, 9, 9, 9, 9, 9, 9, 9, 9, 9, ...
## $ DISP <fct> 4, 8, 2, 3; 6, 3, 3, 6, 3, 3, 8, 6, 8, 6, 6, 6, ...
## $ JMent <fct> 1, 1, 1, 1; NA, NA, NA, 1, NA, NA, 1, 1, 1, 1, 1...
## $ NOJ <fct> 1, 1, 1, 1, 0, 0, 1, 0, 0, 1, 1, 1, 1, 1, 1, 1, ...
## $ Award <dbl> 108628.00, 4069.06, 485154.04, 40000.00, 2200.00...
## $ Eq <fct> "0", "9", "0", "0", "9", "0", "0", "9", "9", "9"...
## $ TermDate <date> 2012-06-14, 2015-11-24, 2012-10-04, 2016-10-26,...
## $ Enforce <fct> 2; 3, 1; 3, 1; 2, 1; 2, 1; 2, 4, 3, 3, 3, 3, 3, ...
## $ case_id <int> 1, 2, 3, 4, 5, 6, 7, 8, 9, 10, 11, 12, 13, 14, 1...
## $ CaseType <chr> "4", "2", "2", "2", "2", "19", "2", "4", "40", "...
## $ CaseType_des <chr> "Contract Claims", "Contract Claims", "Contract ...
## $ lengthOfsuit <dbl> 414, 202, 113, 414, 616, 466, 226, 196, 78, 310,...
## $ def_cat <chr> "Bus", "Neither", "Bus", "Bus", "Bus", "Neither"...
## $ plt_cat <chr> "Bus", "Bus", "Bus", "Bus", "Bus", "Bus", "neith...
## $ business <chr> "both", "only_one", "both", "both", "both", "onl...
## $ government <chr> "neither", "neither", "neither", "neither", "nei...
## $ DISP_des <chr> "jury verdict", "other resolution", "summary jud...
## $ TrialType_des <chr> "jury trial", "no trial", "no trial", "no trial"...
## $ JMent_des <chr> "Pltf", "Pltf", "Pltf", "Settlement", "Settlemen...
## $ plt_winning <chr> "1", "1", "1", NA, NA, NA, "1", NA, NA, "1", "1"...
## $ Jurisdiction <chr> "Belknap", "Belknap", "Belknap", "Belknap", "Bel...
## $ ProSe_rev <chr> "neither", "only_one", "neither", NA, "neither",...Export spreadsheet data
#write.csv(data_processed, "C:/Users/sclee1/OneDrive/Documents/R/legalAnalytics/data/stateCases_spreadsheet.csv")Explore
- While the data includes 667 cases, the processed data has 698 rows. This is because a case can have more than one case type. The data has been transformed so that a row represents one case and one type.
- Most of the data came from 2016.
- Contract Claims was the most frequent nature of suit category in Belknap County in 2016. It made up 66 of 215 all civil cases in 2016.
- It was followed by Tort - Automobile (29 cases), Restraining Order (18), Insurance Coverage (17), Complaint for Injunction (14).
- Tort - Medical Malpractice had only 5 cases.
data_processed %>%
distinct(case_id, year) %>%
ggplot(aes(year)) +
geom_bar(position = "stack")
data_processed %>%
distinct(case_id, year, Jurisdiction) %>%
ggplot(aes(year, fill = Jurisdiction)) +
geom_bar(position = "stack") +
facet_wrap(~Jurisdiction, ncol = 1) +
coord_flip() +
labs(x = NULL,
y = "Number of cases")
data_processed %>%
filter(lengthOfsuit > 0) %>%
ggplot(aes(lengthOfsuit)) +
geom_histogram() +
scale_x_log10()
# without Restraining Order
data_processed %>%
count(CaseType_des, sort = TRUE) %>%
filter(CaseType_des != "Restraining Order") %>%
head(10) %>%
mutate(CaseType_des = str_sub(CaseType_des, 1, 30),
CaseType_des = fct_reorder(CaseType_des, n)) %>%
ggplot(aes(CaseType_des, n)) +
geom_col(fill = "cornflowerblue") +
coord_flip() +
labs(title = "Top 10 Most Common Case Types in New Hampshire",
x = NULL,
y = "Number of Cases")
data_processed %>%
count(CaseType_des, sort = TRUE) %>%
head(10) %>%
mutate(CaseType_des = str_sub(CaseType_des, 1, 30),
CaseType_des = fct_reorder(CaseType_des, n)) %>%
ggplot(aes(CaseType_des, n)) +
geom_col() +
coord_flip()
data_processed %>%
count(DISP_des, sort = TRUE)
## # A tibble: 11 x 2
## DISP_des n
## <chr> <int>
## 1 settled 226
## 2 unknown resolution 121
## 3 other resolution 83
## 4 default judgment 78
## 5 dismissed 73
## 6 voluntary nonsuit 63
## 7 pending 18
## 8 summary judgement 16
## 9 <NA> 11
## 10 bench trial 6
## 11 jury verdict 3
data_processed %>%
count(JMent_des, sort = TRUE)
## # A tibble: 5 x 2
## JMent_des n
## <chr> <int>
## 1 Settlement 432
## 2 Pltf 142
## 3 Def 102
## 4 still open 21
## 5 Both parties 1
data_processed %>%
count(DISP_des, JMent_des, sort = TRUE)
## # A tibble: 24 x 3
## DISP_des JMent_des n
## <chr> <chr> <int>
## 1 settled Settlement 226
## 2 unknown resolution Settlement 109
## 3 default judgment Pltf 69
## 4 dismissed Def 58
## 5 other resolution Pltf 48
## 6 voluntary nonsuit Settlement 38
## 7 other resolution Settlement 24
## 8 voluntary nonsuit Def 24
## 9 pending still open 18
## 10 dismissed Settlement 15
## # ... with 14 more rows
# the plaintiff sought an equitable remedy
data_processed %>%
count(EqReq, sort = TRUE)
## # A tibble: 6 x 2
## EqReq n
## <fct> <int>
## 1 Yes 580
## 2 No 94
## 3 9 14
## 4 Unknown 7
## 5 0 2
## 6 <NA> 1
# the plaintiff sought a monetary remedy
data_processed %>%
count(MonReq, sort = TRUE)
## # A tibble: 180 x 2
## MonReq n
## <dbl> <int>
## 1 1 383
## 2 0 124
## 3 9 11
## 4 35000 3
## 5 25000 2
## 6 1172. 1
## 7 1436. 1
## 8 1501. 1
## 9 1503. 1
## 10 1582. 1
## # ... with 170 more rows
# 1 and 0 are not dollar vlaues
data_processed %>%
filter(MonReq > 10) %>%
ggplot(aes(MonReq)) +
geom_histogram() +
scale_x_log10(label = scales::dollar)
data_processed %>%
filter(!Award %in% c("0","1","9")) %>%
ggplot(aes(Award)) +
geom_histogram(binwidth = 0.5) +
scale_x_log10(labels = scales::dollar)
data_processed %>%
filter(EqReq %in% c("Yes","No"),
MonReq > 10) %>%
count(EqReq, MonReq, sort = TRUE)
## # A tibble: 177 x 3
## EqReq MonReq n
## <fct> <dbl> <int>
## 1 Yes 35000 3
## 2 Yes 25000 2
## 3 Yes 1172. 1
## 4 Yes 1436. 1
## 5 Yes 1501. 1
## 6 Yes 1503. 1
## 7 Yes 1582. 1
## 8 Yes 1670. 1
## 9 Yes 1671. 1
## 10 Yes 1698. 1
## # ... with 167 more rows
data_processed %>%
count(county, sort = TRUE)
## # A tibble: 8 x 2
## county n
## <chr> <int>
## 1 211 342
## 2 216 252
## 3 217 66
## 4 218 14
## 5 212 9
## 6 226 8
## 7 215 4
## 8 219 3
data_processed %>%
count(Jurisdiction, sort = TRUE)
## # A tibble: 3 x 2
## Jurisdiction n
## <chr> <int>
## 1 Belknap 456
## 2 Merrimack 193
## 3 Hills North 49
data_processed %>%
count(def_cat, sort = TRUE)
## # A tibble: 3 x 2
## def_cat n
## <chr> <int>
## 1 Neither 387
## 2 Bus 286
## 3 Gov 25
data_processed %>%
count(plt_cat, sort = TRUE)
## # A tibble: 3 x 2
## plt_cat n
## <chr> <int>
## 1 neither 378
## 2 Bus 304
## 3 Gov 16
data_processed %>%
count(ProSe, sort = TRUE)
## # A tibble: 5 x 2
## ProSe n
## <fct> <int>
## 1 Neither 386
## 2 Defendant 216
## 3 Some DF 40
## 4 Both 32
## 5 Plaintiff 24
data_processed %>%
count(PLaw, sort = TRUE)
## # A tibble: 389 x 2
## PLaw n
## <fct> <int>
## 1 9 45
## 2 Robert L. O'Brien 24
## 3 O'Brien, Robert L 19
## 4 N/A 12
## 5 Arnold Rosenblatt 10
## 6 Michael J. Fontaine 10
## 7 Peter S. Cowan 9
## 8 Benjamin R. Roberge 8
## 9 Edward D. Philpot 8
## 10 Niederman, Jay M 8
## # ... with 379 more rows
# length of complaint pages
data_processed %>%
filter(LoC != 0) %>%
ggplot(aes(LoC)) +
geom_histogram(binwidth = 0.1) +
scale_x_log10()
# A majority had no trial; another large chunk left blank; 89 jury trials vs. 54 bench trials
data_processed %>%
count(TrialType_des, sort = TRUE)
## # A tibble: 6 x 2
## TrialType_des n
## <chr> <int>
## 1 no trial 535
## 2 jury trial 89
## 3 bench trial 54
## 4 still open 17
## 5 N/A 2
## 6 8 1MonReq versus. Award
There is a strong positive one-to-one association between how much money requested and how much awarded. In addition, how much plaintiffs request doesn’t seem to reduce the likelihood of getting awarded.
# When $ was awarded
data_processed %>%
filter(MonReq > 10,
!Award %in% c("0","1","9")) %>%
ggplot(aes(Award, MonReq)) +
geom_point() +
scale_x_log10(labels = scales::dollar) +
scale_y_log10(labels = scales::dollar)
# When $ was awarded versus when not
library(ggbeeswarm)
data_processed %>%
mutate(Award_cat = if_else(Award == "0","No Monetary Award","Monetary Award")) %>%
filter(MonReq > 10,
# filter out undefined amount and I'm not sure what 9 means
!Award %in% c("1","9")) %>%
ggplot(aes(Award_cat, MonReq, color = Award_cat)) +
geom_boxplot(notch = TRUE) +
scale_y_log10(labels = scales::dollar) +
theme(legend.position = "none") +
labs(title = "Table 3: Effect of Monetary Request on Outcome and Award",
x = "Outcome",
y = "Plaintiff’s Monetary Request")
med_MonReqs <-
data_processed %>%
mutate(Award_cat = if_else(Award == "0","No Monetary Award","Monetary Award")) %>%
filter(MonReq > 10,
# filter out undefined amount and I'm not sure what 9 means
!Award %in% c("1","9")) %>%
group_by(Award_cat) %>%
summarise(med_MonReq = median(MonReq))
med_MonReqs
## # A tibble: 2 x 2
## Award_cat med_MonReq
## <chr> <dbl>
## 1 Monetary Award 7170.
## 2 No Monetary Award 14963
data_processed %>%
filter(MonReq > 10,
# filter out undefined amount and I'm not sure what 9 means
!Award %in% c("1","9")) %>%
mutate(Award_cat = if_else(Award == "0","No Monetary Award","Monetary Award")) %>%
ggplot(aes(x="", MonReq, color = Award_cat)) +
geom_quasirandom() +
scale_y_log10(labels = scales::dollar) +
theme(legend.position = "none") +
labs(title = "Table 3: Effect of Monetary Request on Outcome and Award",
x = "Outcome",
y = "Plaintiff’s Monetary Request") +
geom_hline(data = med_MonReqs, aes(yintercept = med_MonReq, color = Award_cat)) +
facet_wrap(~Award_cat)
Award by county, by NOS, by TrialType, and by Plt_Cat and Def_Cat
Award mostly ranges from $1,000 and to $1 million, half of which fall below about $14,000. Interpret with caution as only 85 cases, of 667 cases, have an identified value of award.
- County: award amount appears to be higher in County 217. But the sample size is small.
- CaseType: not enought data to make comparisons
- TrialType: award amount appears to be higher in bench trial. But the sample size i small.
- def_cat: a large award above $100,000 almost exclusively happens when the defendant is business.
- plt_cat: no clear association with award
- other notes: most cases in Belknap County Court (211) don’t go to trial; it’s unlikely to get monetary award by suing governments
# median award by case type
data_processed %>%
filter(!Award %in% c("0","1","9")) %>%
group_by(CaseType_des) %>%
summarise(median(Award, na.rm = TRUE))
## # A tibble: 11 x 2
## CaseType_des `median(Award, na.rm = TRU~
## <chr> <dbl>
## 1 "Complaint for Injunction/To Enjoin Foreclo~ 26481.
## 2 Consumer Protection Act 143552
## 3 Contract Claims 16115.
## 4 Declaratory Judgment 132700.
## 5 "Forfeiture " 2532.
## 6 Interpleder 9525.
## 7 Negligence other than Medical Malpractice 138009.
## 8 Plea of Indemnity 81649.
## 9 "Tort - Intentional " 143552
## 10 "Tort - Medical Malpractice " 125000
## 11 Wage Claim 5500
library(ggbeeswarm)
# Remove Cam's data using row numbers I created early on
data_processed_HilCounty <-
data_processed %>%
filter(case_id <= 425 | case_id >= 619)
data_processed %>%
filter(Award > 10) %>%
summarise(caseN_award = n())
## caseN_award
## 1 85
median_Award <-
data_processed %>%
filter(!Award %in% c("0","1","9")) %>%
summarise(award_median = median(Award, na.rm = TRUE)) %>%
pull()
data_processed %>%
filter(!Award %in% c("0","1","9")) %>%
ggplot(aes(Award)) +
geom_histogram(binwidth = 0.5, fill = "cornflowerblue") +
geom_vline(linetype = "dashed", xintercept = median_Award) +
scale_x_log10(labels = scales::dollar)
data_processed %>%
filter(!Award %in% c("0","1","9"),
county %in% c("211","216","217")) %>%
ggplot(aes(county, Award, color = county)) +
geom_quasirandom(alpha = 0.7,
size = 1.5) +
scale_y_log10(labels = scales::dollar) +
theme(legend.position = "none")
data_processed %>%
mutate(CaseType_des = fct_lump(CaseType_des, 3)) %>%
filter(!Award %in% c("0","1","9"),
CaseType_des != "Other") %>%
ggplot(aes(CaseType_des, Award, color = CaseType_des)) +
geom_quasirandom(alpha = 0.7,
size = 1.5) +
scale_y_log10(labels = scales::dollar) +
theme(legend.position = "none")
data_processed %>%
filter(!Award %in% c("0","1","9"),
TrialType_des %in% c("jury trial","bench trial","no trial")) %>%
ggplot(aes(TrialType_des, Award, color = TrialType)) +
geom_quasirandom(alpha = 0.7,
size = 1.5) +
scale_y_log10(labels = scales::dollar) +
theme(legend.position = "none") 
data_processed %>%
filter(!Award %in% c("0","1","9")) %>%
ggplot(aes(def_cat, Award, color = def_cat)) +
geom_quasirandom(alpha = 0.7,
size = 1.5) +
scale_y_log10(labels = scales::dollar) +
theme(legend.position = "none")
data_processed %>%
filter(!Award %in% c("0","1","9")) %>%
ggplot(aes(plt_cat, Award, color = plt_cat)) +
geom_quasirandom(alpha = 0.7,
size = 1.5) +
scale_y_log10(labels = scales::dollar) +
theme(legend.position = "none")
data_processed %>%
filter(!Award %in% c("0","1","9"),
TrialType_des %in% c("jury trial","bench trial","no trial"),
county %in% c("211","216","217")) %>%
ggplot(aes(TrialType_des, Award, color = TrialType)) +
geom_quasirandom(alpha = 0.7,
size = 1.5) +
theme(legend.position = "none") +
scale_y_log10(labels = scales::dollar) +
facet_grid(county~def_cat) +
labs(subtitle = "def_cat on the horizontal axis and county on vertical axis")
Correlated variables with Judgment
The list of variables that are strongly associated with Judgment includes ProSe, nature of suit, def_cat (whether defendant is business, government or neither), and county.
Belknap Superior Court (211) tends to enter judgment in favor of the plaintiff more than expected by chance.
# reduce # of levels Jment
data_processed_simpleJment <-
data_processed %>%
filter(JMent_des %in% c("Settlement","Pltf","Def")) %>%
mutate(JMent_des = as.factor(JMent_des),
def_cat = as.factor(def_cat),
Jurisdiction = factor(Jurisdiction, levels = c("Hills North","Merrimack","Belknap")))
# business as a predictor
data_processed_simpleJment %>%
ggplot(aes(x = business, fill = JMent_des)) +
geom_bar(position = "fill") +
labs(y = "Proportion") +
coord_flip()
# government as a predictor
data_processed_simpleJment %>%
ggplot(aes(government, fill = JMent_des)) +
geom_bar(position = "fill") +
labs(y = "Proportion") +
coord_flip()
# create a table
tbl <- xtabs(~JMent_des + government, data_processed_simpleJment)
ftable(tbl)
## government both neither only_one
## JMent_des
## Def 0 94 8
## Pltf 0 134 8
## Settlement 1 410 21
library(vcd)
mosaic(tbl,
shade = TRUE,
legend = TRUE,
main = "Court data")
# ProSe as a predictor
data_processed_simpleJment %>%
ggplot(aes(ProSe_rev, fill = JMent_des)) +
geom_bar(position = "fill") +
labs(y = "Proportion") +
coord_flip()
# create a table
tbl <- xtabs(~JMent_des + ProSe_rev, data_processed_simpleJment)
ftable(tbl)
## ProSe_rev both neither only_one
## JMent_des
## Def 9 43 45
## Pltf 14 17 109
## Settlement 8 309 83
library(vcd)
mosaic(tbl,
shade = TRUE,
legend = TRUE,
main = "Court data")
# case type as a predictor
data_processed_simpleJment %>%
mutate(CaseType_des = fct_lump(CaseType_des, 6)) %>%
ggplot(aes(CaseType_des, fill = JMent_des)) +
geom_bar(position = "fill") +
labs(y = "Proportion") +
coord_flip()
# equitable remedy as a predictor
data_processed_simpleJment %>%
filter(EqReq %in% c("Yes","No")) %>%
ggplot(aes(EqReq, fill = JMent_des)) +
geom_bar(position = "fill") +
labs(y = "Proportion") +
coord_flip()
# create a table
tbl <- xtabs(~JMent_des + EqReq, data_processed_simpleJment)
# reduce levels in EqReq
tbl <- tbl[,c("Yes","No")]
ftable(tbl)
## EqReq Yes No
## JMent_des
## Def 80 20
## Pltf 122 17
## Settlement 360 53
library(vcd)
mosaic(tbl,
shade = TRUE,
legend = TRUE,
main = "Court data")
# def_cat as a predictor
data_processed_simpleJment %>%
ggplot(aes(def_cat, fill = JMent_des)) +
geom_bar(position = "fill") +
labs(y = "Proportion") +
coord_flip()
# create a table
tbl <- xtabs(~JMent_des + def_cat, data_processed_simpleJment)
ftable(tbl)
## def_cat Bus Gov Neither
## JMent_des
## Def 26 5 71
## Pltf 42 5 95
## Settlement 210 14 208
library(vcd)
mosaic(tbl,
shade = TRUE,
legend = TRUE,
labeling_args = list(set_varnames = c(def_cat = "Type of Defendant",
JMent_des = "Judgment")),
main = "Table 2: Correlation of Outcome with Type of Defendant")
# plt_cat as a predictor
data_processed_simpleJment %>%
ggplot(aes(plt_cat, fill = JMent_des)) +
geom_bar(position = "fill") +
labs(y = "Proportion") +
coord_flip()
# create a table
tbl <- xtabs(~JMent_des + plt_cat, data_processed_simpleJment)
ftable(tbl)
## plt_cat Bus Gov neither
## JMent_des
## Def 47 3 52
## Pltf 77 3 62
## Settlement 176 9 247
library(vcd)
mosaic(tbl,
shade = TRUE,
legend = TRUE,
main = "Court data")
# county as a predictor
data_processed_simpleJment %>%
filter(county %in% c(211,216,217,218)) %>%
ggplot(aes(county, fill = JMent_des)) +
geom_bar(position = "fill") +
labs(y = "Proportion") +
coord_flip()
# create a table
tbl <- xtabs(~JMent_des + county, data_processed_simpleJment)
# reduce levels in county
tbl <- tbl[,c("211","216","217","218")]
ftable(tbl)
## county 211 216 217 218
## JMent_des
## Def 59 35 7 1
## Pltf 88 46 7 1
## Settlement 181 166 52 12
library(vcd)
mosaic(tbl,
shade = TRUE,
legend = TRUE,
main = "Court data")
Judgment by Jurisdiction and by Case Type
The analysis includes the most common case types: 1) Contract Claims, 2) Tort - Automobile, and 3) Negligence Claims other than Medical Malpractice. Major findings are:
- It doesn’t appear that judgement is associated with jurisdiction alone.
- There is strong evidence that judgement is associated with case type. For example:
- Fewer than expected cases are settled in Contract Claims.
- More than expected cases are settled in Negligence Claims other than Medical Malpractice.
- The plaintiff wins more cases than expected in Contract Claims.
- The plaintiff loses more cases than expected in Negligence Claims other than Medical Malpractice.
- The plaintiff loses more cases than expected in Tort - Automobile.
Case types are abbreviated in the plots.
- Contract: Contract Claims
- Negligence: Negligence Claims other than Medical Malpractice
- Auto: Tort - Automobile
# reduce # of levels Jment and CaseType_des
data_processed_simpleJment_simple_CaseType <-
data_processed %>%
mutate(CaseType_des = fct_lump(CaseType_des, 2)) %>%
filter(JMent_des %in% c("Settlement","Pltf","Def"),
CaseType_des != "Other") %>%
mutate(JMent_des = as.factor(JMent_des),
def_cat = as.factor(def_cat)) %>%
droplevels()
# Jurisdiction as a predictor
data_processed_simpleJment_simple_CaseType %>%
ggplot(aes(Jurisdiction, fill = JMent_des)) +
geom_bar(position = "fill") +
labs(y = "Proportion") +
coord_flip()
# Jurisdiction & CaseType_des as a predictor
data_processed_simpleJment %>%
mutate(JMent_des = as.character(JMent_des),
JMent_des = case_when(
JMent_des == "Def" ~ "Defendant",
JMent_des == "Pltf" ~ "Plaintiff",
TRUE ~ JMent_des
)) %>%
filter(CaseType_des %in% c("Contract Claims",
"Negligence other than Medical Malpractice",
"Declaratory Judgment",
"Real Property Cases")) %>%
ggplot(aes(Jurisdiction, fill = JMent_des)) +
geom_bar(position = "fill") +
labs(y = "Proportion") +
coord_flip() +
facet_wrap(~CaseType_des) +
labs(title = "Table 1: Judgment by Case Type and Jurisdiction",
fill = "Judgment")
# What's wrong with the bar in Hillsborough North & Negligence? It's all settlement!
# Understand Hillsborough N
data_processed_simpleJment %>%
filter(Jurisdiction == "Hills North") %>%
count(CaseType_des, sort = TRUE)
## # A tibble: 10 x 2
## CaseType_des n
## <chr> <int>
## 1 Contract Claims 22
## 2 Declaratory Judgment 8
## 3 Negligence other than Medical Malpractice 8
## 4 Real Property Cases 3
## 5 Replevin 2
## 6 "Complaint for Injunction/To Enjoin Foreclosure " 1
## 7 Foreclosure - Complaint to Enjoin Foreclosure 1
## 8 Plea of Indemnity 1
## 9 "Tort - Medical Malpractice " 1
## 10 Wage Claim 1
data_processed %>%
filter(Jurisdiction == "Hills North") %>%
count(year, sort = TRUE)
## # A tibble: 1 x 2
## year n
## <chr> <int>
## 1 2016 49
data_processed_simpleJment %>%
mutate(JMent_des = as.character(JMent_des),
JMent_des = case_when(
JMent_des == "Def" ~ "Defendant",
JMent_des == "Pltf" ~ "Plaintiff",
TRUE ~ JMent_des
)) %>%
filter(CaseType_des %in% c("Contract Claims",
"Negligence other than Medical Malpractice",
"Declaratory Judgment",
"Real Property Cases")) %>%
ggplot(aes(Jurisdiction, fill = JMent_des)) +
geom_bar(position = "stack") +
labs(y = "Proportion") +
coord_flip() +
facet_wrap(~CaseType_des) +
labs(title = "Table 1: Judgment by Case Type and Jurisdiction",
fill = "Judgment")
# mosaic: judgment vs. jurisdiction
# create a table
tbl <- xtabs(~JMent_des + Jurisdiction,
data_processed_simpleJment_simple_CaseType)
ftable(tbl)
## Jurisdiction Belknap Hills North Merrimack
## JMent_des
## Def 31 1 24
## Pltf 59 2 28
## Settlement 146 27 79
library(vcd)
mosaic(tbl,
shade = TRUE,
legend = TRUE,
labeling_args = list(set_varnames = c(JMent_des = "Judgment")),
main = "Court data")
# mosaic: judgment vs. case type
# create a table
tbl <- xtabs(~JMent_des + CaseType_des,
data_processed_simpleJment_simple_CaseType)
ftable(tbl)
## CaseType_des Contract Claims Negligence other than Medical Malpractice
## JMent_des
## Def 42 14
## Pltf 82 7
## Settlement 131 121
library(vcd)
mosaic(tbl,
shade = TRUE,
legend = TRUE,
labeling_args = list(set_varnames = c(JMent_des = "Judgment",
CaseType_des = "Case Type")),
set_labels = list(CaseType_des = c("Contract",
"Negligence",
"Auto")),
main = "Court data")
Correlated variables with disposition, DISP, outcomes of cases
The list of variables that are strongly associated with Disposition includes business (whether business is either party, both parties, or neither), ProSe, EqReq, nature of suit, def_cat (whether defendant is business, government or neither), plt_cat (whether plaintiff is business, government or neither).
# reduce # of levels DISP
# consider adding more levels:e.g., pending and summary judgement
data_processed_simpleDISP <-
data_processed %>%
filter(DISP_des %in% c("settled","other resolution","default judgment","dismissed","voluntary nonsuit"))
# business as a predictor
data_processed_simpleDISP %>%
ggplot(aes(x = business, fill = DISP_des)) +
geom_bar(position = "fill") +
labs(y = "Proportion") +
coord_flip()
# create a table
tbl <- xtabs(~DISP_des + business, data_processed_simpleDISP)
ftable(tbl)
## business both neither only_one
## DISP_des
## default judgment 13 14 51
## dismissed 9 34 30
## other resolution 9 39 35
## settled 47 79 100
## voluntary nonsuit 14 16 33
library(vcd)
mosaic(tbl,
shade = TRUE,
legend = TRUE,
main = "Court data")
# government as a predictor
data_processed_simpleDISP %>%
ggplot(aes(government, fill = DISP_des)) +
geom_bar(position = "fill") +
labs(y = "Proportion") +
coord_flip()
# create a table
tbl <- xtabs(~DISP_des + government, data_processed_simpleDISP)
ftable(tbl)
## government both neither only_one
## DISP_des
## default judgment 0 75 3
## dismissed 0 67 6
## other resolution 0 74 9
## settled 0 215 11
## voluntary nonsuit 1 60 2
library(vcd)
mosaic(tbl,
shade = TRUE,
legend = TRUE,
main = "Court data")
# ProSe as a predictor
data_processed_simpleDISP %>%
ggplot(aes(ProSe_rev, fill = DISP_des)) +
geom_bar(position = "fill") +
labs(y = "Proportion") +
coord_flip()
# create a table
tbl <- xtabs(~DISP_des + ProSe_rev, data_processed_simpleDISP)
ftable(tbl)
## ProSe_rev both neither only_one
## DISP_des
## default judgment 0 9 66
## dismissed 11 30 27
## other resolution 11 24 44
## settled 3 181 28
## voluntary nonsuit 1 26 34
library(vcd)
mosaic(tbl,
shade = TRUE,
legend = TRUE,
main = "Court data")
# case type as a predictor
data_processed_simpleDISP %>%
mutate(CaseType_des = fct_lump(CaseType_des, 6)) %>%
ggplot(aes(CaseType_des, fill = DISP_des)) +
geom_bar(position = "fill") +
labs(y = "Proportion") +
coord_flip()
# equitable remedy as a predictor
data_processed_simpleDISP %>%
filter(EqReq %in% c("Yes","No")) %>%
ggplot(aes(EqReq, fill = DISP_des)) +
geom_bar(position = "fill") +
labs(y = "Proportion") +
coord_flip()
# create a table
tbl <- xtabs(~DISP_des + EqReq, data_processed_simpleDISP)
# reduce levels in EqReq
tbl <- tbl[,c("Yes","No")]
ftable(tbl)
## EqReq Yes No
## DISP_des
## default judgment 66 10
## dismissed 50 20
## other resolution 58 23
## settled 205 15
## voluntary nonsuit 55 8
library(vcd)
mosaic(tbl,
shade = TRUE,
legend = TRUE,
main = "Court data")
# def_cat as a predictor
data_processed_simpleDISP %>%
ggplot(aes(def_cat, fill = DISP_des)) +
geom_bar(position = "fill") +
labs(y = "Proportion") +
coord_flip()
# create a table
tbl <- xtabs(~DISP_des + def_cat, data_processed_simpleDISP)
ftable(tbl)
## def_cat Bus Gov Neither
## DISP_des
## default judgment 24 1 53
## dismissed 19 4 50
## other resolution 23 7 53
## settled 104 5 117
## voluntary nonsuit 23 2 38
library(vcd)
mosaic(tbl,
shade = TRUE,
legend = TRUE,
main = "Court data")
# plt_cat as a predictor
data_processed_simpleDISP %>%
ggplot(aes(plt_cat, fill = DISP_des)) +
geom_bar(position = "fill") +
labs(y = "Proportion") +
coord_flip()
# create a table
tbl <- xtabs(~DISP_des + plt_cat, data_processed_simpleDISP)
ftable(tbl)
## plt_cat Bus Gov neither
## DISP_des
## default judgment 53 2 23
## dismissed 29 2 42
## other resolution 30 2 51
## settled 90 6 130
## voluntary nonsuit 38 2 23
library(vcd)
mosaic(tbl,
shade = TRUE,
legend = TRUE,
main = "Court data")