Determinants of bus users’ loyalty - continuous
Phương Anh
15/03/2021
1. Read Data
# Read Data directly
t = "F:/NGHIEN CUU SINH/NCS - PHUONG ANH/Part 2-Satisfaction and Loyalty/So lieu/So lieu - PA xu ly/873_Loyaltyofbuspassenger_PAcode_NonBus_Mising data_outliers_forMLM_Analyse.csv"
DataLOY = read.csv(t, header = T)
head(DataLOY)## ID AGE CITY FRE TripPurpose Departure TimeUseonBus TravelTime PSSW PSSS PSAB
## 1 3 1 2 1 5 0 4 3.00 4.9 5.4 5.5
## 2 4 1 2 2 7 0 4 2.00 3.4 2.4 3.5
## 3 5 1 2 1 5 1 4 0.17 3.9 3.4 4.5
## 4 6 1 2 1 5 1 1 4.00 4.1 4.4 5.5
## 5 7 1 2 1 5 1 4 2.00 3.7 2.6 3.5
## 6 8 1 2 1 5 1 6 2.00 4.6 2.7 3.8
## PSEB PSQ SAT LOY IMA PHB PEV ATM PPI SIM PPA SBE EXB EC_Stop WC_Stop EC_Bus
## 1 5.8 4.8 6.0 5.7 5.6 6.6 4.0 4.3 6.0 3.5 4.5 4.0 4.8 2 1 2
## 2 6.0 4.6 4.7 4.3 4.8 4.0 5.3 4.1 4.0 4.0 4.0 4.8 4.2 2 2 1
## 3 4.5 2.7 2.0 3.7 3.0 5.0 5.8 3.1 2.7 3.5 4.5 6.0 4.0 2 2 2
## 4 6.0 4.8 5.0 5.0 5.0 5.0 5.0 5.0 5.0 5.0 5.0 5.0 5.0 2 1 2
## 5 4.5 4.5 4.0 4.6 4.6 5.6 6.0 3.3 3.0 5.5 5.0 5.0 4.2 2 1 2
## 6 4.0 3.9 4.7 4.4 3.0 5.4 5.3 2.7 3.0 2.0 2.5 2.0 5.0 2 1 2
## WC_Bus Gender MarriedStatus Occupation Education Income
## 1 1 2 1 1 2 1
## 2 2 2 1 1 2 1
## 3 2 1 1 1 2 1
## 4 1 1 1 1 2 1
## 5 1 1 1 1 2 1
## 6 1 2 1 1 3 1names(DataLOY)## [1] "ID" "AGE" "CITY" "FRE"
## [5] "TripPurpose" "Departure" "TimeUseonBus" "TravelTime"
## [9] "PSSW" "PSSS" "PSAB" "PSEB"
## [13] "PSQ" "SAT" "LOY" "IMA"
## [17] "PHB" "PEV" "ATM" "PPI"
## [21] "SIM" "PPA" "SBE" "EXB"
## [25] "EC_Stop" "WC_Stop" "EC_Bus" "WC_Bus"
## [29] "Gender" "MarriedStatus" "Occupation" "Education"
## [33] "Income"dim(DataLOY)## [1] 873 332. Desscriptive statistic
# 2.1 Recoding
# Regroup value of variables
DataLOY$SAT[DataLOY$SAT <2.5] <- 1
DataLOY$SAT[DataLOY$SAT >= 2.5 & DataLOY$SAT <5.5] <- 2
DataLOY$SAT[DataLOY$SAT >= 5.5] <- 3
DataLOY$LOY[DataLOY$LOY <2.5] <- 1
DataLOY$LOY[DataLOY$LOY >= 2.5 & DataLOY$LOY <5.5] <- 2
DataLOY$LOY[DataLOY$LOY >= 5.5] <- 3# 2.2. Change type of LOY, SAT and other factor variables
head(DataLOY)## ID AGE CITY FRE TripPurpose Departure TimeUseonBus TravelTime PSSW PSSS PSAB
## 1 3 1 2 1 5 0 4 3.00 4.9 5.4 5.5
## 2 4 1 2 2 7 0 4 2.00 3.4 2.4 3.5
## 3 5 1 2 1 5 1 4 0.17 3.9 3.4 4.5
## 4 6 1 2 1 5 1 1 4.00 4.1 4.4 5.5
## 5 7 1 2 1 5 1 4 2.00 3.7 2.6 3.5
## 6 8 1 2 1 5 1 6 2.00 4.6 2.7 3.8
## PSEB PSQ SAT LOY IMA PHB PEV ATM PPI SIM PPA SBE EXB EC_Stop WC_Stop EC_Bus
## 1 5.8 4.8 3 3 5.6 6.6 4.0 4.3 6.0 3.5 4.5 4.0 4.8 2 1 2
## 2 6.0 4.6 2 2 4.8 4.0 5.3 4.1 4.0 4.0 4.0 4.8 4.2 2 2 1
## 3 4.5 2.7 1 2 3.0 5.0 5.8 3.1 2.7 3.5 4.5 6.0 4.0 2 2 2
## 4 6.0 4.8 2 2 5.0 5.0 5.0 5.0 5.0 5.0 5.0 5.0 5.0 2 1 2
## 5 4.5 4.5 2 2 4.6 5.6 6.0 3.3 3.0 5.5 5.0 5.0 4.2 2 1 2
## 6 4.0 3.9 2 2 3.0 5.4 5.3 2.7 3.0 2.0 2.5 2.0 5.0 2 1 2
## WC_Bus Gender MarriedStatus Occupation Education Income
## 1 1 2 1 1 2 1
## 2 2 2 1 1 2 1
## 3 2 1 1 1 2 1
## 4 1 1 1 1 2 1
## 5 1 1 1 1 2 1
## 6 1 2 1 1 3 1str(DataLOY)## 'data.frame': 873 obs. of 33 variables:
## $ ID : int 3 4 5 6 7 8 10 11 12 13 ...
## $ AGE : int 1 1 1 1 1 1 4 1 1 2 ...
## $ CITY : int 2 2 2 2 2 2 2 2 2 2 ...
## $ FRE : int 1 2 1 1 1 1 1 1 1 1 ...
## $ TripPurpose : int 5 7 5 5 5 5 4 5 5 5 ...
## $ Departure : int 0 0 1 1 1 1 1 1 0 1 ...
## $ TimeUseonBus : int 4 4 4 1 4 6 4 4 3 3 ...
## $ TravelTime : num 3 2 0.17 4 2 2 2 2.5 1.5 2 ...
## $ PSSW : num 4.9 3.4 3.9 4.1 3.7 4.6 3.9 3.4 5.3 4 ...
## $ PSSS : num 5.4 2.4 3.4 4.4 2.6 2.7 4 2.7 3.1 4.7 ...
## $ PSAB : num 5.5 3.5 4.5 5.5 3.5 3.8 6 3.5 4.8 5.5 ...
## $ PSEB : num 5.8 6 4.5 6 4.5 4 4.5 6.3 7 4.5 ...
## $ PSQ : num 4.8 4.6 2.7 4.8 4.5 3.9 4.2 4.1 6 4.5 ...
## $ SAT : num 3 2 1 2 2 2 3 2 3 3 ...
## $ LOY : num 3 2 2 2 2 2 3 2 3 3 ...
## $ IMA : num 5.6 4.8 3 5 4.6 3 5.8 5.8 6.8 6 ...
## $ PHB : num 6.6 4 5 5 5.6 5.4 5.8 5.6 7 6.4 ...
## $ PEV : num 4 5.3 5.8 5 6 5.3 6.3 6.5 7 5.3 ...
## $ ATM : num 4.3 4.1 3.1 5 3.3 2.7 4 4.1 4.4 3.9 ...
## $ PPI : num 6 4 2.7 5 3 3 6 2 3.3 5.3 ...
## $ SIM : num 3.5 4 3.5 5 5.5 2 6 5.3 4.3 6 ...
## $ PPA : num 4.5 4 4.5 5 5 2.5 4 4 4 4.3 ...
## $ SBE : num 4 4.8 6 5 5 2 5.3 4 5 3.8 ...
## $ EXB : num 4.8 4.2 4 5 4.2 5 5.3 5.3 5.3 5 ...
## $ EC_Stop : int 2 2 2 2 2 2 2 2 2 2 ...
## $ WC_Stop : int 1 2 2 1 1 1 2 1 2 2 ...
## $ EC_Bus : int 2 1 2 2 2 2 2 2 2 2 ...
## $ WC_Bus : int 1 2 2 1 1 1 2 1 2 2 ...
## $ Gender : int 2 2 1 1 1 2 2 2 2 1 ...
## $ MarriedStatus: int 1 1 1 1 1 1 2 1 1 1 ...
## $ Occupation : int 1 1 1 1 1 1 2 1 1 7 ...
## $ Education : int 2 2 2 2 2 3 1 2 3 5 ...
## $ Income : int 1 1 1 1 1 1 1 1 1 1 ...attach(DataLOY)
DataLOY = within(DataLOY, {
AGE = factor(AGE, labels = c("16-25", "26-35", "36-45", "46-55", ">55"))
CITY = factor(CITY,labels = c("DaNang", "HoChiMinh"))
FRE = factor(FRE, labels = c(">=3 days/week", "2days/month-2days/week", "2days/year-1day/month", "<2 days/year"))
TripPurpose = factor(TripPurpose, labels = c("Working", "Studying", "Shopping", "Entertaining", "Others"))
Departure = factor(Departure, labels = c("Normal", "Peak-Hour"))
TimeUseonBus = factor(TimeUseonBus, labels = c("Using.telephone", "Reading", "Listening", "Nothing", "Talking", "Others"))
SAT = factor(SAT, labels = c("Not satisfied", "Normal", "Satisfied"))
LOY = factor(LOY, labels = c("Notloyal", "Normal", "Loyal"))
EC_Stop = factor(EC_Stop, labels = c("Ever", "Never"))
WC_Stop = factor(WC_Stop, labels = c("Ever", "Never"))
EC_Bus = factor(EC_Bus, labels = c("Ever", "Never"))
WC_Bus = factor(WC_Bus, labels = c("Ever", "Never"))
Gender = factor(Gender, labels = c("Male", "Female"))
MarriedStatus = factor(MarriedStatus, labels = c("Single", "Married"))
Occupation = factor(Occupation, labels = c("Students/Pupils", "Full.time.job", "Part.time.job", "Retirement", "No.job", "Housewife", "Others"))
Education = factor(Education, labels = c("Secondary.school", "Undergraduate", "High.school", "Postgraduate", "Others"))
Income = factor(Income, labels = c("<5millions", "5-10millions", "10-15millions", ">15millions"))
} )
str(DataLOY)## 'data.frame': 873 obs. of 33 variables:
## $ ID : int 3 4 5 6 7 8 10 11 12 13 ...
## $ AGE : Factor w/ 5 levels "16-25","26-35",..: 1 1 1 1 1 1 3 1 1 2 ...
## $ CITY : Factor w/ 2 levels "DaNang","HoChiMinh": 2 2 2 2 2 2 2 2 2 2 ...
## $ FRE : Factor w/ 4 levels ">=3 days/week",..: 1 2 1 1 1 1 1 1 1 1 ...
## $ TripPurpose : Factor w/ 5 levels "Working","Studying",..: 2 4 2 2 2 2 1 2 2 2 ...
## $ Departure : Factor w/ 2 levels "Normal","Peak-Hour": 1 1 2 2 2 2 2 2 1 2 ...
## $ TimeUseonBus : Factor w/ 6 levels "Using.telephone",..: 4 4 4 1 4 6 4 4 3 3 ...
## $ TravelTime : num 3 2 0.17 4 2 2 2 2.5 1.5 2 ...
## $ PSSW : num 4.9 3.4 3.9 4.1 3.7 4.6 3.9 3.4 5.3 4 ...
## $ PSSS : num 5.4 2.4 3.4 4.4 2.6 2.7 4 2.7 3.1 4.7 ...
## $ PSAB : num 5.5 3.5 4.5 5.5 3.5 3.8 6 3.5 4.8 5.5 ...
## $ PSEB : num 5.8 6 4.5 6 4.5 4 4.5 6.3 7 4.5 ...
## $ PSQ : num 4.8 4.6 2.7 4.8 4.5 3.9 4.2 4.1 6 4.5 ...
## $ SAT : Factor w/ 3 levels "Not satisfied",..: 3 2 1 2 2 2 3 2 3 3 ...
## $ LOY : Factor w/ 3 levels "Notloyal","Normal",..: 3 2 2 2 2 2 3 2 3 3 ...
## $ IMA : num 5.6 4.8 3 5 4.6 3 5.8 5.8 6.8 6 ...
## $ PHB : num 6.6 4 5 5 5.6 5.4 5.8 5.6 7 6.4 ...
## $ PEV : num 4 5.3 5.8 5 6 5.3 6.3 6.5 7 5.3 ...
## $ ATM : num 4.3 4.1 3.1 5 3.3 2.7 4 4.1 4.4 3.9 ...
## $ PPI : num 6 4 2.7 5 3 3 6 2 3.3 5.3 ...
## $ SIM : num 3.5 4 3.5 5 5.5 2 6 5.3 4.3 6 ...
## $ PPA : num 4.5 4 4.5 5 5 2.5 4 4 4 4.3 ...
## $ SBE : num 4 4.8 6 5 5 2 5.3 4 5 3.8 ...
## $ EXB : num 4.8 4.2 4 5 4.2 5 5.3 5.3 5.3 5 ...
## $ EC_Stop : Factor w/ 2 levels "Ever","Never": 2 2 2 2 2 2 2 2 2 2 ...
## $ WC_Stop : Factor w/ 2 levels "Ever","Never": 1 2 2 1 1 1 2 1 2 2 ...
## $ EC_Bus : Factor w/ 2 levels "Ever","Never": 2 1 2 2 2 2 2 2 2 2 ...
## $ WC_Bus : Factor w/ 2 levels "Ever","Never": 1 2 2 1 1 1 2 1 2 2 ...
## $ Gender : Factor w/ 2 levels "Male","Female": 2 2 1 1 1 2 2 2 2 1 ...
## $ MarriedStatus: Factor w/ 2 levels "Single","Married": 1 1 1 1 1 1 2 1 1 1 ...
## $ Occupation : Factor w/ 7 levels "Students/Pupils",..: 1 1 1 1 1 1 2 1 1 7 ...
## $ Education : Factor w/ 5 levels "Secondary.school",..: 2 2 2 2 2 3 1 2 3 5 ...
## $ Income : Factor w/ 4 levels "<5millions","5-10millions",..: 1 1 1 1 1 1 1 1 1 1 ...dim(DataLOY)## [1] 873 33# 2.3. Descritive Table
library(tableone)
require(tableone)
library(magrittr)
summary(DataLOY)## ID AGE CITY FRE
## Min. : 3.0 16-25:425 DaNang :422 >=3 days/week :508
## 1st Qu.:273.0 26-35:172 HoChiMinh:451 2days/month-2days/week:168
## Median :526.0 36-45:105 2days/year-1day/month : 99
## Mean :521.7 46-55: 77 <2 days/year : 98
## 3rd Qu.:769.0 >55 : 94
## Max. :993.0
##
## TripPurpose Departure TimeUseonBus TravelTime
## Working :305 Normal :296 Using.telephone:198 Min. : 0.000
## Studying :303 Peak-Hour:577 Reading : 53 1st Qu.: 0.500
## Shopping : 60 Listening :138 Median : 1.000
## Entertaining:100 Nothing :428 Mean : 1.291
## Others :105 Talking : 34 3rd Qu.: 2.000
## Others : 22 Max. :20.000
##
## PSSW PSSS PSAB PSEB
## Min. :1.000 Min. :1.000 Min. :1.000 Min. :1.000
## 1st Qu.:4.300 1st Qu.:3.700 1st Qu.:4.000 1st Qu.:5.500
## Median :4.900 Median :4.600 Median :5.500 Median :6.000
## Mean :4.768 Mean :4.478 Mean :5.047 Mean :5.786
## 3rd Qu.:5.400 3rd Qu.:5.400 3rd Qu.:6.000 3rd Qu.:6.300
## Max. :7.000 Max. :7.000 Max. :7.000 Max. :7.000
##
## PSQ SAT LOY IMA
## Min. :1.000 Not satisfied: 16 Notloyal: 12 Min. :1.000
## 1st Qu.:4.800 Normal :347 Normal :342 1st Qu.:4.800
## Median :5.400 Satisfied :510 Loyal :519 Median :5.600
## Mean :5.226 Mean :5.338
## 3rd Qu.:5.800 3rd Qu.:6.000
## Max. :7.000 Max. :7.000
##
## PHB PEV ATM PPI
## Min. :1.000 Min. :1.000 Min. :1.000 Min. :1.000
## 1st Qu.:5.200 1st Qu.:5.300 1st Qu.:4.300 1st Qu.:3.700
## Median :5.800 Median :6.000 Median :5.100 Median :4.300
## Mean :5.611 Mean :5.619 Mean :4.979 Mean :4.373
## 3rd Qu.:6.200 3rd Qu.:6.300 3rd Qu.:5.900 3rd Qu.:5.300
## Max. :7.000 Max. :7.000 Max. :7.000 Max. :7.000
##
## SIM PPA SBE EXB EC_Stop
## Min. :1.000 Min. :1.000 Min. :1.80 Min. :2.00 Ever : 36
## 1st Qu.:4.000 1st Qu.:4.000 1st Qu.:4.30 1st Qu.:5.00 Never:837
## Median :5.000 Median :5.000 Median :5.30 Median :5.80
## Mean :4.901 Mean :4.957 Mean :5.18 Mean :5.59
## 3rd Qu.:6.000 3rd Qu.:5.800 3rd Qu.:6.00 3rd Qu.:6.00
## Max. :7.000 Max. :7.000 Max. :7.00 Max. :7.00
##
## WC_Stop EC_Bus WC_Bus Gender MarriedStatus
## Ever :121 Ever : 42 Ever :116 Male :364 Single :535
## Never:752 Never:831 Never:757 Female:509 Married:338
##
##
##
##
##
## Occupation Education Income
## Students/Pupils:368 Secondary.school: 57 <5millions :465
## Full.time.job :305 Undergraduate :287 5-10millions :249
## Part.time.job : 69 High.school :368 10-15millions:122
## Retirement : 46 Postgraduate :106 >15millions : 37
## No.job : 3 Others : 55
## Housewife : 54
## Others : 28library(table1)##
## Attaching package: 'table1'## The following objects are masked from 'package:base':
##
## units, units<-Tab1_LOY <- table1(~ PSSW + PSSS + PSAB + PSEB + PSQ + SAT + IMA + PHB + PEV + ATM + PPI + SIM + PPA + SBE + EXB + EC_Stop + WC_Stop + EC_Bus + WC_Bus + Gender + MarriedStatus + Occupation + Education + Income + AGE + CITY + FRE + TripPurpose + Departure + TimeUseonBus + TravelTime| LOY , data = DataLOY)
Tab1_LOY| Notloyal (N=12) |
Normal (N=342) |
Loyal (N=519) |
Overall (N=873) |
|
|---|---|---|---|---|
| PSSW | ||||
| Mean (SD) | 3.48 (1.29) | 4.40 (0.996) | 5.04 (0.908) | 4.77 (1.01) |
| Median [Min, Max] | 3.25 [2.00, 6.40] | 4.60 [1.00, 7.00] | 5.10 [1.00, 7.00] | 4.90 [1.00, 7.00] |
| PSSS | ||||
| Mean (SD) | 3.08 (1.04) | 4.02 (1.02) | 4.81 (1.15) | 4.48 (1.17) |
| Median [Min, Max] | 2.95 [2.00, 5.30] | 4.00 [1.00, 6.40] | 4.90 [1.00, 7.00] | 4.60 [1.00, 7.00] |
| PSAB | ||||
| Mean (SD) | 3.53 (1.61) | 4.51 (1.28) | 5.44 (1.07) | 5.05 (1.26) |
| Median [Min, Max] | 3.00 [2.00, 7.00] | 4.50 [1.00, 7.00] | 5.80 [1.00, 7.00] | 5.50 [1.00, 7.00] |
| PSEB | ||||
| Mean (SD) | 4.49 (1.46) | 5.44 (1.07) | 6.04 (0.727) | 5.79 (0.948) |
| Median [Min, Max] | 4.50 [2.30, 7.00] | 5.80 [1.00, 7.00] | 6.00 [1.00, 7.00] | 6.00 [1.00, 7.00] |
| PSQ | ||||
| Mean (SD) | 3.06 (0.986) | 4.71 (0.834) | 5.61 (0.607) | 5.23 (0.871) |
| Median [Min, Max] | 3.10 [1.20, 4.30] | 4.80 [1.00, 7.00] | 5.70 [2.90, 7.00] | 5.40 [1.00, 7.00] |
| SAT | ||||
| Not satisfied | 6 (50.0%) | 10 (2.9%) | 0 (0%) | 16 (1.8%) |
| Normal | 6 (50.0%) | 254 (74.3%) | 87 (16.8%) | 347 (39.7%) |
| Satisfied | 0 (0%) | 78 (22.8%) | 432 (83.2%) | 510 (58.4%) |
| IMA | ||||
| Mean (SD) | 3.75 (1.29) | 4.75 (0.971) | 5.76 (0.660) | 5.34 (0.964) |
| Median [Min, Max] | 3.40 [2.20, 6.00] | 4.80 [1.00, 7.00] | 6.00 [1.60, 7.00] | 5.60 [1.00, 7.00] |
| PHB | ||||
| Mean (SD) | 4.28 (1.38) | 5.05 (1.10) | 6.01 (0.666) | 5.61 (1.00) |
| Median [Min, Max] | 4.30 [2.00, 6.00] | 5.20 [1.00, 7.00] | 6.00 [2.00, 7.00] | 5.80 [1.00, 7.00] |
| PEV | ||||
| Mean (SD) | 4.97 (1.47) | 5.08 (1.38) | 5.99 (0.906) | 5.62 (1.21) |
| Median [Min, Max] | 5.75 [2.00, 6.50] | 5.30 [1.00, 7.00] | 6.00 [1.00, 7.00] | 6.00 [1.00, 7.00] |
| ATM | ||||
| Mean (SD) | 3.38 (1.43) | 4.37 (1.03) | 5.42 (1.02) | 4.98 (1.16) |
| Median [Min, Max] | 3.35 [1.00, 6.00] | 4.40 [1.00, 6.40] | 5.70 [1.00, 7.00] | 5.10 [1.00, 7.00] |
| PPI | ||||
| Mean (SD) | 3.37 (1.47) | 3.95 (1.33) | 4.67 (1.39) | 4.37 (1.41) |
| Median [Min, Max] | 2.85 [1.70, 6.00] | 4.00 [1.00, 7.00] | 5.00 [1.00, 7.00] | 4.30 [1.00, 7.00] |
| SIM | ||||
| Mean (SD) | 4.08 (1.63) | 4.42 (1.07) | 5.24 (1.05) | 4.90 (1.14) |
| Median [Min, Max] | 4.00 [1.80, 6.00] | 4.30 [1.00, 7.00] | 5.50 [1.00, 7.00] | 5.00 [1.00, 7.00] |
| PPA | ||||
| Mean (SD) | 3.96 (1.30) | 4.60 (0.876) | 5.21 (1.02) | 4.96 (1.02) |
| Median [Min, Max] | 4.00 [2.00, 6.00] | 4.50 [1.00, 7.00] | 5.50 [1.00, 7.00] | 5.00 [1.00, 7.00] |
| SBE | ||||
| Mean (SD) | 4.32 (1.23) | 4.88 (0.990) | 5.40 (1.02) | 5.18 (1.05) |
| Median [Min, Max] | 4.15 [2.00, 6.00] | 5.00 [1.80, 7.00] | 5.80 [1.80, 7.00] | 5.30 [1.80, 7.00] |
| EXB | ||||
| Mean (SD) | 3.63 (1.10) | 5.03 (0.837) | 6.00 (0.595) | 5.59 (0.881) |
| Median [Min, Max] | 3.65 [2.00, 6.00] | 5.00 [2.00, 7.00] | 6.00 [3.70, 7.00] | 5.80 [2.00, 7.00] |
| EC_Stop | ||||
| Ever | 1 (8.3%) | 21 (6.1%) | 14 (2.7%) | 36 (4.1%) |
| Never | 11 (91.7%) | 321 (93.9%) | 505 (97.3%) | 837 (95.9%) |
| WC_Stop | ||||
| Ever | 2 (16.7%) | 65 (19.0%) | 54 (10.4%) | 121 (13.9%) |
| Never | 10 (83.3%) | 277 (81.0%) | 465 (89.6%) | 752 (86.1%) |
| EC_Bus | ||||
| Ever | 2 (16.7%) | 20 (5.8%) | 20 (3.9%) | 42 (4.8%) |
| Never | 10 (83.3%) | 322 (94.2%) | 499 (96.1%) | 831 (95.2%) |
| WC_Bus | ||||
| Ever | 2 (16.7%) | 58 (17.0%) | 56 (10.8%) | 116 (13.3%) |
| Never | 10 (83.3%) | 284 (83.0%) | 463 (89.2%) | 757 (86.7%) |
| Gender | ||||
| Male | 6 (50.0%) | 152 (44.4%) | 206 (39.7%) | 364 (41.7%) |
| Female | 6 (50.0%) | 190 (55.6%) | 313 (60.3%) | 509 (58.3%) |
| MarriedStatus | ||||
| Single | 8 (66.7%) | 247 (72.2%) | 280 (53.9%) | 535 (61.3%) |
| Married | 4 (33.3%) | 95 (27.8%) | 239 (46.1%) | 338 (38.7%) |
| Occupation | ||||
| Students/Pupils | 7 (58.3%) | 178 (52.0%) | 183 (35.3%) | 368 (42.2%) |
| Full.time.job | 5 (41.7%) | 118 (34.5%) | 182 (35.1%) | 305 (34.9%) |
| Part.time.job | 0 (0%) | 19 (5.6%) | 50 (9.6%) | 69 (7.9%) |
| Retirement | 0 (0%) | 9 (2.6%) | 37 (7.1%) | 46 (5.3%) |
| No.job | 0 (0%) | 0 (0%) | 3 (0.6%) | 3 (0.3%) |
| Housewife | 0 (0%) | 12 (3.5%) | 42 (8.1%) | 54 (6.2%) |
| Others | 0 (0%) | 6 (1.8%) | 22 (4.2%) | 28 (3.2%) |
| Education | ||||
| Secondary.school | 0 (0%) | 14 (4.1%) | 43 (8.3%) | 57 (6.5%) |
| Undergraduate | 2 (16.7%) | 110 (32.2%) | 175 (33.7%) | 287 (32.9%) |
| High.school | 10 (83.3%) | 157 (45.9%) | 201 (38.7%) | 368 (42.2%) |
| Postgraduate | 0 (0%) | 43 (12.6%) | 63 (12.1%) | 106 (12.1%) |
| Others | 0 (0%) | 18 (5.3%) | 37 (7.1%) | 55 (6.3%) |
| Income | ||||
| <5millions | 5 (41.7%) | 187 (54.7%) | 273 (52.6%) | 465 (53.3%) |
| 5-10millions | 3 (25.0%) | 101 (29.5%) | 145 (27.9%) | 249 (28.5%) |
| 10-15millions | 3 (25.0%) | 39 (11.4%) | 80 (15.4%) | 122 (14.0%) |
| >15millions | 1 (8.3%) | 15 (4.4%) | 21 (4.0%) | 37 (4.2%) |
| AGE | ||||
| 16-25 | 8 (66.7%) | 214 (62.6%) | 203 (39.1%) | 425 (48.7%) |
| 26-35 | 2 (16.7%) | 68 (19.9%) | 102 (19.7%) | 172 (19.7%) |
| 36-45 | 1 (8.3%) | 27 (7.9%) | 77 (14.8%) | 105 (12.0%) |
| 46-55 | 1 (8.3%) | 21 (6.1%) | 55 (10.6%) | 77 (8.8%) |
| >55 | 0 (0%) | 12 (3.5%) | 82 (15.8%) | 94 (10.8%) |
| CITY | ||||
| DaNang | 2 (16.7%) | 136 (39.8%) | 284 (54.7%) | 422 (48.3%) |
| HoChiMinh | 10 (83.3%) | 206 (60.2%) | 235 (45.3%) | 451 (51.7%) |
| FRE | ||||
| >=3 days/week | 4 (33.3%) | 179 (52.3%) | 325 (62.6%) | 508 (58.2%) |
| 2days/month-2days/week | 2 (16.7%) | 62 (18.1%) | 104 (20.0%) | 168 (19.2%) |
| 2days/year-1day/month | 2 (16.7%) | 49 (14.3%) | 48 (9.2%) | 99 (11.3%) |
| <2 days/year | 4 (33.3%) | 52 (15.2%) | 42 (8.1%) | 98 (11.2%) |
| TripPurpose | ||||
| Working | 4 (33.3%) | 114 (33.3%) | 187 (36.0%) | 305 (34.9%) |
| Studying | 4 (33.3%) | 135 (39.5%) | 164 (31.6%) | 303 (34.7%) |
| Shopping | 0 (0%) | 13 (3.8%) | 47 (9.1%) | 60 (6.9%) |
| Entertaining | 1 (8.3%) | 45 (13.2%) | 54 (10.4%) | 100 (11.5%) |
| Others | 3 (25.0%) | 35 (10.2%) | 67 (12.9%) | 105 (12.0%) |
| Departure | ||||
| Normal | 5 (41.7%) | 105 (30.7%) | 186 (35.8%) | 296 (33.9%) |
| Peak-Hour | 7 (58.3%) | 237 (69.3%) | 333 (64.2%) | 577 (66.1%) |
| TimeUseonBus | ||||
| Using.telephone | 3 (25.0%) | 79 (23.1%) | 116 (22.4%) | 198 (22.7%) |
| Reading | 1 (8.3%) | 19 (5.6%) | 33 (6.4%) | 53 (6.1%) |
| Listening | 1 (8.3%) | 64 (18.7%) | 73 (14.1%) | 138 (15.8%) |
| Nothing | 5 (41.7%) | 156 (45.6%) | 267 (51.4%) | 428 (49.0%) |
| Talking | 1 (8.3%) | 11 (3.2%) | 22 (4.2%) | 34 (3.9%) |
| Others | 1 (8.3%) | 13 (3.8%) | 8 (1.5%) | 22 (2.5%) |
| TravelTime | ||||
| Mean (SD) | 1.67 (2.69) | 1.35 (1.65) | 1.24 (1.29) | 1.29 (1.46) |
| Median [Min, Max] | 1.00 [0, 10.0] | 1.00 [0, 20.0] | 1.00 [0.100, 20.0] | 1.00 [0, 20.0] |
library(compareGroups)
Des_LOY <- compareGroups(LOY ~ PSSW + PSSS + PSAB + PSEB + PSQ + SAT + IMA + PHB + PEV + ATM + PPI + SIM + PPA + SBE + EXB + EC_Stop + WC_Stop + EC_Bus + WC_Bus + Gender + MarriedStatus + Occupation + Education + Income + AGE + CITY + FRE + TripPurpose + Departure + TimeUseonBus + TravelTime, data = DataLOY)
createTable(Des_LOY)##
## --------Summary descriptives table by 'LOY'---------
##
## ________________________________________________________________________
## Notloyal Normal Loyal p.overall
## N=12 N=342 N=519
## ¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯
## PSSW 3.48 (1.29) 4.40 (1.00) 5.04 (0.91) <0.001
## PSSS 3.08 (1.04) 4.02 (1.02) 4.81 (1.15) <0.001
## PSAB 3.53 (1.61) 4.51 (1.28) 5.44 (1.07) <0.001
## PSEB 4.49 (1.46) 5.44 (1.07) 6.04 (0.73) <0.001
## PSQ 3.06 (0.99) 4.71 (0.83) 5.61 (0.61) <0.001
## SAT: <0.001
## Not satisfied 6 (50.0%) 10 (2.92%) 0 (0.00%)
## Normal 6 (50.0%) 254 (74.3%) 87 (16.8%)
## Satisfied 0 (0.00%) 78 (22.8%) 432 (83.2%)
## IMA 3.75 (1.29) 4.75 (0.97) 5.76 (0.66) <0.001
## PHB 4.28 (1.38) 5.05 (1.10) 6.01 (0.67) <0.001
## PEV 4.97 (1.47) 5.08 (1.38) 5.99 (0.91) <0.001
## ATM 3.38 (1.43) 4.37 (1.03) 5.42 (1.02) <0.001
## PPI 3.37 (1.47) 3.95 (1.33) 4.67 (1.39) <0.001
## SIM 4.08 (1.63) 4.42 (1.07) 5.24 (1.05) <0.001
## PPA 3.96 (1.30) 4.60 (0.88) 5.21 (1.02) <0.001
## SBE 4.32 (1.23) 4.88 (0.99) 5.40 (1.02) <0.001
## EXB 3.63 (1.10) 5.03 (0.84) 6.00 (0.60) <0.001
## EC_Stop: 0.026
## Ever 1 (8.33%) 21 (6.14%) 14 (2.70%)
## Never 11 (91.7%) 321 (93.9%) 505 (97.3%)
## WC_Stop: 0.001
## Ever 2 (16.7%) 65 (19.0%) 54 (10.4%)
## Never 10 (83.3%) 277 (81.0%) 465 (89.6%)
## EC_Bus: 0.058
## Ever 2 (16.7%) 20 (5.85%) 20 (3.85%)
## Never 10 (83.3%) 322 (94.2%) 499 (96.1%)
## WC_Bus: 0.024
## Ever 2 (16.7%) 58 (17.0%) 56 (10.8%)
## Never 10 (83.3%) 284 (83.0%) 463 (89.2%)
## Gender: 0.323
## Male 6 (50.0%) 152 (44.4%) 206 (39.7%)
## Female 6 (50.0%) 190 (55.6%) 313 (60.3%)
## MarriedStatus: <0.001
## Single 8 (66.7%) 247 (72.2%) 280 (53.9%)
## Married 4 (33.3%) 95 (27.8%) 239 (46.1%)
## Occupation: .
## Students/Pupils 7 (58.3%) 178 (52.0%) 183 (35.3%)
## Full.time.job 5 (41.7%) 118 (34.5%) 182 (35.1%)
## Part.time.job 0 (0.00%) 19 (5.56%) 50 (9.63%)
## Retirement 0 (0.00%) 9 (2.63%) 37 (7.13%)
## No.job 0 (0.00%) 0 (0.00%) 3 (0.58%)
## Housewife 0 (0.00%) 12 (3.51%) 42 (8.09%)
## Others 0 (0.00%) 6 (1.75%) 22 (4.24%)
## Education: .
## Secondary.school 0 (0.00%) 14 (4.09%) 43 (8.29%)
## Undergraduate 2 (16.7%) 110 (32.2%) 175 (33.7%)
## High.school 10 (83.3%) 157 (45.9%) 201 (38.7%)
## Postgraduate 0 (0.00%) 43 (12.6%) 63 (12.1%)
## Others 0 (0.00%) 18 (5.26%) 37 (7.13%)
## Income: .
## <5millions 5 (41.7%) 187 (54.7%) 273 (52.6%)
## 5-10millions 3 (25.0%) 101 (29.5%) 145 (27.9%)
## 10-15millions 3 (25.0%) 39 (11.4%) 80 (15.4%)
## >15millions 1 (8.33%) 15 (4.39%) 21 (4.05%)
## AGE: .
## 16-25 8 (66.7%) 214 (62.6%) 203 (39.1%)
## 26-35 2 (16.7%) 68 (19.9%) 102 (19.7%)
## 36-45 1 (8.33%) 27 (7.89%) 77 (14.8%)
## 46-55 1 (8.33%) 21 (6.14%) 55 (10.6%)
## >55 0 (0.00%) 12 (3.51%) 82 (15.8%)
## CITY: <0.001
## DaNang 2 (16.7%) 136 (39.8%) 284 (54.7%)
## HoChiMinh 10 (83.3%) 206 (60.2%) 235 (45.3%)
## FRE: .
## >=3 days/week 4 (33.3%) 179 (52.3%) 325 (62.6%)
## 2days/month-2days/week 2 (16.7%) 62 (18.1%) 104 (20.0%)
## 2days/year-1day/month 2 (16.7%) 49 (14.3%) 48 (9.25%)
## <2 days/year 4 (33.3%) 52 (15.2%) 42 (8.09%)
## TripPurpose: .
## Working 4 (33.3%) 114 (33.3%) 187 (36.0%)
## Studying 4 (33.3%) 135 (39.5%) 164 (31.6%)
## Shopping 0 (0.00%) 13 (3.80%) 47 (9.06%)
## Entertaining 1 (8.33%) 45 (13.2%) 54 (10.4%)
## Others 3 (25.0%) 35 (10.2%) 67 (12.9%)
## Departure: 0.244
## Normal 5 (41.7%) 105 (30.7%) 186 (35.8%)
## Peak-Hour 7 (58.3%) 237 (69.3%) 333 (64.2%)
## TimeUseonBus: .
## Using.telephone 3 (25.0%) 79 (23.1%) 116 (22.4%)
## Reading 1 (8.33%) 19 (5.56%) 33 (6.36%)
## Listening 1 (8.33%) 64 (18.7%) 73 (14.1%)
## Nothing 5 (41.7%) 156 (45.6%) 267 (51.4%)
## Talking 1 (8.33%) 11 (3.22%) 22 (4.24%)
## Others 1 (8.33%) 13 (3.80%) 8 (1.54%)
## TravelTime 1.67 (2.69) 1.35 (1.65) 1.24 (1.29) 0.393
## ¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯3. Describe Data by graph
library(magrittr)
library(tidyverse)## -- Attaching packages --------------------------------------- tidyverse 1.3.0 --## v ggplot2 3.3.2 v purrr 0.3.4
## v tibble 3.0.4 v dplyr 1.0.2
## v tidyr 1.1.2 v stringr 1.4.0
## v readr 1.4.0 v forcats 0.5.0## -- Conflicts ------------------------------------------ tidyverse_conflicts() --
## x tidyr::extract() masks magrittr::extract()
## x dplyr::filter() masks stats::filter()
## x dplyr::lag() masks stats::lag()
## x purrr::set_names() masks magrittr::set_names()library(ggplot2)
library(car)## Warning: package 'car' was built under R version 4.0.4## Loading required package: carData##
## Attaching package: 'car'## The following object is masked from 'package:dplyr':
##
## recode## The following object is masked from 'package:purrr':
##
## somelibrary(psych)##
## Attaching package: 'psych'## The following object is masked from 'package:car':
##
## logit## The following objects are masked from 'package:ggplot2':
##
## %+%, alphalibrary(gridExtra)##
## Attaching package: 'gridExtra'## The following object is masked from 'package:dplyr':
##
## combinehist(DataLOY$PSSW)
# Correlations between LOY and continuous variables of users' perception
Cor1_1 = data.frame(DataLOY$LOY, DataLOY$PSSW, DataLOY$PSSS, DataLOY$PSAB, DataLOY$PSEB, DataLOY$PSQ, DataLOY$SAT, DataLOY$IMA)
pairs.panels(Cor1_1)
Cor1_2 = data.frame(DataLOY$LOY, DataLOY$PHB, DataLOY$PEV, DataLOY$ATM, DataLOY$PPI, DataLOY$SIM, DataLOY$PPA, DataLOY$SBE, DataLOY$EXB)
pairs.panels(Cor1_2)
# Correlation between LOY and variables of users experiences
Cor2 = data.frame(DataLOY$LOY, DataLOY$EC_Stop, DataLOY$WC_Stop, DataLOY$EC_Bus, DataLOY$WC_Bus)
pairs.panels(Cor2)
# Correlation between LOY and variables of travel characteristics
Cor3 = data.frame(DataLOY$LOY, DataLOY$CITY, DataLOY$FRE, DataLOY$TripPurpose, DataLOY$Departure, DataLOY$TimeUseonBus, DataLOY$TravelTime)
pairs.panels(Cor3)
# Correlation between LOY and variables of socioeconomics
Cor4 = data.frame(DataLOY$LOY, DataLOY$Gender, DataLOY$MarriedStatus, DataLOY$Occupation, DataLOY$Education, DataLOY$Income, DataLOY$AGE)
pairs.panels(Cor4)
# Boxplot of variables/CITY
# Chia cot theo cach khac : library(gridExtra) ; grid.arrange(p1, p2, ncol=2)
par(mfrow = c(1,2))
boxplot (DataLOY$PSSW ~ DataLOY$CITY, main = "Perceived Security & Safety on the way to/from bus stops", xlab = "PSSW", ylab = "Point (1-7)", col = c("Red", "blue"), notch = T, names = c("Da Nang", "Ho Chi Minh"))
boxplot (DataLOY$PSSW ~ DataLOY$Gender, main = "Perceived Security & Safety on the way to/from bus stops", xlab = "PSSW", ylab = "Point (1-7)", col = c("Red", "blue"), notch = T, names = c("Male", "Female"))
boxplot (DataLOY$PSSS ~ DataLOY$CITY, main = "Perceived Security & Safety at bus Stations", xlab = "PSSS", ylab = "Point (1-7)", col = c("Red", "blue"), notch = T, names = c("Da Nang", "Ho Chi Minh"))
boxplot (DataLOY$PSSS ~ DataLOY$Gender, main = "Perceived Security & Safety at bus Stations", xlab = "PSSS", ylab = "Point (1-7)", col = c("Red", "blue"), notch = T, names = c("Male", "Female"))
boxplot (DataLOY$PSAB ~ DataLOY$CITY, main = "Perceived Safety on Buses", xlab = "PSAB", ylab = "Point (1-7)", col = c("Red", "blue"), notch = T, names = c("Da Nang", "Ho Chi Minh"))
boxplot (DataLOY$PSAB ~ DataLOY$Gender, main = "Perceived Safety on Buses", xlab = "PSAB", ylab = "Point (1-7)", col = c("Red", "blue"), notch = T, names = c("Male", "Female"))
boxplot (DataLOY$PSEB ~ DataLOY$CITY, main = "Perceived Security on Buses", xlab = "PSEB", ylab = "Point (1-7)", col = c("Red", "blue"), notch = T, names = c("Da Nang", "Ho Chi Minh"))## Warning in bxp(list(stats = structure(c(4.8, 5.8, 6, 6.5, 7, 3.5, 5, 6, : some
## notches went outside hinges ('box'): maybe set notch=FALSEboxplot (DataLOY$PSEB ~ DataLOY$Gender, main = "Perceived Security on Buses", xlab = "PSEB", ylab = "Point (1-7)", col = c("Red", "blue"), notch = T, names = c("Male", "Female"))
boxplot (DataLOY$PSQ ~ DataLOY$CITY, main = "Perceived Service Quality", xlab = "PSQ", ylab = "Point (1-7)", col = c("Red", "blue"), notch = T, names = c("Da Nang", "Ho Chi Minh"))
boxplot (DataLOY$PSQ ~ DataLOY$Gender, main = "Perceived Service Quality", xlab = "PSQ", ylab = "Point (1-7)", col = c("Red", "blue"), notch = T, names = c("Male", "Female"))
boxplot (DataLOY$IMA ~ DataLOY$CITY, main = "Perceived Image", xlab = "IMA", ylab = "Point (1-7)", col = c("Red", "blue"), notch = T, names = c("Da Nang", "Ho Chi Minh"))
boxplot (DataLOY$IMA ~ DataLOY$Gender, main = "Perceived Image", xlab = "IMA", ylab = "Point (1-7)", col = c("Red", "blue"), notch = T, names = c("Male", "Female"))
boxplot (DataLOY$PHB ~ DataLOY$CITY, main = "Perceived Health Benefits", xlab = "PHB", ylab = "Point (1-7)", col = c("Red", "blue"), notch = T, names = c("Da Nang", "Ho Chi Minh"))## Warning in bxp(list(stats = structure(c(4.4, 5.4, 5.8, 6.2, 7, 3.6, 5, 6, : some
## notches went outside hinges ('box'): maybe set notch=FALSEboxplot (DataLOY$PHB ~ DataLOY$Gender, main = "Perceived Health Benefits", xlab = "PHB", ylab = "Point (1-7)", col = c("Red", "blue"), notch = T, names = c("Male", "Female"))
boxplot (DataLOY$PEV ~ DataLOY$CITY, main = "Perceived Environment Value/Benefits", xlab = "PEV", ylab = "Point (1-7)", col = c("Red", "blue"), notch = T, names = c("Da Nang", "Ho Chi Minh"))
boxplot (DataLOY$PEV ~ DataLOY$Gender, main = "Perceived Environment Value/Benefits", xlab = "PEV", ylab = "Point (1-7)", col = c("Red", "blue"), notch = T, names = c("Male", "Female"))
boxplot (DataLOY$ATM ~ DataLOY$CITY, main = "Perceived Atmospheric", xlab = "ATM", ylab = "Point (1-7)", col = c("Red", "blue"), notch = T, names = c("Da Nang", "Ho Chi Minh"))
boxplot (DataLOY$ATM ~ DataLOY$Gender, main = "Perceived Atmospheric", xlab = "ATM", ylab = "Point (1-7)", col = c("Red", "blue"), notch = T, names = c("Male", "Female"))
boxplot (DataLOY$PPI ~ DataLOY$CITY, main = "Passenger to Passenger Interaction", xlab = "PPI", ylab = "Point (1-7)", col = c("Red", "blue"), notch = T, names = c("Da Nang", "Ho Chi Minh"))
boxplot (DataLOY$PPI ~ DataLOY$Gender, main = "Passenger to Passenger Interaction", xlab = "PPI", ylab = "Point (1-7)", col = c("Red", "blue"), notch = T, names = c("Male", "Female"))
boxplot (DataLOY$SIM ~ DataLOY$CITY, main = "Similarity", xlab = "SIM", ylab = "Point (1-7)", col = c("Red", "blue"), notch = T, names = c("Da Nang", "Ho Chi Minh"))
boxplot (DataLOY$SIM ~ DataLOY$Gender, main = "Similarity", xlab = "SIM", ylab = "Point (1-7)", col = c("Red", "blue"), notch = T, names = c("Male", "Female"))
boxplot (DataLOY$PPA ~ DataLOY$CITY, main = "Perceived Physical Appearance", xlab = "PPA", ylab = "Point (1-7)", col = c("Red", "blue"), notch = T, names = c("Da Nang", "Ho Chi Minh"))
boxplot (DataLOY$PPA ~ DataLOY$Gender, main = "Perceived Physical Appearance", xlab = "PPA", ylab = "Point (1-7)", col = c("Red", "blue"), notch = T, names = c("Male", "Female"))
boxplot (DataLOY$SBE ~ DataLOY$CITY, main = "Perceived Suitable Behavior", xlab = "SBE", ylab = "Point (1-7)", col = c("Red", "blue"), notch = T, names = c("Da Nang", "Ho Chi Minh"))
boxplot (DataLOY$SBE ~ DataLOY$Gender, main = "Perceived Suitable Behavior", xlab = "SBE", ylab = "Point (1-7)", col = c("Red", "blue"), notch = T, names = c("Male", "Female"))
boxplot (DataLOY$EXB ~ DataLOY$CITY, main = "Experience on the Bus", xlab = "EXB", ylab = "Point (1-7)", col = c("Red", "blue"), notch = T, names = c("Da Nang", "Ho Chi Minh"))
boxplot (DataLOY$EXB ~ DataLOY$Gender, main = "Experience on the Bus", xlab = "EXB", ylab = "Point (1-7)", col = c("Red", "blue"), notch = T, names = c("Male", "Female"))
## LOY ~ EC_Stop
DataLOY %>%
group_by(LOY, EC_Stop) %>%
count() %>%
ggplot(aes(EC_Stop, n, fill = LOY)) +
geom_col(position = "fill") +
xlab("Experienced crime at bus stops") +
ylab("Perceived loyalty") +
ggtitle("Experienced crime at bus stops ~ Perceived loyalty")
## LOY ~ WC_Stop
DataLOY %>%
group_by(LOY, WC_Stop) %>%
count() %>%
ggplot(aes(WC_Stop, n, fill = LOY)) +
geom_col(position = "fill") +
xlab("Witnessed crime at bus stops") +
ylab("Perceived loyalty") +
ggtitle("Witnessed crime at bus stops ~ Perceived loyalty")
## LOY ~ EC_Bus
DataLOY %>%
group_by(LOY, EC_Bus) %>%
count() %>%
ggplot(aes(EC_Bus, n, fill = LOY)) +
geom_col(position = "fill") +
xlab("Experienced crime on bus") +
ylab("Perceived loyalty") +
ggtitle("Experienced crime on bus ~ Perceived loyalty")
## LOY ~ WC_Bus
DataLOY %>%
group_by(LOY, WC_Bus) %>%
count() %>%
ggplot(aes(WC_Bus, n, fill = LOY)) +
geom_col(position = "fill") +
xlab("Witnessed crime on bus") +
ylab("Perceived loyalty") +
ggtitle("Witnessed crime on bus ~ Perceived loyalty")
4. Descriptive statistical analysis
summary(DataLOY)## ID AGE CITY FRE
## Min. : 3.0 16-25:425 DaNang :422 >=3 days/week :508
## 1st Qu.:273.0 26-35:172 HoChiMinh:451 2days/month-2days/week:168
## Median :526.0 36-45:105 2days/year-1day/month : 99
## Mean :521.7 46-55: 77 <2 days/year : 98
## 3rd Qu.:769.0 >55 : 94
## Max. :993.0
##
## TripPurpose Departure TimeUseonBus TravelTime
## Working :305 Normal :296 Using.telephone:198 Min. : 0.000
## Studying :303 Peak-Hour:577 Reading : 53 1st Qu.: 0.500
## Shopping : 60 Listening :138 Median : 1.000
## Entertaining:100 Nothing :428 Mean : 1.291
## Others :105 Talking : 34 3rd Qu.: 2.000
## Others : 22 Max. :20.000
##
## PSSW PSSS PSAB PSEB
## Min. :1.000 Min. :1.000 Min. :1.000 Min. :1.000
## 1st Qu.:4.300 1st Qu.:3.700 1st Qu.:4.000 1st Qu.:5.500
## Median :4.900 Median :4.600 Median :5.500 Median :6.000
## Mean :4.768 Mean :4.478 Mean :5.047 Mean :5.786
## 3rd Qu.:5.400 3rd Qu.:5.400 3rd Qu.:6.000 3rd Qu.:6.300
## Max. :7.000 Max. :7.000 Max. :7.000 Max. :7.000
##
## PSQ SAT LOY IMA
## Min. :1.000 Not satisfied: 16 Notloyal: 12 Min. :1.000
## 1st Qu.:4.800 Normal :347 Normal :342 1st Qu.:4.800
## Median :5.400 Satisfied :510 Loyal :519 Median :5.600
## Mean :5.226 Mean :5.338
## 3rd Qu.:5.800 3rd Qu.:6.000
## Max. :7.000 Max. :7.000
##
## PHB PEV ATM PPI
## Min. :1.000 Min. :1.000 Min. :1.000 Min. :1.000
## 1st Qu.:5.200 1st Qu.:5.300 1st Qu.:4.300 1st Qu.:3.700
## Median :5.800 Median :6.000 Median :5.100 Median :4.300
## Mean :5.611 Mean :5.619 Mean :4.979 Mean :4.373
## 3rd Qu.:6.200 3rd Qu.:6.300 3rd Qu.:5.900 3rd Qu.:5.300
## Max. :7.000 Max. :7.000 Max. :7.000 Max. :7.000
##
## SIM PPA SBE EXB EC_Stop
## Min. :1.000 Min. :1.000 Min. :1.80 Min. :2.00 Ever : 36
## 1st Qu.:4.000 1st Qu.:4.000 1st Qu.:4.30 1st Qu.:5.00 Never:837
## Median :5.000 Median :5.000 Median :5.30 Median :5.80
## Mean :4.901 Mean :4.957 Mean :5.18 Mean :5.59
## 3rd Qu.:6.000 3rd Qu.:5.800 3rd Qu.:6.00 3rd Qu.:6.00
## Max. :7.000 Max. :7.000 Max. :7.00 Max. :7.00
##
## WC_Stop EC_Bus WC_Bus Gender MarriedStatus
## Ever :121 Ever : 42 Ever :116 Male :364 Single :535
## Never:752 Never:831 Never:757 Female:509 Married:338
##
##
##
##
##
## Occupation Education Income
## Students/Pupils:368 Secondary.school: 57 <5millions :465
## Full.time.job :305 Undergraduate :287 5-10millions :249
## Part.time.job : 69 High.school :368 10-15millions:122
## Retirement : 46 Postgraduate :106 >15millions : 37
## No.job : 3 Others : 55
## Housewife : 54
## Others : 28table(LOY) ## LOY
## 1 2 3
## 12 342 519# Descriptive Statistics of categorical variables
with(DataLOY, do.call(rbind, tapply(TravelTime, LOY, function(x) c(M = mean(x), SD = sd(x)))))## M SD
## Notloyal 1.666667 2.692151
## Normal 1.348763 1.647059
## Loyal 1.243487 1.289686with(DataLOY, do.call(rbind, tapply(PSSW, LOY, function(x) c(M = mean(x), SD = sd(x)))))## M SD
## Notloyal 3.475000 1.2899084
## Normal 4.404678 0.9961399
## Loyal 5.037380 0.90761995.1. Estimate Multinominal Logit Regression Model - DataLOY for 2 cities
library(mlogitBMA)## Warning: package 'mlogitBMA' was built under R version 4.0.4## Loading required package: BMA## Loading required package: survival## Loading required package: leaps## Loading required package: robustbase##
## Attaching package: 'robustbase'## The following object is masked from 'package:survival':
##
## heart## Loading required package: inline## Loading required package: rrcov## Scalable Robust Estimators with High Breakdown Point (version 1.5-5)## Loading required package: abind## Loading required package: maxLik## Warning: package 'maxLik' was built under R version 4.0.4## Loading required package: miscTools## Warning: package 'miscTools' was built under R version 4.0.4##
## Attaching package: 'miscTools'## The following objects are masked from 'package:robustbase':
##
## colMedians, rowMedians##
## Please cite the 'maxLik' package as:
## Henningsen, Arne and Toomet, Ott (2011). maxLik: A package for maximum likelihood estimation in R. Computational Statistics 26(3), 443-458. DOI 10.1007/s00180-010-0217-1.
##
## If you have questions, suggestions, or comments regarding the 'maxLik' package, please use a forum or 'tracker' at maxLik's R-Forge site:
## https://r-forge.r-project.org/projects/maxlik/attach(DataLOY)## The following objects are masked from DataLOY (pos = 30):
##
## AGE, ATM, CITY, Departure, EC_Bus, EC_Stop, Education, EXB, FRE,
## Gender, ID, IMA, Income, LOY, MarriedStatus, Occupation, PEV, PHB,
## PPA, PPI, PSAB, PSEB, PSQ, PSSS, PSSW, SAT, SBE, SIM, TimeUseonBus,
## TravelTime, TripPurpose, WC_Bus, WC_Stopstr(DataLOY)## 'data.frame': 873 obs. of 33 variables:
## $ ID : int 3 4 5 6 7 8 10 11 12 13 ...
## $ AGE : Factor w/ 5 levels "16-25","26-35",..: 1 1 1 1 1 1 3 1 1 2 ...
## $ CITY : Factor w/ 2 levels "DaNang","HoChiMinh": 2 2 2 2 2 2 2 2 2 2 ...
## $ FRE : Factor w/ 4 levels ">=3 days/week",..: 1 2 1 1 1 1 1 1 1 1 ...
## $ TripPurpose : Factor w/ 5 levels "Working","Studying",..: 2 4 2 2 2 2 1 2 2 2 ...
## $ Departure : Factor w/ 2 levels "Normal","Peak-Hour": 1 1 2 2 2 2 2 2 1 2 ...
## $ TimeUseonBus : Factor w/ 6 levels "Using.telephone",..: 4 4 4 1 4 6 4 4 3 3 ...
## $ TravelTime : num 3 2 0.17 4 2 2 2 2.5 1.5 2 ...
## $ PSSW : num 4.9 3.4 3.9 4.1 3.7 4.6 3.9 3.4 5.3 4 ...
## $ PSSS : num 5.4 2.4 3.4 4.4 2.6 2.7 4 2.7 3.1 4.7 ...
## $ PSAB : num 5.5 3.5 4.5 5.5 3.5 3.8 6 3.5 4.8 5.5 ...
## $ PSEB : num 5.8 6 4.5 6 4.5 4 4.5 6.3 7 4.5 ...
## $ PSQ : num 4.8 4.6 2.7 4.8 4.5 3.9 4.2 4.1 6 4.5 ...
## $ SAT : Factor w/ 3 levels "Not satisfied",..: 3 2 1 2 2 2 3 2 3 3 ...
## $ LOY : Factor w/ 3 levels "Notloyal","Normal",..: 3 2 2 2 2 2 3 2 3 3 ...
## $ IMA : num 5.6 4.8 3 5 4.6 3 5.8 5.8 6.8 6 ...
## $ PHB : num 6.6 4 5 5 5.6 5.4 5.8 5.6 7 6.4 ...
## $ PEV : num 4 5.3 5.8 5 6 5.3 6.3 6.5 7 5.3 ...
## $ ATM : num 4.3 4.1 3.1 5 3.3 2.7 4 4.1 4.4 3.9 ...
## $ PPI : num 6 4 2.7 5 3 3 6 2 3.3 5.3 ...
## $ SIM : num 3.5 4 3.5 5 5.5 2 6 5.3 4.3 6 ...
## $ PPA : num 4.5 4 4.5 5 5 2.5 4 4 4 4.3 ...
## $ SBE : num 4 4.8 6 5 5 2 5.3 4 5 3.8 ...
## $ EXB : num 4.8 4.2 4 5 4.2 5 5.3 5.3 5.3 5 ...
## $ EC_Stop : Factor w/ 2 levels "Ever","Never": 2 2 2 2 2 2 2 2 2 2 ...
## $ WC_Stop : Factor w/ 2 levels "Ever","Never": 1 2 2 1 1 1 2 1 2 2 ...
## $ EC_Bus : Factor w/ 2 levels "Ever","Never": 2 1 2 2 2 2 2 2 2 2 ...
## $ WC_Bus : Factor w/ 2 levels "Ever","Never": 1 2 2 1 1 1 2 1 2 2 ...
## $ Gender : Factor w/ 2 levels "Male","Female": 2 2 1 1 1 2 2 2 2 1 ...
## $ MarriedStatus: Factor w/ 2 levels "Single","Married": 1 1 1 1 1 1 2 1 1 1 ...
## $ Occupation : Factor w/ 7 levels "Students/Pupils",..: 1 1 1 1 1 1 2 1 1 7 ...
## $ Education : Factor w/ 5 levels "Secondary.school",..: 2 2 2 2 2 3 1 2 3 5 ...
## $ Income : Factor w/ 4 levels "<5millions","5-10millions",..: 1 1 1 1 1 1 1 1 1 1 ...# All variables - rempve: EC_Stop + WC_Stop + EC_Bus + WC_Bus + SAT and variable have cor > 0.7 - AGE, PPA
bma.LOY <- bic.mlogit(LOY ~ PSSW + PSSS + PSAB + PSEB + PSQ + IMA + PHB + PEV + ATM + PPI + SIM + SBE + EXB + Gender + MarriedStatus + Occupation + Education + Income + CITY + FRE + TripPurpose + Departure + TimeUseonBus + TravelTime, data = DataLOY)## Warning in new.output.names[names(output.names)[new.idx[i - 1]]] <-
## output.names[[i - : number of items to replace is not a multiple of replacement
## length## Warning in new.output.names[names(output.names)[new.idx[i - 1]]] <-
## output.names[[i - : number of items to replace is not a multiple of replacement
## length
## Warning in new.output.names[names(output.names)[new.idx[i - 1]]] <-
## output.names[[i - : number of items to replace is not a multiple of replacement
## length
## Warning in new.output.names[names(output.names)[new.idx[i - 1]]] <-
## output.names[[i - : number of items to replace is not a multiple of replacement
## length
## Warning in new.output.names[names(output.names)[new.idx[i - 1]]] <-
## output.names[[i - : number of items to replace is not a multiple of replacement
## length
## Warning in new.output.names[names(output.names)[new.idx[i - 1]]] <-
## output.names[[i - : number of items to replace is not a multiple of replacement
## length
## Warning in new.output.names[names(output.names)[new.idx[i - 1]]] <-
## output.names[[i - : number of items to replace is not a multiple of replacement
## length
## Warning in new.output.names[names(output.names)[new.idx[i - 1]]] <-
## output.names[[i - : number of items to replace is not a multiple of replacement
## length
## Warning in new.output.names[names(output.names)[new.idx[i - 1]]] <-
## output.names[[i - : number of items to replace is not a multiple of replacement
## length
## Warning in new.output.names[names(output.names)[new.idx[i - 1]]] <-
## output.names[[i - : number of items to replace is not a multiple of replacement
## lengthsummary(bma.LOY)##
## Call:
## bic.mlogit(f = LOY ~ PSSW + PSSS + PSAB + PSEB + PSQ + IMA + PHB + PEV + ATM + PPI + SIM + SBE + EXB + Gender + MarriedStatus + Occupation + Education + Income + CITY + FRE + TripPurpose + Departure + TimeUseonBus + TravelTime, data = DataLOY)
##
##
## 3 models were selected
## Best 3 models (cumulative posterior probability = 1 ):
##
## p!=0 EV SD model 1 model 2 model 3
## Intercept 100 19.7131 1.3528 19.5830 20.0530 19.4016
## Intercept.Notloyal 100.0 15.0010 0.5669 14.8929 15.3089 14.6009
## CITY 0.0
## . 0.0000 0.0000 . . .
## FRE 0.0
## . 0.0000 0.0000 . . .
## TripPurpose 0.0
## . 0.0000 0.0000 . . .
## Departure 0.0
## . 0.0000 0.0000 . . .
## TimeUseonBus 0.0
## . 0.0000 0.0000 . . .
## TravelTime 0.0 0.0000 0.0000 . . .
## PSSW 0.0 0.0000 0.0000 . . .
## PSSS 0.0 0.0000 0.0000 . . .
## PSAB 0.0 0.0000 0.0000 . . .
## PSEB 0.0 0.0000 0.0000 . . .
## PSQ 100.0 0.0000 0.0000 . . .
## IMA 100.0 0.0000 0.0000 . . .
## PHB 35.1 0.0000 0.0000 . . .
## PEV 94.6 0.0000 0.0000 . . .
## ATM 0.0 0.0000 0.0000 . . .
## PPI 0.0 0.0000 0.0000 . . .
## SIM 100.0 0.0000 0.0000 . . .
## SBE 100.0 0.0000 0.0000 . . .
## EXB 100.0 0.0000 0.0000 . . .
## Gender 0.0
## . -1.3102 0.1733 -1.3157 -1.3009 -1.2951
## MarriedStatus 0.0
## . -0.8294 0.1504 -0.8442 -0.8039 -0.7912
## Occupation 0.0
## . -0.1430 0.2167 . -0.3744 -0.5882
## Education 0.0
## . -0.3601 0.1367 -0.4210 -0.2922 .
## Income 0.0
## . 0.0000 0.0000 . . .
##
## nVar 4 5 4
## BIC 819.9786 821.5370 824.9627
## post prob 0.649 0.298 0.054
##
## MNL specification:
## ==================
## Response variable: LOY
## Base choice name: Loyal
## Base choice index: 1
## Frequency of alternatives:
## Loyal Normal Notloyal
## 519 342 12
##
## Equations:
## ----------
## alternative intercept 1 2 3 4
## 1 Loyal:
## 2 Normal: asc.Normal + CITY + FRE + TripPurpose + Departure
## 3 Notloyal: asc.Notloyal + CITY + FRE + TripPurpose + Departure
## 5 6 7 8 9 10 11 12 13
## 1
## 2 + TimeUseonBus + TravelTime + PSSW + PSSS + PSAB + PSEB + PSQ + IMA + PHB
## 3 + TimeUseonBus + TravelTime + PSSW + PSSS + PSAB + PSEB + PSQ + IMA + PHB
## 14 15 16 17 18 19 20 21 22
## 1
## 2 + PEV + ATM + PPI + SIM + SBE + EXB + Gender + MarriedStatus + Occupation
## 3 + PEV + ATM + PPI + SIM + SBE + EXB + Gender + MarriedStatus + Occupation
## 23 24
## 1
## 2 + Education + Income
## 3 + Education + Income# All variables - rempve: EC_Stop + WC_Stop + EC_Bus + WC_Bus + SAT and variable have cor > 0.7 - AGE, PPA and cor > 0.6 : PHB, PSQ, PSSW, PSAB
bma.LOY3 <- bic.mlogit(LOY ~ PSSS + PSEB + IMA + PEV + ATM + PPI + SIM + SBE + EXB + Gender + MarriedStatus + Occupation + Education + Income + CITY + FRE + TripPurpose + Departure + TimeUseonBus + TravelTime, data = DataLOY)## Warning in new.output.names[names(output.names)[new.idx[i - 1]]] <-
## output.names[[i - : number of items to replace is not a multiple of replacement
## length
## Warning in new.output.names[names(output.names)[new.idx[i - 1]]] <-
## output.names[[i - : number of items to replace is not a multiple of replacement
## length
## Warning in new.output.names[names(output.names)[new.idx[i - 1]]] <-
## output.names[[i - : number of items to replace is not a multiple of replacement
## length
## Warning in new.output.names[names(output.names)[new.idx[i - 1]]] <-
## output.names[[i - : number of items to replace is not a multiple of replacement
## length
## Warning in new.output.names[names(output.names)[new.idx[i - 1]]] <-
## output.names[[i - : number of items to replace is not a multiple of replacement
## length
## Warning in new.output.names[names(output.names)[new.idx[i - 1]]] <-
## output.names[[i - : number of items to replace is not a multiple of replacement
## length
## Warning in new.output.names[names(output.names)[new.idx[i - 1]]] <-
## output.names[[i - : number of items to replace is not a multiple of replacement
## length
## Warning in new.output.names[names(output.names)[new.idx[i - 1]]] <-
## output.names[[i - : number of items to replace is not a multiple of replacement
## length
## Warning in new.output.names[names(output.names)[new.idx[i - 1]]] <-
## output.names[[i - : number of items to replace is not a multiple of replacement
## length
## Warning in new.output.names[names(output.names)[new.idx[i - 1]]] <-
## output.names[[i - : number of items to replace is not a multiple of replacement
## lengthsummary(bma.LOY3)##
## Call:
## bic.mlogit(f = LOY ~ PSSS + PSEB + IMA + PEV + ATM + PPI + SIM + SBE + EXB + Gender + MarriedStatus + Occupation + Education + Income + CITY + FRE + TripPurpose + Departure + TimeUseonBus + TravelTime, data = DataLOY)
##
##
## 1 models were selected
## Best 1 models (cumulative posterior probability = 1 ):
##
## p!=0 EV SD model 1
## Intercept 100 17.4991 1.20366 17.4991
## Intercept.Notloyal 100 13.0409 0.48623 13.0409
## CITY 0
## . 0.0000 0.00000 .
## FRE 0
## . 0.0000 0.00000 .
## TripPurpose 0
## . 0.0000 0.00000 .
## Departure 0
## . 0.0000 0.00000 .
## TimeUseonBus 0
## . 0.0000 0.00000 .
## TravelTime 0 0.0000 0.00000 .
## PSSS 100 0.0000 0.00000 .
## PSEB 0 0.0000 0.00000 .
## IMA 100 0.0000 0.00000 .
## PEV 100 0.0000 0.00000 .
## ATM 0 0.0000 0.00000 .
## PPI 0 0.0000 0.00000 .
## SIM 100 0.0000 0.00000 .
## SBE 100 0.0000 0.00000 .
## EXB 100 0.0000 0.00000 .
## Gender 0
## . -0.3857 0.09075 -0.3857
## MarriedStatus 0
## . 0.0000 0.00000 .
## Occupation 0
## . -1.2094 0.13820 -1.2094
## Education 0
## . -0.3820 0.08698 -0.3820
## Income 0
## . 0.0000 0.00000 .
##
## nVar 4
## BIC 849.7514
## post prob 1
##
## MNL specification:
## ==================
## Response variable: LOY
## Base choice name: Loyal
## Base choice index: 1
## Frequency of alternatives:
## Loyal Normal Notloyal
## 519 342 12
##
## Equations:
## ----------
## alternative intercept 1 2 3 4
## 1 Loyal:
## 2 Normal: asc.Normal + CITY + FRE + TripPurpose + Departure
## 3 Notloyal: asc.Notloyal + CITY + FRE + TripPurpose + Departure
## 5 6 7 8 9 10 11 12 13 14
## 1
## 2 + TimeUseonBus + TravelTime + PSSS + PSEB + IMA + PEV + ATM + PPI + SIM + SBE
## 3 + TimeUseonBus + TravelTime + PSSS + PSEB + IMA + PEV + ATM + PPI + SIM + SBE
## 15 16 17 18 19 20
## 1
## 2 + EXB + Gender + MarriedStatus + Occupation + Education + Income
## 3 + EXB + Gender + MarriedStatus + Occupation + Education + Income# Remove 3 variables non significant : Gender, Departure, Travel time compare to bma.LOY
bma.LOY1 <- bic.mlogit(LOY ~ PSSW + PSSS + PSAB + PSEB + PSQ + IMA + PHB + PEV + ATM + PPI + SIM + SBE + EXB + MarriedStatus + Occupation + Education + Income + CITY + FRE + TripPurpose + TimeUseonBus, data = DataLOY)## Warning in new.output.names[names(output.names)[new.idx[i - 1]]] <-
## output.names[[i - : number of items to replace is not a multiple of replacement
## length
## Warning in new.output.names[names(output.names)[new.idx[i - 1]]] <-
## output.names[[i - : number of items to replace is not a multiple of replacement
## length
## Warning in new.output.names[names(output.names)[new.idx[i - 1]]] <-
## output.names[[i - : number of items to replace is not a multiple of replacement
## length
## Warning in new.output.names[names(output.names)[new.idx[i - 1]]] <-
## output.names[[i - : number of items to replace is not a multiple of replacement
## length
## Warning in new.output.names[names(output.names)[new.idx[i - 1]]] <-
## output.names[[i - : number of items to replace is not a multiple of replacement
## length
## Warning in new.output.names[names(output.names)[new.idx[i - 1]]] <-
## output.names[[i - : number of items to replace is not a multiple of replacement
## length
## Warning in new.output.names[names(output.names)[new.idx[i - 1]]] <-
## output.names[[i - : number of items to replace is not a multiple of replacement
## length
## Warning in new.output.names[names(output.names)[new.idx[i - 1]]] <-
## output.names[[i - : number of items to replace is not a multiple of replacement
## lengthsummary(bma.LOY1)##
## Call:
## bic.mlogit(f = LOY ~ PSSW + PSSS + PSAB + PSEB + PSQ + IMA + PHB + PEV + ATM + PPI + SIM + SBE + EXB + MarriedStatus + Occupation + Education + Income + CITY + FRE + TripPurpose + TimeUseonBus, data = DataLOY)
##
##
## 3 models were selected
## Best 3 models (cumulative posterior probability = 1 ):
##
## p!=0 EV SD model 1 model 2 model 3
## Intercept 100 19.7131 1.3528 19.5830 20.0530 19.4016
## Intercept.Notloyal 100.0 15.0010 0.5669 14.8929 15.3089 14.6009
## CITY 0.0
## . 0.0000 0.0000 . . .
## FRE 0.0
## . 0.0000 0.0000 . . .
## TripPurpose 0.0
## . 0.0000 0.0000 . . .
## TimeUseonBus 0.0
## . 0.0000 0.0000 . . .
## PSSW 0.0 0.0000 0.0000 . . .
## PSSS 0.0 0.0000 0.0000 . . .
## PSAB 0.0 0.0000 0.0000 . . .
## PSEB 0.0 0.0000 0.0000 . . .
## PSQ 100.0 0.0000 0.0000 . . .
## IMA 100.0 0.0000 0.0000 . . .
## PHB 35.1 0.0000 0.0000 . . .
## PEV 94.6 0.0000 0.0000 . . .
## ATM 0.0 0.0000 0.0000 . . .
## PPI 0.0 0.0000 0.0000 . . .
## SIM 100.0 0.0000 0.0000 . . .
## SBE 100.0 0.0000 0.0000 . . .
## EXB 100.0 0.0000 0.0000 . . .
## MarriedStatus 0.0
## . -1.3102 0.1733 -1.3157 -1.3009 -1.2951
## Occupation 0.0
## . -0.8294 0.1504 -0.8442 -0.8039 -0.7912
## Education 0.0
## . -0.1430 0.2167 . -0.3744 -0.5882
## Income 0.0
## . -0.3601 0.1367 -0.4210 -0.2922 .
##
## nVar 4 5 4
## BIC 819.9786 821.5370 824.9627
## post prob 0.649 0.298 0.054
##
## MNL specification:
## ==================
## Response variable: LOY
## Base choice name: Loyal
## Base choice index: 1
## Frequency of alternatives:
## Loyal Normal Notloyal
## 519 342 12
##
## Equations:
## ----------
## alternative intercept 1 2 3 4 5
## 1 Loyal:
## 2 Normal: asc.Normal + CITY + FRE + TripPurpose + TimeUseonBus + PSSW
## 3 Notloyal: asc.Notloyal + CITY + FRE + TripPurpose + TimeUseonBus + PSSW
## 6 7 8 9 10 11 12 13 14 15 16 17
## 1
## 2 + PSSS + PSAB + PSEB + PSQ + IMA + PHB + PEV + ATM + PPI + SIM + SBE + EXB
## 3 + PSSS + PSAB + PSEB + PSQ + IMA + PHB + PEV + ATM + PPI + SIM + SBE + EXB
## 18 19 20 21
## 1
## 2 + MarriedStatus + Occupation + Education + Income
## 3 + MarriedStatus + Occupation + Education + Income# Only variables related to personal perception , remove SAT, PPI
bma.LOY2 <- bic.mlogit(LOY ~ PSSW + PSSS + PSAB + PSEB + PSQ + IMA + PHB + PEV + ATM + SIM + PPA + SBE + EXB + CITY, data = DataLOY)## Warning in new.output.names[names(output.names)[new.idx[i - 1]]] <-
## output.names[[i - : number of items to replace is not a multiple of replacement
## lengthsummary(bma.LOY2)##
## Call:
## bic.mlogit(f = LOY ~ PSSW + PSSS + PSAB + PSEB + PSQ + IMA + PHB + PEV + ATM + SIM + PPA + SBE + EXB + CITY, data = DataLOY)
##
##
## 10 models were selected
## Best 5 models (cumulative posterior probability = 0.8542 ):
##
## p!=0 EV SD model 1 model 2 model 3
## Intercept 100 19.806445 1.37103 19.5830 20.0530 19.9257
## Intercept.Notloyal 100.0 15.085372 0.59154 14.8929 15.3089 15.1966
## CITY 0.0
## . 0.000000 0.00000 . . .
## PSSW 2.1 -0.002589 0.02413 . . .
## PSSS 3.8 -0.005154 0.03229 . . .
## PSAB 0.0 0.000000 0.00000 . . .
## PSEB 2.4 0.003422 0.02956 . . .
## PSQ 100.0 -1.299253 0.17615 -1.3157 -1.3009 -1.2918
## IMA 100.0 -0.834828 0.15059 -0.8442 -0.8039 -0.8530
## PHB 32.4 -0.129220 0.20704 . -0.3744 .
## PEV 96.6 -0.375862 0.12819 -0.4210 -0.2922 -0.4439
## ATM 0.0 0.000000 0.00000 . . .
## SIM 93.3 -0.356105 0.14446 -0.4087 -0.4019 -0.3067
## PPA 28.9 -0.119721 0.20887 . . -0.3713
## SBE 100.0 0.610799 0.15560 0.5693 0.5401 0.7451
## EXB 100.0 -1.255380 0.18204 -1.3008 -1.1675 -1.2763
##
## nVar 7 8 8
## BIC 819.9786 821.5370 822.0100
## post prob 0.407 0.187 0.147
## model 4 model 5
## Intercept 20.3971 19.6036
## Intercept.Notloyal 15.6208 14.8775
## CITY
## . . .
## PSSW . .
## PSSS . -0.1343
## PSAB . .
## PSEB . .
## PSQ -1.2735 -1.2285
## IMA -0.8131 -0.8588
## PHB -0.3781 .
## PEV -0.3171 -0.4009
## ATM . .
## SIM -0.2971 -0.4031
## PPA -0.3795 .
## SBE 0.7220 0.5827
## EXB -1.1409 -1.3032
##
## nVar 9 8
## BIC 823.3629 824.7009
## post prob 0.075 0.038
##
## MNL specification:
## ==================
## Response variable: LOY
## Base choice name: Loyal
## Base choice index: 1
## Frequency of alternatives:
## Loyal Normal Notloyal
## 519 342 12
##
## Equations:
## ----------
## alternative intercept 1 2 3 4 5 6 7 8
## 1 Loyal:
## 2 Normal: asc.Normal + CITY + PSSW + PSSS + PSAB + PSEB + PSQ + IMA + PHB
## 3 Notloyal: asc.Notloyal + CITY + PSSW + PSSS + PSAB + PSEB + PSQ + IMA + PHB
## 9 10 11 12 13 14
## 1
## 2 + PEV + ATM + SIM + PPA + SBE + EXB
## 3 + PEV + ATM + SIM + PPA + SBE + EXB# Multinominal Logit Model
## Way 1 - Use mlogit (in mlogit package) - Khong dung
## Way 2 - Use multinom (in nnet package)
library(nnet)
attach(DataLOY)## The following objects are masked from DataLOY (pos = 4):
##
## AGE, ATM, CITY, Departure, EC_Bus, EC_Stop, Education, EXB, FRE,
## Gender, ID, IMA, Income, LOY, MarriedStatus, Occupation, PEV, PHB,
## PPA, PPI, PSAB, PSEB, PSQ, PSSS, PSSW, SAT, SBE, SIM, TimeUseonBus,
## TravelTime, TripPurpose, WC_Bus, WC_Stop## The following objects are masked from DataLOY (pos = 32):
##
## AGE, ATM, CITY, Departure, EC_Bus, EC_Stop, Education, EXB, FRE,
## Gender, ID, IMA, Income, LOY, MarriedStatus, Occupation, PEV, PHB,
## PPA, PPI, PSAB, PSEB, PSQ, PSSS, PSSW, SAT, SBE, SIM, TimeUseonBus,
## TravelTime, TripPurpose, WC_Bus, WC_Stop# Model - All variables - rempve: EC_Stop + WC_Stop + EC_Bus + WC_Bus + SAT and variable have cor > 0.7 - AGE, PPA
DataLOY$LOY <- relevel (DataLOY$LOY, ref = "Normal")
mlm <- multinom(LOY ~ PSSW + PSSS + PSAB + PSEB + PSQ + IMA + PHB + PEV + ATM + PPI + SIM + SBE + EXB + Gender + MarriedStatus + Occupation + Education + Income + CITY + FRE + TripPurpose + Departure + TimeUseonBus + TravelTime, data = DataLOY)## # weights: 135 (88 variable)
## initial value 959.088528
## iter 10 value 453.950608
## iter 20 value 352.250834
## iter 30 value 332.503203
## iter 40 value 316.998887
## iter 50 value 304.252876
## iter 60 value 295.668752
## iter 70 value 290.059617
## iter 80 value 290.007156
## final value 290.007100
## convergedsummary(mlm)## Warning in sqrt(diag(vc)): NaNs produced## Call:
## multinom(formula = LOY ~ PSSW + PSSS + PSAB + PSEB + PSQ + IMA +
## PHB + PEV + ATM + PPI + SIM + SBE + EXB + Gender + MarriedStatus +
## Occupation + Education + Income + CITY + FRE + TripPurpose +
## Departure + TimeUseonBus + TravelTime, data = DataLOY)
##
## Coefficients:
## (Intercept) PSSW PSSS PSAB PSEB
## Notloyal 321.65863 -3.998656e+01 -13.2782936 32.11904148 11.5846313
## Loyal -19.61627 2.841991e-05 0.1305778 -0.09042967 -0.1085531
## PSQ IMA PHB PEV ATM PPI
## Notloyal -83.678637 -11.975128 -40.0186353 43.1496270 8.5890689 -4.78048461
## Loyal 1.124745 1.085532 0.3697361 0.4033366 0.1020622 -0.01930062
## SIM SBE EXB GenderFemale MarriedStatusMarried
## Notloyal 10.8172060 48.1616619 -85.7155032 -14.2277213 185.7306743
## Loyal 0.5259931 -0.6753868 0.9851403 0.2167796 0.5413487
## OccupationFull.time.job OccupationPart.time.job OccupationRetirement
## Notloyal -246.4583127 -254.311302 -131.17934
## Loyal 0.4449424 1.367044 1.35764
## OccupationNo.job OccupationHousewife OccupationOthers
## Notloyal 103.0764 -115.847663 -319.479805
## Loyal 265.5093 0.806828 0.446782
## EducationUndergraduate EducationHigh.school EducationPostgraduate
## Notloyal -32.8082866 128.9417664 30.353026
## Loyal -0.2352182 -0.7134174 -0.719562
## EducationOthers Income5-10millions Income10-15millions
## Notloyal 115.180634 31.3949930 157.8417421
## Loyal -0.875494 -0.3906108 0.5066374
## Income>15millions CITYHoChiMinh FRE2days/month-2days/week
## Notloyal -187.4191561 32.2056836 3.6049167
## Loyal 0.1473969 0.0112766 -0.1518838
## FRE2days/year-1day/month FRE<2 days/year TripPurposeStudying
## Notloyal -43.718210 -4.070659 6.7120707
## Loyal -1.313819 -1.240355 0.4223585
## TripPurposeShopping TripPurposeEntertaining TripPurposeOthers
## Notloyal 137.56006575 53.3667563 53.9688266
## Loyal 0.01141173 -0.2367265 0.0153024
## DeparturePeak-Hour TimeUseonBusReading TimeUseonBusListening
## Notloyal -90.0779462 -63.6792007 -45.2935586
## Loyal -0.2274571 0.9514671 0.1684674
## TimeUseonBusNothing TimeUseonBusTalking TimeUseonBusOthers TravelTime
## Notloyal -34.47060377 49.654156 -42.99992612 0.8095944
## Loyal -0.08913576 -0.220101 0.05227977 -0.2182342
##
## Std. Errors:
## (Intercept) PSSW PSSS PSAB PSEB PSQ
## Notloyal 2.715091 7.8419680 7.1134876 11.7136338 14.1565669 12.3642321
## Loyal 1.763570 0.1571821 0.1359787 0.1251969 0.1447274 0.2257357
## IMA PHB PEV ATM PPI SIM
## Notloyal 12.3830689 13.66067 14.7215674 12.2046501 11.31223342 12.4896476
## Loyal 0.1819016 0.16518 0.1199103 0.1356465 0.09432754 0.1225799
## SBE EXB GenderFemale MarriedStatusMarried
## Notloyal 13.8147540 13.4473089 2.1788845 1.4150799
## Loyal 0.1586131 0.1945776 0.2375803 0.3419758
## OccupationFull.time.job OccupationPart.time.job OccupationRetirement
## Notloyal 1.4625207 NaN NaN
## Loyal 0.5360589 0.5870125 0.6809328
## OccupationNo.job OccupationHousewife OccupationOthers
## Notloyal 1.740837e-12 1.757861e-11 0.002632499
## Loyal 1.615448e-32 6.723880e-01 0.876885108
## EducationUndergraduate EducationHigh.school EducationPostgraduate
## Notloyal 0.3734500 2.339009 0.0001052767
## Loyal 0.5178708 0.525864 0.6036034852
## EducationOthers Income5-10millions Income10-15millions
## Notloyal 0.002527222 0.8578368 0.6073165
## Loyal 0.665109294 0.3757588 0.5016978
## Income>15millions CITYHoChiMinh FRE2days/month-2days/week
## Notloyal 0.1591323 2.0806846 0.0006658904
## Loyal 0.7389645 0.2747704 0.3037308445
## FRE2days/year-1day/month FRE<2 days/year TripPurposeStudying
## Notloyal 0.8580611 1.2794331 0.5948624
## Loyal 0.4262825 0.4456464 0.4168705
## TripPurposeShopping TripPurposeEntertaining TripPurposeOthers
## Notloyal 0.5546039 1.3547296 0.1591309
## Loyal 0.5443258 0.4544956 0.4768115
## DeparturePeak-Hour TimeUseonBusReading TimeUseonBusListening
## Notloyal 0.4870490 0.002527223 4.777538e-06
## Loyal 0.2408429 0.580925599 3.427102e-01
## TimeUseonBusNothing TimeUseonBusTalking TimeUseonBusOthers TravelTime
## Notloyal 1.0874769 NaN 1.874612e-13 2.0575057
## Loyal 0.2768653 0.620583 8.166636e-01 0.0985532
##
## Residual Deviance: 580.0142
## AIC: 756.0142exp(coef(mlm))## (Intercept) PSSW PSSS PSAB PSEB
## Notloyal 4.949585e+139 4.305825e-18 1.711238e-06 8.894519e+13 1.074339e+05
## Loyal 3.025259e-09 1.000028e+00 1.139487e+00 9.135386e-01 8.971313e-01
## PSQ IMA PHB PEV ATM
## Notloyal 4.558582e-37 6.298945e-06 4.169918e-18 5.490917e+18 5372.608999
## Loyal 3.079433e+00 2.961015e+00 1.447353e+00 1.496811e+00 1.107452
## PPI SIM SBE EXB GenderFemale
## Notloyal 0.008391931 49871.551072 8.247912e+20 5.946069e-38 6.621847e-07
## Loyal 0.980884447 1.692138 5.089595e-01 2.678188e+00 1.242070e+00
## MarriedStatusMarried OccupationFull.time.job OccupationPart.time.job
## Notloyal 4.589940e+80 9.215412e-108 3.581004e-111
## Loyal 1.718323e+00 1.560400e+00 3.923737e+00
## OccupationRetirement OccupationNo.job OccupationHousewife
## Notloyal 1.070374e-57 5.827626e+44 4.875274e-51
## Loyal 3.887008e+00 2.038165e+115 2.240789e+00
## OccupationOthers EducationUndergraduate EducationHigh.school
## Notloyal 1.785186e-139 5.643418e-15 9.970057e+55
## Loyal 1.563274e+00 7.903983e-01 4.899669e-01
## EducationPostgraduate EducationOthers Income5-10millions
## Notloyal 1.521079e+13 1.052722e+50 4.311935e+13
## Loyal 4.869655e-01 4.166561e-01 6.766434e-01
## Income10-15millions Income>15millions CITYHoChiMinh
## Notloyal 3.546481e+68 4.026194e-82 9.69953e+13
## Loyal 1.659701e+00 1.158814e+00 1.01134e+00
## FRE2days/month-2days/week FRE2days/year-1day/month FRE<2 days/year
## Notloyal 36.7786203 1.031389e-19 0.01706614
## Loyal 0.8590881 2.687915e-01 0.28928140
## TripPurposeStudying TripPurposeShopping TripPurposeEntertaining
## Notloyal 822.271557 5.515406e+59 1.502754e+23
## Loyal 1.525555 1.011477e+00 7.892071e-01
## TripPurposeOthers DeparturePeak-Hour TimeUseonBusReading
## Notloyal 2.74387e+23 7.579578e-40 2.210419e-28
## Loyal 1.01542e+00 7.965566e-01 2.589506e+00
## TimeUseonBusListening TimeUseonBusNothing TimeUseonBusTalking
## Notloyal 2.13431e-20 1.070550e-15 3.668816e+21
## Loyal 1.18349e+00 9.147214e-01 8.024378e-01
## TimeUseonBusOthers TravelTime
## Notloyal 2.115287e-19 2.2469964
## Loyal 1.053670e+00 0.8039371exp(confint(mlm))## Warning in sqrt(diag(vcov(object))): NaNs produced## , , Notloyal
##
## 2.5 % 97.5 %
## (Intercept) 2.418126e+137 1.013115e+142
## PSSW 9.098318e-25 2.037754e-11
## PSSS 1.507644e-12 1.942325e+00
## PSAB 9.516108e+03 8.313533e+23
## PSEB 9.573039e-08 1.205683e+17
## PSQ 1.362625e-47 1.525047e-26
## IMA 1.814598e-16 2.186529e+05
## PHB 9.820717e-30 1.770565e-06
## PEV 1.616680e+06 1.864944e+31
## ATM 2.195676e-07 1.314626e+14
## PPI 1.971839e-12 3.571515e+07
## SIM 1.165858e-06 2.133341e+15
## SBE 1.436164e+09 4.736787e+32
## EXB 2.127452e-49 1.661881e-26
## GenderFemale 9.253570e-09 4.738588e-05
## MarriedStatusMarried 2.866083e+79 7.350640e+81
## OccupationFull.time.job 5.243423e-109 1.619626e-106
## OccupationPart.time.job NaN NaN
## OccupationRetirement NaN NaN
## OccupationNo.job 5.827626e+44 5.827626e+44
## OccupationHousewife 4.875274e-51 4.875274e-51
## OccupationOthers 1.775999e-139 1.794420e-139
## EducationUndergraduate 2.714319e-15 1.173339e-14
## EducationHigh.school 1.017958e+54 9.764847e+57
## EducationPostgraduate 1.520766e+13 1.521393e+13
## EducationOthers 1.047521e+50 1.057950e+50
## Income5-10millions 8.025650e+12 2.316670e+14
## Income10-15millions 1.078571e+68 1.166128e+69
## Income>15millions 2.947413e-82 5.499818e-82
## CITYHoChiMinh 1.643118e+12 5.725753e+15
## FRE2days/month-2days/week 3.673065e+01 3.682665e+01
## FRE2days/year-1day/month 1.918844e-20 5.543774e-19
## FRE<2 days/year 1.390208e-03 2.095032e-01
## TripPurposeStudying 2.562521e+02 2.638536e+03
## TripPurposeShopping 1.859935e+59 1.635525e+60
## TripPurposeEntertaining 1.056185e+22 2.138138e+24
## TripPurposeOthers 2.008682e+23 3.748143e+23
## DeparturePeak-Hour 2.917885e-40 1.968892e-39
## TimeUseonBusReading 2.199497e-28 2.221395e-28
## TimeUseonBusListening 2.134290e-20 2.134330e-20
## TimeUseonBusNothing 1.270413e-16 9.021297e-15
## TimeUseonBusTalking NaN NaN
## TimeUseonBusOthers 2.115287e-19 2.115287e-19
## TravelTime 3.983367e-02 1.267519e+02
##
## , , Loyal
##
## 2.5 % 97.5 %
## (Intercept) 9.541323e-11 9.592159e-08
## PSSW 7.348838e-01 1.360837e+00
## PSSS 8.728989e-01 1.487491e+00
## PSAB 7.147581e-01 1.167602e+00
## PSEB 6.755598e-01 1.191374e+00
## PSQ 1.978446e+00 4.793109e+00
## IMA 2.073030e+00 4.229368e+00
## PHB 1.047063e+00 2.000672e+00
## PEV 1.183311e+00 1.893366e+00
## ATM 8.489118e-01 1.444733e+00
## PPI 8.153163e-01 1.180075e+00
## SIM 1.330748e+00 2.151672e+00
## SBE 3.729680e-01 6.945363e-01
## EXB 1.829011e+00 3.921623e+00
## GenderFemale 7.796819e-01 1.978677e+00
## MarriedStatusMarried 8.790517e-01 3.358884e+00
## OccupationFull.time.job 5.456848e-01 4.462006e+00
## OccupationPart.time.job 1.241749e+00 1.239840e+01
## OccupationRetirement 1.023303e+00 1.476476e+01
## OccupationNo.job 2.038165e+115 2.038165e+115
## OccupationHousewife 5.998784e-01 8.370255e+00
## OccupationOthers 2.803038e-01 8.718483e+00
## EducationUndergraduate 2.864399e-01 2.181014e+00
## EducationHigh.school 1.748037e-01 1.373356e+00
## EducationPostgraduate 1.491798e-01 1.589595e+00
## EducationOthers 1.131451e-01 1.534334e+00
## Income5-10millions 3.239764e-01 1.413209e+00
## Income10-15millions 6.208460e-01 4.436861e+00
## Income>15millions 2.722738e-01 4.931981e+00
## CITYHoChiMinh 5.902179e-01 1.732935e+00
## FRE2days/month-2days/week 4.736980e-01 1.558023e+00
## FRE2days/year-1day/month 1.165636e-01 6.198232e-01
## FRE<2 days/year 1.207774e-01 6.928758e-01
## TripPurposeStudying 6.738872e-01 3.453574e+00
## TripPurposeShopping 3.480368e-01 2.939592e+00
## TripPurposeEntertaining 3.238349e-01 1.923350e+00
## TripPurposeOthers 3.988257e-01 2.585285e+00
## DeparturePeak-Hour 4.968334e-01 1.277093e+00
## TimeUseonBusReading 8.293391e-01 8.085403e+00
## TimeUseonBusListening 6.045733e-01 2.316754e+00
## TimeUseonBusNothing 5.316437e-01 1.573827e+00
## TimeUseonBusTalking 2.377772e-01 2.708024e+00
## TimeUseonBusOthers 2.125982e-01 5.222159e+00
## TravelTime 6.627251e-01 9.752382e-01z <- summary(mlm)$coefficients/summary(mlm)$standard.errors## Warning in sqrt(diag(vc)): NaNs produced## Warning in sqrt(diag(vc)): NaNs producedz## (Intercept) PSSW PSSS PSAB PSEB PSQ
## Notloyal 118.47065 -5.0990469034 -1.8666362 2.7420220 0.8183221 -6.767799
## Loyal -11.12304 0.0001808088 0.9602811 -0.7222996 -0.7500523 4.982578
## IMA PHB PEV ATM PPI SIM SBE
## Notloyal -0.9670566 -2.929478 2.931048 0.7037538 -0.4225942 0.8660938 3.486248
## Loyal 5.9676874 2.238384 3.363653 0.7524136 -0.2046128 4.2910219 -4.258077
## EXB GenderFemale MarriedStatusMarried OccupationFull.time.job
## Notloyal -6.374175 -6.5298190 131.251018 -168.5161139
## Loyal 5.062968 0.9124477 1.583003 0.8300253
## OccupationPart.time.job OccupationRetirement OccupationNo.job
## Notloyal NaN NaN 5.921080e+13
## Loyal 2.328817 1.993794 1.643565e+34
## OccupationHousewife OccupationOthers EducationUndergraduate
## Notloyal -6.590262e+12 -1.213599e+05 -87.8518845
## Loyal 1.199944e+00 5.095104e-01 -0.4542025
## EducationHigh.school EducationPostgraduate EducationOthers
## Notloyal 55.126673 288316.56960 45575.979271
## Loyal -1.356657 -1.19211 -1.316316
## Income5-10millions Income10-15millions Income>15millions CITYHoChiMinh
## Notloyal 36.597862 259.900321 -1177.7569939 15.47840727
## Loyal -1.039525 1.009846 0.1994641 0.04104007
## FRE2days/month-2days/week FRE2days/year-1day/month FRE<2 days/year
## Notloyal 5413.6785334 -50.949997 -3.181612
## Loyal -0.5000605 -3.082039 -2.783273
## TripPurposeStudying TripPurposeShopping TripPurposeEntertaining
## Notloyal 11.283401 248.03298331 39.3929219
## Loyal 1.013165 0.02096489 -0.5208555
## TripPurposeOthers DeparturePeak-Hour TimeUseonBusReading
## Notloyal 339.14727038 -184.9463820 -25197.305047
## Loyal 0.03209318 -0.9444209 1.637847
## TimeUseonBusListening TimeUseonBusNothing TimeUseonBusTalking
## Notloyal -9.480523e+06 -31.6977807 NaN
## Loyal 4.915739e-01 -0.3219463 -0.3546681
## TimeUseonBusOthers TravelTime
## Notloyal -2.293804e+14 0.3934834
## Loyal 6.401628e-02 -2.2143800pvalue <- (1-pnorm(abs(z), 0, 1))*2
pvalue## (Intercept) PSSW PSSS PSAB PSEB PSQ
## Notloyal 0 3.413680e-07 0.06195241 0.006106225 0.4131733 1.307554e-11
## Loyal 0 9.998557e-01 0.33691378 0.470110290 0.4532232 6.274279e-07
## IMA PHB PEV ATM PPI SIM
## Notloyal 3.335157e-01 0.003395318 0.0033782010 0.4815861 0.6725913 3.864387e-01
## Loyal 2.406398e-09 0.025196052 0.0007691808 0.4518024 0.8378747 1.778528e-05
## SBE EXB GenderFemale MarriedStatusMarried
## Notloyal 4.898462e-04 1.839502e-10 6.584933e-11 0.0000000
## Loyal 2.061927e-05 4.127785e-07 3.615331e-01 0.1134208
## OccupationFull.time.job OccupationPart.time.job OccupationRetirement
## Notloyal 0.0000000 NaN NaN
## Loyal 0.4065245 0.01986878 0.04617459
## OccupationNo.job OccupationHousewife OccupationOthers
## Notloyal 0 0.0000000 0.0000000
## Loyal 0 0.2301611 0.6103945
## EducationUndergraduate EducationHigh.school EducationPostgraduate
## Notloyal 0.0000000 0.0000000 0.0000000
## Loyal 0.6496831 0.1748901 0.2332179
## EducationOthers Income5-10millions Income10-15millions
## Notloyal 0.000000 0.0000000 0.0000000
## Loyal 0.188068 0.2985604 0.3125691
## Income>15millions CITYHoChiMinh FRE2days/month-2days/week
## Notloyal 0.0000000 0.000000 0.0000000
## Loyal 0.8418997 0.967264 0.6170325
## FRE2days/year-1day/month FRE<2 days/year TripPurposeStudying
## Notloyal 0.000000000 0.001464581 0.0000000
## Loyal 0.002055878 0.005381356 0.3109814
## TripPurposeShopping TripPurposeEntertaining TripPurposeOthers
## Notloyal 0.0000000 0.0000000 0.0000000
## Loyal 0.9832737 0.6024674 0.9743977
## DeparturePeak-Hour TimeUseonBusReading TimeUseonBusListening
## Notloyal 0.0000000 0.0000000 0.0000000
## Loyal 0.3449546 0.1014537 0.6230206
## TimeUseonBusNothing TimeUseonBusTalking TimeUseonBusOthers TravelTime
## Notloyal 0.0000000 NaN 0.0000000 0.69396248
## Loyal 0.7474934 0.7228383 0.9489573 0.02680265# All variables - rempve: EC_Stop + WC_Stop + EC_Bus + WC_Bus + SAT and variable have cor > 0.7 - AGE, PPA and cor > 0.6 : PHB, PSQ, PSSW, PSAB
mlm3 <- multinom(LOY ~ PSSS + PSEB + IMA + PEV + ATM + PPI + SIM + SBE + EXB + Gender + MarriedStatus + Occupation + Education + Income + CITY + FRE + TripPurpose + Departure + TimeUseonBus + TravelTime, data = DataLOY)## # weights: 123 (80 variable)
## initial value 959.088528
## iter 10 value 490.718472
## iter 20 value 377.902263
## iter 30 value 352.518574
## iter 40 value 338.392823
## iter 50 value 326.169984
## iter 60 value 321.139194
## iter 70 value 312.047173
## iter 80 value 305.693354
## iter 90 value 305.680024
## final value 305.680009
## convergedsummary(mlm3)## Warning in sqrt(diag(vc)): NaNs produced## Call:
## multinom(formula = LOY ~ PSSS + PSEB + IMA + PEV + ATM + PPI +
## SIM + SBE + EXB + Gender + MarriedStatus + Occupation + Education +
## Income + CITY + FRE + TripPurpose + Departure + TimeUseonBus +
## TravelTime, data = DataLOY)
##
## Coefficients:
## (Intercept) PSSS PSEB IMA PEV
## Notloyal 983.98562 -174.0270435 -342.068512 -30.670326 33.4599816
## Loyal -18.17361 0.2684894 0.131721 1.286097 0.5003621
## ATM PPI SIM SBE EXB
## Notloyal -392.3821864 123.21998922 261.7459918 621.0000351 -955.097695
## Loyal 0.2699704 -0.00964319 0.4885282 -0.5561363 1.133247
## GenderFemale MarriedStatusMarried OccupationFull.time.job
## Notloyal 311.5579788 2296.638912 -3548.9250264
## Loyal 0.2860572 0.561053 0.6001944
## OccupationPart.time.job OccupationRetirement OccupationNo.job
## Notloyal -3405.131231 -4560.650666 1205.459
## Loyal 1.467401 1.424432 3808.448
## OccupationHousewife OccupationOthers EducationUndergraduate
## Notloyal -2961.2716099 -4759.9957260 890.1436161
## Loyal 0.9033854 0.5660432 -0.2724886
## EducationHigh.school EducationPostgraduate EducationOthers
## Notloyal 2188.3431334 -187.621234 504.5678333
## Loyal -0.6833763 -0.756985 -0.8010366
## Income5-10millions Income10-15millions Income>15millions CITYHoChiMinh
## Notloyal 1638.7933664 2723.5089257 -960.65630342 -216.3973773
## Loyal -0.4029312 0.4068685 0.03074978 0.1936299
## FRE2days/month-2days/week FRE2days/year-1day/month FRE<2 days/year
## Notloyal 411.70301226 -71.539793 754.709339
## Loyal -0.02303295 -1.138717 -1.060629
## TripPurposeStudying TripPurposeShopping TripPurposeEntertaining
## Notloyal -1066.258889 -738.181270 -317.0859753
## Loyal 0.439192 0.121196 -0.1262873
## TripPurposeOthers DeparturePeak-Hour TimeUseonBusReading
## Notloyal -148.09011855 -547.9195346 322.3412650
## Loyal 0.08358628 -0.2713393 0.9524749
## TimeUseonBusListening TimeUseonBusNothing TimeUseonBusTalking
## Notloyal -253.8454239 -297.3637081 -292.6306714
## Loyal 0.2709124 -0.1346034 -0.1231011
## TimeUseonBusOthers TravelTime
## Notloyal 1.333061e+02 122.2186518
## Loyal 7.072144e-03 -0.1851648
##
## Std. Errors:
## (Intercept) PSSS PSEB IMA PEV ATM
## Notloyal 3.580116 11.0561585 19.7689560 17.015055 21.4940103 13.9388656
## Loyal 1.643273 0.1106245 0.1279661 0.169772 0.1050239 0.1278413
## PPI SIM SBE EXB GenderFemale
## Notloyal 11.62924805 14.2362260 17.3939321 14.6095279 1.0114609
## Loyal 0.09162656 0.1170518 0.1467181 0.1865282 0.2285929
## MarriedStatusMarried OccupationFull.time.job OccupationPart.time.job
## Notloyal 0.05675354 0.0208823 NaN
## Loyal 0.32753779 0.5181017 0.569897
## OccupationRetirement OccupationNo.job OccupationHousewife
## Notloyal NaN NaN 2.769343e-11
## Loyal 0.6668581 0 6.677160e-01
## OccupationOthers EducationUndergraduate EducationHigh.school
## Notloyal 2.570646e-11 0.01676111 3.5633546
## Loyal 8.502461e-01 0.50223499 0.5107926
## EducationPostgraduate EducationOthers Income5-10millions
## Notloyal 1.558077e-11 6.095908e-12 0.8100033
## Loyal 5.836837e-01 6.464915e-01 0.3662780
## Income10-15millions Income>15millions CITYHoChiMinh
## Notloyal 0.001223154 0.06454067 1.0754318
## Loyal 0.477556853 0.67294860 0.2569506
## FRE2days/month-2days/week FRE2days/year-1day/month FRE<2 days/year
## Notloyal 2.5053014 5.827012e-12 1.0359255
## Loyal 0.2924999 4.060779e-01 0.4224086
## TripPurposeStudying TripPurposeShopping TripPurposeEntertaining
## Notloyal 1.8874514 NaN 0.01680792
## Loyal 0.4067056 0.525103 0.45331538
## TripPurposeOthers DeparturePeak-Hour TimeUseonBusReading
## Notloyal 1.671735 1.7906433 0.001187232
## Loyal 0.465053 0.2350279 0.555775266
## TimeUseonBusListening TimeUseonBusNothing TimeUseonBusTalking
## Notloyal 1.032094e-10 1.8903881 NaN
## Loyal 3.312029e-01 0.2651796 0.602188
## TimeUseonBusOthers TravelTime
## Notloyal 1.6454858 7.88215994
## Loyal 0.8028598 0.09742529
##
## Residual Deviance: 611.36
## AIC: 771.36# Model 1 - Remove 3 variables non significant : Gender, Departure, Travel time compare to mlm
mlm1 <- multinom(LOY ~ PSSW + PSSS + PSAB + PSEB + PSQ + IMA + PHB + PEV + ATM + PPI + SIM + SBE + EXB + MarriedStatus + Occupation + Education + Income + CITY + FRE + TripPurpose + TimeUseonBus, data = DataLOY)## # weights: 126 (82 variable)
## initial value 959.088528
## iter 10 value 467.143426
## iter 20 value 353.848576
## iter 30 value 330.540175
## iter 40 value 319.814747
## iter 50 value 305.931630
## iter 60 value 295.221188
## iter 70 value 292.833987
## iter 80 value 292.822387
## final value 292.822369
## convergedsummary(mlm1)## Warning in sqrt(diag(vc)): NaNs produced## Call:
## multinom(formula = LOY ~ PSSW + PSSS + PSAB + PSEB + PSQ + IMA +
## PHB + PEV + ATM + PPI + SIM + SBE + EXB + MarriedStatus +
## Occupation + Education + Income + CITY + FRE + TripPurpose +
## TimeUseonBus, data = DataLOY)
##
## Coefficients:
## (Intercept) PSSW PSSS PSAB PSEB PSQ
## Notloyal 323.20808 -85.1411169 -23.9370240 57.3849388 45.8611458 -123.541324
## Loyal -19.40988 -0.0210442 0.1286796 -0.1020307 -0.1090966 1.139398
## IMA PHB PEV ATM PPI SIM
## Notloyal -20.625356 -27.352013 25.5627070 -35.45570334 -0.31313855 -3.4318447
## Loyal 1.073455 0.327174 0.3875495 0.08007645 -0.04539722 0.5287091
## SBE EXB MarriedStatusMarried OccupationFull.time.job
## Notloyal 81.7652946 -106.431404 254.5549231 -421.3617298
## Loyal -0.6628959 1.005245 0.5941146 0.4890313
## OccupationPart.time.job OccupationRetirement OccupationNo.job
## Notloyal -523.017501 -288.366158 166.4984
## Loyal 1.427174 1.315342 449.6795
## OccupationHousewife OccupationOthers EducationUndergraduate
## Notloyal -284.2746810 -476.7155758 -49.2267281
## Loyal 0.8579015 0.4794359 -0.2958962
## EducationHigh.school EducationPostgraduate EducationOthers
## Notloyal 305.9735391 12.4481264 295.2808450
## Loyal -0.6951871 -0.8623468 -0.8138199
## Income5-10millions Income10-15millions Income>15millions CITYHoChiMinh
## Notloyal 124.0549573 284.6574730 -191.9674979 74.58257986
## Loyal -0.3686488 0.4438381 -0.1629409 -0.06422769
## FRE2days/month-2days/week FRE2days/year-1day/month FRE<2 days/year
## Notloyal 34.403038 -26.108525 -22.801133
## Loyal -0.101386 -1.189073 -1.121525
## TripPurposeStudying TripPurposeShopping TripPurposeEntertaining
## Notloyal 18.9090198 325.9364410 113.8412896
## Loyal 0.4558836 0.1304685 -0.1489088
## TripPurposeOthers TimeUseonBusReading TimeUseonBusListening
## Notloyal 126.85770158 -68.1132027 -26.1397836
## Loyal 0.05640188 0.8994296 0.1686483
## TimeUseonBusNothing TimeUseonBusTalking TimeUseonBusOthers
## Notloyal 0.3955566 61.1801334 -87.1159454
## Loyal -0.1047600 -0.1459001 -0.2573989
##
## Std. Errors:
## (Intercept) PSSW PSSS PSAB PSEB PSQ
## Notloyal 1.483124 3.4161391 2.0823546 8.5970241 8.1552481 8.1531224
## Loyal 1.713274 0.1555265 0.1351471 0.1244643 0.1440654 0.2236143
## IMA PHB PEV ATM PPI SIM SBE
## Notloyal 7.1155808 9.1886425 8.8952036 8.8920846 9.92818396 9.3389618 9.6337013
## Loyal 0.1813382 0.1634773 0.1188479 0.1336689 0.09343782 0.1195174 0.1574458
## EXB MarriedStatusMarried OccupationFull.time.job
## Notloyal 8.5965310 1.4805925 1.4804611
## Loyal 0.1932893 0.3332808 0.5344822
## OccupationPart.time.job OccupationRetirement OccupationNo.job
## Notloyal 5.440314e-12 NaN NaN
## Loyal 5.846004e-01 0.6755306 0
## OccupationHousewife OccupationOthers EducationUndergraduate
## Notloyal NaN 1.050980e-12 4.661934e-05
## Loyal 0.6639656 8.975743e-01 5.166424e-01
## EducationHigh.school EducationPostgraduate EducationOthers
## Notloyal 1.4830772 2.615932e-12 NaN
## Loyal 0.5238524 5.998233e-01 0.6560544
## Income5-10millions Income10-15millions Income>15millions CITYHoChiMinh
## Notloyal 1.480451 9.968968e-06 0.0001407498 1.4830495
## Loyal 0.378699 4.982579e-01 0.7332674044 0.2689951
## FRE2days/month-2days/week FRE2days/year-1day/month FRE<2 days/year
## Notloyal 0.0000611189 1.4804521 0.002413144
## Loyal 0.2972394431 0.4170424 0.433790016
## TripPurposeStudying TripPurposeShopping TripPurposeEntertaining
## Notloyal 0.001602951 6.797734e-07 1.4814499
## Loyal 0.415162970 5.329705e-01 0.4459175
## TripPurposeOthers TimeUseonBusReading TimeUseonBusListening
## Notloyal 2.559843e-12 8.120761e-07 3.216733e-12
## Loyal 4.759997e-01 5.726479e-01 3.402315e-01
## TimeUseonBusNothing TimeUseonBusTalking TimeUseonBusOthers
## Notloyal 0.002661274 3.398440e-14 1.466051e-13
## Loyal 0.274437661 6.134246e-01 8.040931e-01
##
## Residual Deviance: 585.6447
## AIC: 749.6447mlm.chon <- multinom(LOY ~ PSQ + IMA + PHB + PEV + SIM + SBE + EXB + MarriedStatus + FRE, data = DataLOY)## # weights: 39 (24 variable)
## initial value 959.088528
## iter 10 value 447.521930
## iter 20 value 365.754637
## iter 30 value 337.634601
## iter 40 value 335.910970
## final value 335.910128
## convergedsummary(mlm.chon)## Call:
## multinom(formula = LOY ~ PSQ + IMA + PHB + PEV + SIM + SBE +
## EXB + MarriedStatus + FRE, data = DataLOY)
##
## Coefficients:
## (Intercept) PSQ IMA PHB PEV SIM
## Notloyal 4.267974 -1.498441 0.1001002 -0.5211413 0.7548364 0.09905232
## Loyal -19.555787 1.157054 0.9410593 0.3238565 0.3294089 0.48888198
## SBE EXB MarriedStatusMarried FRE2days/month-2days/week
## Notloyal 0.5021210 -1.605856 0.4721092 1.0918113
## Loyal -0.5720311 1.034042 0.8651999 -0.1551092
## FRE2days/year-1day/month FRE<2 days/year
## Notloyal 0.2989249 1.4920046
## Loyal -1.0257128 -0.8659766
##
## Std. Errors:
## (Intercept) PSQ IMA PHB PEV SIM
## Notloyal 2.215573 0.4219775 0.4896395 0.4782836 0.4003960 0.3675357
## Loyal 1.439342 0.1836982 0.1615532 0.1504298 0.1054378 0.1076867
## SBE EXB MarriedStatusMarried FRE2days/month-2days/week
## Notloyal 0.4193698 0.5636659 0.9377554 1.3035991
## Loyal 0.1341555 0.1820965 0.2243892 0.2597373
## FRE2days/year-1day/month FRE<2 days/year
## Notloyal 1.2188064 0.9797358
## Loyal 0.3464703 0.3541791
##
## Residual Deviance: 671.8203
## AIC: 719.8203# Model 2 - bma2 - 9 variables (including LOY), remove cor > 0.7: PPA
mlm2 <- multinom(LOY ~ PSSW + PSSS + PSAB + PSEB + PSQ + IMA + PHB + PEV + ATM + SIM + SBE + EXB + CITY, data = DataLOY)## # weights: 45 (28 variable)
## initial value 959.088528
## iter 10 value 452.411172
## iter 20 value 374.998611
## iter 30 value 355.405641
## iter 40 value 346.088394
## final value 346.071312
## convergedsummary(mlm2)## Call:
## multinom(formula = LOY ~ PSSW + PSSS + PSAB + PSEB + PSQ + IMA +
## PHB + PEV + ATM + SIM + SBE + EXB + CITY, data = DataLOY)
##
## Coefficients:
## (Intercept) PSSW PSSS PSAB PSEB PSQ
## Notloyal 4.732102 -0.12041084 -0.02960159 0.12566353 0.6064778 -2.009214
## Loyal -18.932702 -0.01977747 0.17224302 -0.01347676 -0.1673042 1.096446
## IMA PHB PEV ATM SIM SBE
## Notloyal -0.008939119 -0.6986138 0.6965186 0.19464308 0.1287331 0.4470136
## Loyal 0.936785582 0.3686742 0.2863793 0.04451399 0.4317126 -0.5923468
## EXB CITYHoChiMinh
## Notloyal -1.510156 -0.22936715
## Loyal 1.101410 -0.09227602
##
## Std. Errors:
## (Intercept) PSSW PSSS PSAB PSEB PSQ
## Notloyal 2.635756 0.4994537 0.5411987 0.5048315 0.4673654 0.6300213
## Loyal 1.486636 0.1451729 0.1232509 0.1147958 0.1361819 0.2028641
## IMA PHB PEV ATM SIM SBE EXB
## Notloyal 0.5500393 0.5278843 0.3870319 0.4674947 0.3741817 0.4378694 0.5692588
## Loyal 0.1638865 0.1537574 0.1047847 0.1209207 0.1054288 0.1397219 0.1805568
## CITYHoChiMinh
## Notloyal 1.0995331
## Loyal 0.2364972
##
## Residual Deviance: 692.1426
## AIC: 748.1426# CHON: Model mlm1 - all varibles , removing : EC_Stop + WC_Stop + EC_Bus + WC_Bus + SAT and variable have cor > 0.7 - AGE, PPA and nonsignificant variables : Gender, Departure, Travel time
summary(mlm1)## Warning in sqrt(diag(vc)): NaNs produced## Call:
## multinom(formula = LOY ~ PSSW + PSSS + PSAB + PSEB + PSQ + IMA +
## PHB + PEV + ATM + PPI + SIM + SBE + EXB + MarriedStatus +
## Occupation + Education + Income + CITY + FRE + TripPurpose +
## TimeUseonBus, data = DataLOY)
##
## Coefficients:
## (Intercept) PSSW PSSS PSAB PSEB PSQ
## Notloyal 323.20808 -85.1411169 -23.9370240 57.3849388 45.8611458 -123.541324
## Loyal -19.40988 -0.0210442 0.1286796 -0.1020307 -0.1090966 1.139398
## IMA PHB PEV ATM PPI SIM
## Notloyal -20.625356 -27.352013 25.5627070 -35.45570334 -0.31313855 -3.4318447
## Loyal 1.073455 0.327174 0.3875495 0.08007645 -0.04539722 0.5287091
## SBE EXB MarriedStatusMarried OccupationFull.time.job
## Notloyal 81.7652946 -106.431404 254.5549231 -421.3617298
## Loyal -0.6628959 1.005245 0.5941146 0.4890313
## OccupationPart.time.job OccupationRetirement OccupationNo.job
## Notloyal -523.017501 -288.366158 166.4984
## Loyal 1.427174 1.315342 449.6795
## OccupationHousewife OccupationOthers EducationUndergraduate
## Notloyal -284.2746810 -476.7155758 -49.2267281
## Loyal 0.8579015 0.4794359 -0.2958962
## EducationHigh.school EducationPostgraduate EducationOthers
## Notloyal 305.9735391 12.4481264 295.2808450
## Loyal -0.6951871 -0.8623468 -0.8138199
## Income5-10millions Income10-15millions Income>15millions CITYHoChiMinh
## Notloyal 124.0549573 284.6574730 -191.9674979 74.58257986
## Loyal -0.3686488 0.4438381 -0.1629409 -0.06422769
## FRE2days/month-2days/week FRE2days/year-1day/month FRE<2 days/year
## Notloyal 34.403038 -26.108525 -22.801133
## Loyal -0.101386 -1.189073 -1.121525
## TripPurposeStudying TripPurposeShopping TripPurposeEntertaining
## Notloyal 18.9090198 325.9364410 113.8412896
## Loyal 0.4558836 0.1304685 -0.1489088
## TripPurposeOthers TimeUseonBusReading TimeUseonBusListening
## Notloyal 126.85770158 -68.1132027 -26.1397836
## Loyal 0.05640188 0.8994296 0.1686483
## TimeUseonBusNothing TimeUseonBusTalking TimeUseonBusOthers
## Notloyal 0.3955566 61.1801334 -87.1159454
## Loyal -0.1047600 -0.1459001 -0.2573989
##
## Std. Errors:
## (Intercept) PSSW PSSS PSAB PSEB PSQ
## Notloyal 1.483124 3.4161391 2.0823546 8.5970241 8.1552481 8.1531224
## Loyal 1.713274 0.1555265 0.1351471 0.1244643 0.1440654 0.2236143
## IMA PHB PEV ATM PPI SIM SBE
## Notloyal 7.1155808 9.1886425 8.8952036 8.8920846 9.92818396 9.3389618 9.6337013
## Loyal 0.1813382 0.1634773 0.1188479 0.1336689 0.09343782 0.1195174 0.1574458
## EXB MarriedStatusMarried OccupationFull.time.job
## Notloyal 8.5965310 1.4805925 1.4804611
## Loyal 0.1932893 0.3332808 0.5344822
## OccupationPart.time.job OccupationRetirement OccupationNo.job
## Notloyal 5.440314e-12 NaN NaN
## Loyal 5.846004e-01 0.6755306 0
## OccupationHousewife OccupationOthers EducationUndergraduate
## Notloyal NaN 1.050980e-12 4.661934e-05
## Loyal 0.6639656 8.975743e-01 5.166424e-01
## EducationHigh.school EducationPostgraduate EducationOthers
## Notloyal 1.4830772 2.615932e-12 NaN
## Loyal 0.5238524 5.998233e-01 0.6560544
## Income5-10millions Income10-15millions Income>15millions CITYHoChiMinh
## Notloyal 1.480451 9.968968e-06 0.0001407498 1.4830495
## Loyal 0.378699 4.982579e-01 0.7332674044 0.2689951
## FRE2days/month-2days/week FRE2days/year-1day/month FRE<2 days/year
## Notloyal 0.0000611189 1.4804521 0.002413144
## Loyal 0.2972394431 0.4170424 0.433790016
## TripPurposeStudying TripPurposeShopping TripPurposeEntertaining
## Notloyal 0.001602951 6.797734e-07 1.4814499
## Loyal 0.415162970 5.329705e-01 0.4459175
## TripPurposeOthers TimeUseonBusReading TimeUseonBusListening
## Notloyal 2.559843e-12 8.120761e-07 3.216733e-12
## Loyal 4.759997e-01 5.726479e-01 3.402315e-01
## TimeUseonBusNothing TimeUseonBusTalking TimeUseonBusOthers
## Notloyal 0.002661274 3.398440e-14 1.466051e-13
## Loyal 0.274437661 6.134246e-01 8.040931e-01
##
## Residual Deviance: 585.6447
## AIC: 749.6447 ## Calculate OR and CI
exp(coef(mlm1))## (Intercept) PSSW PSSS PSAB PSEB
## Notloyal 2.330691e+140 1.056046e-37 4.020523e-11 8.355304e+24 8.264994e+19
## Loyal 3.718751e-09 9.791757e-01 1.137326e+00 9.030018e-01 8.966438e-01
## PSQ IMA PHB PEV ATM
## Notloyal 2.221696e-54 1.102863e-09 1.321818e-12 1.263987e+11 3.997456e-16
## Loyal 3.124887e+00 2.925471e+00 1.387043e+00 1.473366e+00 1.083370e+00
## PPI SIM SBE EXB MarriedStatusMarried
## Notloyal 0.7311486 0.03232725 3.237548e+35 5.990027e-47 3.562857e+110
## Loyal 0.9556178 1.69674059 5.153568e-01 2.732576e+00 1.811426e+00
## OccupationFull.time.job OccupationPart.time.job OccupationRetirement
## Notloyal 1.011407e-183 7.184317e-228 5.809904e-126
## Loyal 1.630736e+00 4.166908e+00 3.726025e+00
## OccupationNo.job OccupationHousewife OccupationOthers
## Notloyal 2.038590e+72 3.475960e-124 9.226903e-208
## Loyal 1.964886e+195 2.358207e+00 1.615163e+00
## EducationUndergraduate EducationHigh.school EducationPostgraduate
## Notloyal 4.179301e-22 7.631671e+132 2.547725e+05
## Loyal 7.438646e-01 4.989811e-01 4.221702e-01
## EducationOthers Income5-10millions Income10-15millions
## Notloyal 1.733172e+128 7.522867e+53 4.218613e+123
## Loyal 4.431620e-01 6.916683e-01 1.558678e+00
## Income>15millions CITYHoChiMinh FRE2days/month-2days/week
## Notloyal 4.261622e-84 2.459251e+32 8.730706e+14
## Loyal 8.496414e-01 9.377915e-01 9.035842e-01
## FRE2days/year-1day/month FRE<2 days/year TripPurposeStudying
## Notloyal 4.583654e-12 1.251969e-10 1.629607e+08
## Loyal 3.045034e-01 3.257826e-01 1.577567e+00
## TripPurposeShopping TripPurposeEntertaining TripPurposeOthers
## Notloyal 3.567778e+141 2.758315e+49 1.240509e+55
## Loyal 1.139362e+00 8.616477e-01 1.058023e+00
## TimeUseonBusReading TimeUseonBusListening TimeUseonBusNothing
## Notloyal 2.623082e-30 4.442590e-12 1.4852106
## Loyal 2.458201e+00 1.183704e+00 0.9005406
## TimeUseonBusTalking TimeUseonBusOthers
## Notloyal 3.717015e+26 1.465634e-38
## Loyal 8.642440e-01 7.730598e-01exp(confint(mlm1))## Warning in sqrt(diag(vcov(object))): NaNs produced## , , Notloyal
##
## 2.5 % 97.5 %
## (Intercept) 1.273642e+139 4.265030e+141
## PSSW 1.305750e-40 8.540940e-35
## PSSS 6.788583e-13 2.381146e-09
## PSAB 4.019408e+17 1.736850e+32
## PSEB 9.451108e+12 7.227737e+26
## PSQ 2.551140e-61 1.934795e-47
## IMA 9.676724e-16 1.256941e-03
## PHB 1.994274e-20 8.761091e-05
## PEV 3.389468e+03 4.713610e+18
## ATM 1.078518e-23 1.481630e-08
## PPI 2.588929e-09 2.064863e+08
## SIM 3.632705e-10 2.876786e+06
## SBE 2.041713e+27 5.133785e+43
## EXB 2.884352e-54 1.243968e-39
## MarriedStatusMarried 1.956662e+109 6.487555e+111
## OccupationFull.time.job 5.555909e-185 1.841181e-182
## OccupationPart.time.job 7.184317e-228 7.184317e-228
## OccupationRetirement NaN NaN
## OccupationNo.job NaN NaN
## OccupationHousewife NaN NaN
## OccupationOthers 9.226903e-208 9.226903e-208
## EducationUndergraduate 4.178919e-22 4.179683e-22
## EducationHigh.school 4.170826e+131 1.396424e+134
## EducationPostgraduate 2.547725e+05 2.547725e+05
## EducationOthers NaN NaN
## Income5-10millions 4.132579e+52 1.369448e+55
## Income10-15millions 4.218531e+123 4.218696e+123
## Income>15millions 4.260447e-84 4.262798e-84
## CITYHoChiMinh 1.344092e+31 4.499631e+33
## FRE2days/month-2days/week 8.729660e+14 8.731752e+14
## FRE2days/year-1day/month 2.517960e-13 8.344012e-11
## FRE<2 days/year 1.246062e-10 1.257904e-10
## TripPurposeStudying 1.624496e+08 1.634735e+08
## TripPurposeShopping 3.567773e+141 3.567782e+141
## TripPurposeEntertaining 1.512277e+48 5.031024e+50
## TripPurposeOthers 1.240509e+55 1.240509e+55
## TimeUseonBusReading 2.623078e-30 2.623087e-30
## TimeUseonBusListening 4.442590e-12 4.442590e-12
## TimeUseonBusNothing 1.477484e+00 1.492978e+00
## TimeUseonBusTalking 3.717015e+26 3.717015e+26
## TimeUseonBusOthers 1.465634e-38 1.465634e-38
##
## , , Loyal
##
## 2.5 % 97.5 %
## (Intercept) 1.294362e-10 1.068411e-07
## PSSW 7.218987e-01 1.328143e+00
## PSSS 8.726647e-01 1.482253e+00
## PSAB 7.075292e-01 1.152479e+00
## PSEB 6.760693e-01 1.189183e+00
## PSQ 2.016013e+00 4.843676e+00
## IMA 2.050409e+00 4.173987e+00
## PHB 1.006787e+00 1.910918e+00
## PEV 1.167205e+00 1.859834e+00
## ATM 8.336765e-01 1.407849e+00
## PPI 7.957008e-01 1.147674e+00
## SIM 1.342400e+00 2.144612e+00
## SBE 3.785209e-01 7.016590e-01
## EXB 1.870872e+00 3.991172e+00
## MarriedStatusMarried 9.426090e-01 3.481046e+00
## OccupationFull.time.job 5.720468e-01 4.648744e+00
## OccupationPart.time.job 1.324955e+00 1.310469e+01
## OccupationRetirement 9.913637e-01 1.400420e+01
## OccupationNo.job 1.964886e+195 1.964886e+195
## OccupationHousewife 6.418201e-01 8.664639e+00
## OccupationOthers 2.780992e-01 9.380652e+00
## EducationUndergraduate 2.702260e-01 2.047673e+00
## EducationHigh.school 1.787229e-01 1.393118e+00
## EducationPostgraduate 1.302918e-01 1.367912e+00
## EducationOthers 1.224977e-01 1.603234e+00
## Income5-10millions 3.292674e-01 1.452938e+00
## Income10-15millions 5.870006e-01 4.138799e+00
## Income>15millions 2.018726e-01 3.575971e+00
## CITYHoChiMinh 5.535250e-01 1.588822e+00
## FRE2days/month-2days/week 5.046124e-01 1.618003e+00
## FRE2days/year-1day/month 1.344637e-01 6.895715e-01
## FRE<2 days/year 1.392147e-01 7.623785e-01
## TripPurposeStudying 6.991983e-01 3.559386e+00
## TripPurposeShopping 4.008634e-01 3.238374e+00
## TripPurposeEntertaining 3.595540e-01 2.064883e+00
## TripPurposeOthers 4.162204e-01 2.689470e+00
## TimeUseonBusReading 8.001631e-01 7.551898e+00
## TimeUseonBusListening 6.076276e-01 2.305943e+00
## TimeUseonBusNothing 5.258980e-01 1.542073e+00
## TimeUseonBusTalking 2.597098e-01 2.875970e+00
## TimeUseonBusOthers 1.598703e-01 3.738165e+00est1 <- estimate.mlogit(LOY ~ PSSW + PSSS + PSAB + PSEB + PSQ + IMA + PHB + PEV + ATM + PPI + SIM + SBE + EXB + MarriedStatus + Occupation + Education + Income + CITY + FRE + TripPurpose + TimeUseonBus, data = DataLOY)
summary(est1)##
## successive function values within relative tolerance limit (reltol)
##
## Log-Likelihood: -355.3
## Null Log-Likelihood: -959.1
## Likelihood ratio index: 0.6296
## AIC: 756.5488
## BIC: 866.3033
## Sample size: 873
## Iterations: 14
## Suggested |t-value| > 2.602294
## Convergence statistics: 2.748882e-06
##
## Estimated using BHHH in 0.7609339 s.
##
## Coefficients :
## Estimate Std. Error t-value Pr(>|t|)
## asc.Normal 18.636551 1.58334 11.77038 0.000e+00
## asc.Notloyal 15.286711 1.60564 9.52065 0.000e+00
## CITY 0.138567 0.25831 0.53644 5.917e-01
## FRE 0.407963 0.13732 2.97086 2.970e-03
## TripPurpose -0.001332 0.09614 -0.01385 9.889e-01
## TimeUseonBus 0.049141 0.07977 0.61606 5.379e-01
## PSSW 0.034578 0.15418 0.22427 8.225e-01
## PSSS -0.202934 0.13259 -1.53052 1.259e-01
## PSAB 0.115619 0.11538 1.00205 3.163e-01
## PSEB 0.115551 0.13627 0.84795 3.965e-01
## PSQ -1.129345 0.22845 -4.94350 7.673e-07
## IMA -0.991808 0.15201 -6.52449 6.823e-11
## PHB -0.378955 0.16131 -2.34919 1.881e-02
## PEV -0.309982 0.11101 -2.79249 5.230e-03
## ATM -0.082888 0.12819 -0.64660 5.179e-01
## PPI 0.041199 0.08628 0.47750 6.330e-01
## SIM -0.477622 0.10857 -4.39901 1.087e-05
## SBE 0.618264 0.14746 4.19275 2.756e-05
## EXB -1.022911 0.18731 -5.46097 4.735e-08
## MarriedStatus -0.579582 0.28668 -2.02167 4.321e-02
## Occupation -0.161793 0.09066 -1.78462 7.432e-02
## Education 0.214111 0.11939 1.79333 7.292e-02
## Income -0.033166 0.15368 -0.21581 8.291e-01
##
## Response variable: LOY
## Base choice name: Loyal
## Base choice index: 1
## Frequency of alternatives:
## Loyal Normal Notloyal
## 519 342 12
##
## Equations:
## ----------
## alternative intercept 1 2 3 4 5
## 1 Loyal:
## 2 Normal: asc.Normal + CITY + FRE + TripPurpose + TimeUseonBus + PSSW
## 3 Notloyal: asc.Notloyal + CITY + FRE + TripPurpose + TimeUseonBus + PSSW
## 6 7 8 9 10 11 12 13 14 15 16 17
## 1
## 2 + PSSS + PSAB + PSEB + PSQ + IMA + PHB + PEV + ATM + PPI + SIM + SBE + EXB
## 3 + PSSS + PSAB + PSEB + PSQ + IMA + PHB + PEV + ATM + PPI + SIM + SBE + EXB
## 18 19 20 21
## 1
## 2 + MarriedStatus + Occupation + Education + Income
## 3 + MarriedStatus + Occupation + Education + Income# so voi mlm1 (est1) bo CITY, TripPurpose, TimeUseonBus, PSSW, PSSS, PSAB, PSEB, ATM, PPI, Occupation, Education, Income
est.chon <- estimate.mlogit(LOY ~ PSQ + IMA + PHB + PEV + SIM + SBE + EXB + MarriedStatus + FRE, data = DataLOY)
summary(est.chon)##
## successive function values within relative tolerance limit (reltol)
##
## Log-Likelihood: -361.1
## Null Log-Likelihood: -959.1
## Likelihood ratio index: 0.6235
## AIC: 744.1447
## BIC: 796.636
## Sample size: 873
## Iterations: 10
## Suggested |t-value| > 2.602294
## Convergence statistics: 4.87819e-07
##
## Estimated using BHHH in 0.339057 s.
##
## Coefficients :
## Estimate Std. Error t-value Pr(>|t|)
## asc.Normal 20.0251 1.35944 14.730 0.000e+00
## asc.Notloyal 16.6753 1.38317 12.056 0.000e+00
## FRE 0.3414 0.11614 2.940 3.287e-03
## PSQ -1.1710 0.18521 -6.322 2.574e-10
## IMA -0.9309 0.14292 -6.513 7.354e-11
## PHB -0.3178 0.15084 -2.107 3.512e-02
## PEV -0.3283 0.10252 -3.203 1.361e-03
## SIM -0.4785 0.09967 -4.800 1.583e-06
## SBE 0.5710 0.12713 4.492 7.072e-06
## EXB -1.0561 0.18229 -5.794 6.887e-09
## MarriedStatus -0.8192 0.22476 -3.645 2.677e-04
##
## Response variable: LOY
## Base choice name: Loyal
## Base choice index: 1
## Frequency of alternatives:
## Loyal Normal Notloyal
## 519 342 12
##
## Equations:
## ----------
## alternative intercept 1 2 3 4 5 6 7 8
## 1 Loyal:
## 2 Normal: asc.Normal + FRE + PSQ + IMA + PHB + PEV + SIM + SBE + EXB
## 3 Notloyal: asc.Notloyal + FRE + PSQ + IMA + PHB + PEV + SIM + SBE + EXB
## 9
## 1
## 2 + MarriedStatus
## 3 + MarriedStatus# THU
est.thu <- estimate.mlogit(LOY ~ PSQ + IMA + PHB + PEV + SIM + SBE + EXB + FRE + AGE, data = DataLOY)
summary(est.thu)##
## successive function values within relative tolerance limit (reltol)
##
## Log-Likelihood: -358.4
## Null Log-Likelihood: -959.1
## Likelihood ratio index: 0.6263
## AIC: 738.8591
## BIC: 791.3504
## Sample size: 873
## Iterations: 10
## Suggested |t-value| > 2.602294
## Convergence statistics: 4.304734e-07
##
## Estimated using BHHH in 0.360038 s.
##
## Coefficients :
## Estimate Std. Error t-value Pr(>|t|)
## asc.Normal 19.1986 1.32747 14.463 0.000e+00
## asc.Notloyal 15.8488 1.35061 11.735 0.000e+00
## AGE -0.3685 0.08599 -4.285 1.825e-05
## FRE 0.3567 0.11663 3.058 2.226e-03
## PSQ -1.1343 0.18211 -6.228 4.716e-10
## IMA -0.9388 0.14304 -6.563 5.260e-11
## PHB -0.3262 0.15181 -2.149 3.164e-02
## PEV -0.3578 0.10671 -3.353 7.986e-04
## SIM -0.4545 0.10047 -4.524 6.075e-06
## SBE 0.5714 0.12973 4.405 1.059e-05
## EXB -0.9889 0.18751 -5.274 1.336e-07
##
## Response variable: LOY
## Base choice name: Loyal
## Base choice index: 1
## Frequency of alternatives:
## Loyal Normal Notloyal
## 519 342 12
##
## Equations:
## ----------
## alternative intercept 1 2 3 4 5 6 7 8
## 1 Loyal:
## 2 Normal: asc.Normal + AGE + FRE + PSQ + IMA + PHB + PEV + SIM + SBE
## 3 Notloyal: asc.Notloyal + AGE + FRE + PSQ + IMA + PHB + PEV + SIM + SBE
## 9
## 1
## 2 + EXB
## 3 + EXBmlm.thu <- multinom(LOY ~ PSQ + IMA + PHB + PEV + SIM + SBE + EXB + FRE + AGE, data = DataLOY)## # weights: 48 (30 variable)
## initial value 959.088528
## iter 10 value 477.740893
## iter 20 value 362.737560
## iter 30 value 335.877884
## iter 40 value 330.609579
## iter 50 value 330.542482
## iter 60 value 330.536117
## final value 330.536051
## convergedsummary(mlm.thu)## Call:
## multinom(formula = LOY ~ PSQ + IMA + PHB + PEV + SIM + SBE +
## EXB + FRE + AGE, data = DataLOY)
##
## Coefficients:
## (Intercept) PSQ IMA PHB PEV SIM
## Notloyal 4.114588 -1.575694 0.09199217 -0.5684943 0.7036342 0.1162092
## Loyal -19.188824 1.138328 0.92964636 0.3295300 0.3579844 0.4664955
## SBE EXB FRE2days/month-2days/week
## Notloyal 0.6517687 -1.6080943 1.1895773
## Loyal -0.5748037 0.9759594 -0.1522929
## FRE2days/year-1day/month FRE<2 days/year AGE26-35 AGE36-45
## Notloyal 0.2565518 1.7674033 1.2166689 -0.7154157
## Loyal -1.0947415 -0.9531502 0.2433404 1.0958006
## AGE46-55 AGE>55
## Notloyal 1.5196208 -6.773994
## Loyal 0.7241241 1.726935
##
## Std. Errors:
## (Intercept) PSQ IMA PHB PEV SIM
## Notloyal 2.285067 0.4252624 0.5129671 0.5054483 0.3798203 0.3545387
## Loyal 1.437813 0.1870156 0.1663416 0.1519978 0.1069438 0.1084361
## SBE EXB FRE2days/month-2days/week FRE2days/year-1day/month
## Notloyal 0.4327721 0.5579435 1.3362777 1.2761417
## Loyal 0.1350506 0.1849188 0.2648883 0.3557135
## FRE<2 days/year AGE26-35 AGE36-45 AGE46-55 AGE>55
## Notloyal 1.0272813 1.0799892 1.973501 1.8389826 0.0002228833
## Loyal 0.3628632 0.2596796 0.357309 0.3920418 0.4605939293
##
## Residual Deviance: 661.0721
## AIC: 721.0721z.thu <- summary(mlm.thu)$coefficients/summary(mlm.thu)$standard.errors
z.thu## (Intercept) PSQ IMA PHB PEV SIM SBE
## Notloyal 1.800642 -3.705228 0.1793335 -1.124733 1.852545 0.3277757 1.506032
## Loyal -13.345838 6.086806 5.5887780 2.167992 3.347408 4.3020302 -4.256209
## EXB FRE2days/month-2days/week FRE2days/year-1day/month
## Notloyal -2.882181 0.8902171 0.2010371
## Loyal 5.277772 -0.5749325 -3.0775936
## FRE<2 days/year AGE26-35 AGE36-45 AGE46-55 AGE>55
## Notloyal 1.720467 1.1265566 -0.3625109 0.8263377 -30392.565961
## Loyal -2.626748 0.9370792 3.0668151 1.8470582 3.749366pvalue.thu <- (1-pnorm(abs(z.thu), 0, 1))*2
pvalue.thu## (Intercept) PSQ IMA PHB PEV
## Notloyal 0.07175933 2.112010e-04 8.576759e-01 0.26070224 0.0639476213
## Loyal 0.00000000 1.151855e-09 2.286731e-08 0.03015932 0.0008157111
## SIM SBE EXB FRE2days/month-2days/week
## Notloyal 7.430813e-01 1.320588e-01 3.949326e-03 0.3733493
## Loyal 1.692402e-05 2.079224e-05 1.307641e-07 0.5653369
## FRE2days/year-1day/month FRE<2 days/year AGE26-35 AGE36-45
## Notloyal 0.840669599 0.085347650 0.2599300 0.716970280
## Loyal 0.002086793 0.008620508 0.3487178 0.002163527
## AGE46-55 AGE>55
## Notloyal 0.40861253 0.0000000000
## Loyal 0.06473871 0.0001772825 ## Calculate p-value
z1 <- summary(mlm1)$coefficients/summary(mlm1)$standard.errors## Warning in sqrt(diag(vc)): NaNs produced
## Warning in sqrt(diag(vc)): NaNs producedz1## (Intercept) PSSW PSSS PSAB PSEB PSQ
## Notloyal 217.92387 -24.9231995 -11.4951720 6.6749771 5.6235132 -15.152639
## Loyal -11.32912 -0.1353095 0.9521445 -0.8197588 -0.7572714 5.095372
## IMA PHB PEV ATM PPI SIM
## Notloyal -2.898619 -2.976720 2.873763 -3.9873331 -0.03154036 -0.367476
## Loyal 5.919631 2.001343 3.260886 0.5990655 -0.48585487 4.423701
## SBE EXB MarriedStatusMarried OccupationFull.time.job
## Notloyal 8.487423 -12.380739 171.927739 -284.6152064
## Loyal -4.210312 5.200725 1.782625 0.9149627
## OccupationPart.time.job OccupationRetirement OccupationNo.job
## Notloyal -9.613738e+13 NaN NaN
## Loyal 2.441281e+00 1.947124 Inf
## OccupationHousewife OccupationOthers EducationUndergraduate
## Notloyal NaN -4.535915e+14 -1.055929e+06
## Loyal 1.292087 5.341461e-01 -5.727293e-01
## EducationHigh.school EducationPostgraduate EducationOthers
## Notloyal 206.309923 4.758582e+12 NaN
## Loyal -1.327067 -1.437668e+00 -1.240476
## Income5-10millions Income10-15millions Income>15millions CITYHoChiMinh
## Notloyal 83.7953773 2.855436e+07 -1.363892e+06 50.290015
## Loyal -0.9734612 8.907799e-01 -2.222121e-01 -0.238769
## FRE2days/month-2days/week FRE2days/year-1day/month FRE<2 days/year
## Notloyal 562887.079483 -17.635508 -9448.72568
## Loyal -0.341092 -2.851204 -2.58541
## TripPurposeStudying TripPurposeShopping TripPurposeEntertaining
## Notloyal 11796.379061 4.794780e+08 76.8445062
## Loyal 1.098083 2.447949e-01 -0.3339379
## TripPurposeOthers TimeUseonBusReading TimeUseonBusListening
## Notloyal 4.955682e+13 -8.387539e+07 -8.126190e+12
## Loyal 1.184914e-01 1.570650e+00 4.956869e-01
## TimeUseonBusNothing TimeUseonBusTalking TimeUseonBusOthers
## Notloyal 148.6342737 1.800242e+15 -5.942218e+14
## Loyal -0.3817261 -2.378453e-01 -3.201107e-01pvalue1 <- (1-pnorm(abs(z1), 0, 1))*2
pvalue1## (Intercept) PSSW PSSS PSAB PSEB PSQ
## Notloyal 0 0.0000000 0.0000000 2.472711e-11 1.871124e-08 0.000000e+00
## Loyal 0 0.8923672 0.3410237 4.123536e-01 4.488873e-01 3.480578e-07
## IMA PHB PEV ATM PPI
## Notloyal 3.748103e-03 0.002913502 0.004056132 6.682017e-05 0.9748386
## Loyal 3.226638e-09 0.045355492 0.001110645 5.491292e-01 0.6270701
## SIM SBE EXB MarriedStatusMarried
## Notloyal 7.132640e-01 0.000000e+00 0.000000e+00 0.0000000
## Loyal 9.702432e-06 2.550182e-05 1.985122e-07 0.0746474
## OccupationFull.time.job OccupationPart.time.job OccupationRetirement
## Notloyal 0.0000000 0.00000000 NaN
## Loyal 0.3602112 0.01463525 0.05151987
## OccupationNo.job OccupationHousewife OccupationOthers
## Notloyal NaN NaN 0.0000000
## Loyal 0 0.1963269 0.5932404
## EducationUndergraduate EducationHigh.school EducationPostgraduate
## Notloyal 0.000000 0.0000000 0.0000000
## Loyal 0.566828 0.1844865 0.1505283
## EducationOthers Income5-10millions Income10-15millions
## Notloyal NaN 0.0000000 0.0000000
## Loyal 0.2147993 0.3303242 0.3730473
## Income>15millions CITYHoChiMinh FRE2days/month-2days/week
## Notloyal 0.0000000 0.0000000 0.0000000
## Loyal 0.8241488 0.8112847 0.7330343
## FRE2days/year-1day/month FRE<2 days/year TripPurposeStudying
## Notloyal 0.000000000 0.000000000 0.000000
## Loyal 0.004355403 0.009726336 0.272168
## TripPurposeShopping TripPurposeEntertaining TripPurposeOthers
## Notloyal 0.0000000 0.0000000 0.0000000
## Loyal 0.8066152 0.7384264 0.9056783
## TimeUseonBusReading TimeUseonBusListening TimeUseonBusNothing
## Notloyal 0.0000000 0.0000000 0.0000000
## Loyal 0.1162639 0.6201153 0.7026645
## TimeUseonBusTalking TimeUseonBusOthers
## Notloyal 0.0000000 0.0000000
## Loyal 0.8120011 0.7488844z.chon <- summary(mlm.chon)$coefficients/summary(mlm.chon)$standard.errors
z.chon## (Intercept) PSQ IMA PHB PEV SIM SBE
## Notloyal 1.926352 -3.550998 0.2044365 -1.089607 1.885224 0.269504 1.197323
## Loyal -13.586621 6.298665 5.8250745 2.152874 3.124202 4.539857 -4.263940
## EXB MarriedStatusMarried FRE2days/month-2days/week
## Notloyal -2.848951 0.503446 0.8375361
## Loyal 5.678538 3.855801 -0.5971770
## FRE2days/year-1day/month FRE<2 days/year
## Notloyal 0.2452604 1.522864
## Loyal -2.9604636 -2.445024pvalue.chon <- (1-pnorm(abs(z.chon), 0, 1))*2
pvalue.chon## (Intercept) PSQ IMA PHB PEV
## Notloyal 0.05406044 3.837732e-04 8.380124e-01 0.27588622 0.059399542
## Loyal 0.00000000 3.002201e-10 5.708712e-09 0.03132854 0.001782879
## SIM SBE EXB MarriedStatusMarried
## Notloyal 7.875419e-01 2.311808e-01 4.386365e-03 0.6146507700
## Loyal 5.629247e-06 2.008535e-05 1.358511e-08 0.0001153515
## FRE2days/month-2days/week FRE2days/year-1day/month FRE<2 days/year
## Notloyal 0.4022913 0.806254805 0.12779269
## Loyal 0.5503892 0.003071765 0.01448424exp(coef(mlm.chon))## (Intercept) PSQ IMA PHB PEV SIM SBE
## Notloyal 7.137688e+01 0.2234782 1.105282 0.5938424 2.127263 1.104124 1.652222
## Loyal 3.213879e-09 3.1805481 2.562695 1.3824489 1.390146 1.630492 0.564378
## EXB MarriedStatusMarried FRE2days/month-2days/week
## Notloyal 0.2007176 1.603372 2.9796662
## Loyal 2.8124099 2.375481 0.8563217
## FRE2days/year-1day/month FRE<2 days/year
## Notloyal 1.3484084 4.4459992
## Loyal 0.3585408 0.4206406exp(confint(mlm.chon))## , , Notloyal
##
## 2.5 % 97.5 %
## (Intercept) 0.92823549 5488.5417774
## PSQ 0.09773438 0.5110026
## IMA 0.42334173 2.8857243
## PHB 0.23257095 1.5163064
## PEV 0.97051792 4.6627163
## SIM 0.53724322 2.2691584
## SBE 0.72627346 3.7586903
## EXB 0.06649549 0.6058690
## MarriedStatusMarried 0.25516173 10.0751910
## FRE2days/month-2days/week 0.23149521 38.3524594
## FRE2days/year-1day/month 0.12370054 14.6984423
## FRE<2 days/year 0.65165347 30.3334664
##
## , , Loyal
##
## 2.5 % 97.5 %
## (Intercept) 1.913638e-10 5.397581e-08
## PSQ 2.218900e+00 4.558964e+00
## IMA 1.867164e+00 3.517314e+00
## PHB 1.029445e+00 1.856501e+00
## PEV 1.130607e+00 1.709264e+00
## SIM 1.320249e+00 2.013640e+00
## SBE 4.338870e-01 7.341139e-01
## EXB 1.968239e+00 4.018643e+00
## MarriedStatusMarried 1.530210e+00 3.687669e+00
## FRE2days/month-2days/week 5.146929e-01 1.424708e+00
## FRE2days/year-1day/month 1.818120e-01 7.070574e-01
## FRE<2 days/year 2.101035e-01 8.421490e-01# Tinh xs uoc luong cho danh gia trung thanh cua 6 quan sat cuoi
xs <- fitted(mlm.chon)
tail(xs)## Normal Notloyal Loyal
## 868 0.45498404 2.719525e-03 0.54229643
## 869 0.34905208 3.300121e-05 0.65091492
## 870 0.09663928 2.342619e-06 0.90335838
## 871 0.02144760 5.231910e-06 0.97854717
## 872 0.04974320 5.491224e-06 0.95025131
## 873 0.97676754 5.593650e-03 0.01763881# Uoc tinh xs uoc luong TB (averaged predicted probabilities) cua long trung thanh tuong ung voi nhom doc than va co gia dinh voi cac gia tri khac nhau cua su cam nhan ve gia tri doi voi moi truong
Data.cu <- DataLOY[1:854,]
head(Data.cu)## ID AGE CITY FRE TripPurpose Departure
## 1 3 16-25 HoChiMinh >=3 days/week Studying Normal
## 2 4 16-25 HoChiMinh 2days/month-2days/week Entertaining Normal
## 3 5 16-25 HoChiMinh >=3 days/week Studying Peak-Hour
## 4 6 16-25 HoChiMinh >=3 days/week Studying Peak-Hour
## 5 7 16-25 HoChiMinh >=3 days/week Studying Peak-Hour
## 6 8 16-25 HoChiMinh >=3 days/week Studying Peak-Hour
## TimeUseonBus TravelTime PSSW PSSS PSAB PSEB PSQ SAT LOY IMA
## 1 Nothing 3.00 4.9 5.4 5.5 5.8 4.8 Satisfied Loyal 5.6
## 2 Nothing 2.00 3.4 2.4 3.5 6.0 4.6 Normal Normal 4.8
## 3 Nothing 0.17 3.9 3.4 4.5 4.5 2.7 Not satisfied Normal 3.0
## 4 Using.telephone 4.00 4.1 4.4 5.5 6.0 4.8 Normal Normal 5.0
## 5 Nothing 2.00 3.7 2.6 3.5 4.5 4.5 Normal Normal 4.6
## 6 Others 2.00 4.6 2.7 3.8 4.0 3.9 Normal Normal 3.0
## PHB PEV ATM PPI SIM PPA SBE EXB EC_Stop WC_Stop EC_Bus WC_Bus Gender
## 1 6.6 4.0 4.3 6.0 3.5 4.5 4.0 4.8 Never Ever Never Ever Female
## 2 4.0 5.3 4.1 4.0 4.0 4.0 4.8 4.2 Never Never Ever Never Female
## 3 5.0 5.8 3.1 2.7 3.5 4.5 6.0 4.0 Never Never Never Never Male
## 4 5.0 5.0 5.0 5.0 5.0 5.0 5.0 5.0 Never Ever Never Ever Male
## 5 5.6 6.0 3.3 3.0 5.5 5.0 5.0 4.2 Never Ever Never Ever Male
## 6 5.4 5.3 2.7 3.0 2.0 2.5 2.0 5.0 Never Ever Never Ever Female
## MarriedStatus Occupation Education Income
## 1 Single Students/Pupils Undergraduate <5millions
## 2 Single Students/Pupils Undergraduate <5millions
## 3 Single Students/Pupils Undergraduate <5millions
## 4 Single Students/Pupils Undergraduate <5millions
## 5 Single Students/Pupils Undergraduate <5millions
## 6 Single Students/Pupils High.school <5millionsdim(Data.cu)## [1] 854 33str(Data.cu)## 'data.frame': 854 obs. of 33 variables:
## $ ID : int 3 4 5 6 7 8 10 11 12 13 ...
## $ AGE : Factor w/ 5 levels "16-25","26-35",..: 1 1 1 1 1 1 3 1 1 2 ...
## $ CITY : Factor w/ 2 levels "DaNang","HoChiMinh": 2 2 2 2 2 2 2 2 2 2 ...
## $ FRE : Factor w/ 4 levels ">=3 days/week",..: 1 2 1 1 1 1 1 1 1 1 ...
## $ TripPurpose : Factor w/ 5 levels "Working","Studying",..: 2 4 2 2 2 2 1 2 2 2 ...
## $ Departure : Factor w/ 2 levels "Normal","Peak-Hour": 1 1 2 2 2 2 2 2 1 2 ...
## $ TimeUseonBus : Factor w/ 6 levels "Using.telephone",..: 4 4 4 1 4 6 4 4 3 3 ...
## $ TravelTime : num 3 2 0.17 4 2 2 2 2.5 1.5 2 ...
## $ PSSW : num 4.9 3.4 3.9 4.1 3.7 4.6 3.9 3.4 5.3 4 ...
## $ PSSS : num 5.4 2.4 3.4 4.4 2.6 2.7 4 2.7 3.1 4.7 ...
## $ PSAB : num 5.5 3.5 4.5 5.5 3.5 3.8 6 3.5 4.8 5.5 ...
## $ PSEB : num 5.8 6 4.5 6 4.5 4 4.5 6.3 7 4.5 ...
## $ PSQ : num 4.8 4.6 2.7 4.8 4.5 3.9 4.2 4.1 6 4.5 ...
## $ SAT : Factor w/ 3 levels "Not satisfied",..: 3 2 1 2 2 2 3 2 3 3 ...
## $ LOY : Factor w/ 3 levels "Normal","Notloyal",..: 3 1 1 1 1 1 3 1 3 3 ...
## $ IMA : num 5.6 4.8 3 5 4.6 3 5.8 5.8 6.8 6 ...
## $ PHB : num 6.6 4 5 5 5.6 5.4 5.8 5.6 7 6.4 ...
## $ PEV : num 4 5.3 5.8 5 6 5.3 6.3 6.5 7 5.3 ...
## $ ATM : num 4.3 4.1 3.1 5 3.3 2.7 4 4.1 4.4 3.9 ...
## $ PPI : num 6 4 2.7 5 3 3 6 2 3.3 5.3 ...
## $ SIM : num 3.5 4 3.5 5 5.5 2 6 5.3 4.3 6 ...
## $ PPA : num 4.5 4 4.5 5 5 2.5 4 4 4 4.3 ...
## $ SBE : num 4 4.8 6 5 5 2 5.3 4 5 3.8 ...
## $ EXB : num 4.8 4.2 4 5 4.2 5 5.3 5.3 5.3 5 ...
## $ EC_Stop : Factor w/ 2 levels "Ever","Never": 2 2 2 2 2 2 2 2 2 2 ...
## $ WC_Stop : Factor w/ 2 levels "Ever","Never": 1 2 2 1 1 1 2 1 2 2 ...
## $ EC_Bus : Factor w/ 2 levels "Ever","Never": 2 1 2 2 2 2 2 2 2 2 ...
## $ WC_Bus : Factor w/ 2 levels "Ever","Never": 1 2 2 1 1 1 2 1 2 2 ...
## $ Gender : Factor w/ 2 levels "Male","Female": 2 2 1 1 1 2 2 2 2 1 ...
## $ MarriedStatus: Factor w/ 2 levels "Single","Married": 1 1 1 1 1 1 2 1 1 1 ...
## $ Occupation : Factor w/ 7 levels "Students/Pupils",..: 1 1 1 1 1 1 2 1 1 7 ...
## $ Education : Factor w/ 5 levels "Secondary.school",..: 2 2 2 2 2 3 1 2 3 5 ...
## $ Income : Factor w/ 4 levels "<5millions","5-10millions",..: 1 1 1 1 1 1 1 1 1 1 ...attach(Data.cu)## The following objects are masked from DataLOY (pos = 3):
##
## AGE, ATM, CITY, Departure, EC_Bus, EC_Stop, Education, EXB, FRE,
## Gender, ID, IMA, Income, LOY, MarriedStatus, Occupation, PEV, PHB,
## PPA, PPI, PSAB, PSEB, PSQ, PSSS, PSSW, SAT, SBE, SIM, TimeUseonBus,
## TravelTime, TripPurpose, WC_Bus, WC_Stop## The following objects are masked from DataLOY (pos = 5):
##
## AGE, ATM, CITY, Departure, EC_Bus, EC_Stop, Education, EXB, FRE,
## Gender, ID, IMA, Income, LOY, MarriedStatus, Occupation, PEV, PHB,
## PPA, PPI, PSAB, PSEB, PSQ, PSSS, PSSW, SAT, SBE, SIM, TimeUseonBus,
## TravelTime, TripPurpose, WC_Bus, WC_Stop## The following objects are masked from DataLOY (pos = 33):
##
## AGE, ATM, CITY, Departure, EC_Bus, EC_Stop, Education, EXB, FRE,
## Gender, ID, IMA, Income, LOY, MarriedStatus, Occupation, PEV, PHB,
## PPA, PPI, PSAB, PSEB, PSQ, PSSS, PSSW, SAT, SBE, SIM, TimeUseonBus,
## TravelTime, TripPurpose, WC_Bus, WC_Stopmlm.chon1 <- multinom(LOY ~ PSQ + IMA + PHB + PEV + SIM + SBE + EXB + MarriedStatus + FRE, data = Data.cu)## # weights: 39 (24 variable)
## initial value 938.214895
## iter 10 value 470.687536
## iter 20 value 366.514280
## iter 30 value 331.951176
## iter 40 value 329.268243
## final value 329.265939
## convergedsummary(mlm.chon1)## Call:
## multinom(formula = LOY ~ PSQ + IMA + PHB + PEV + SIM + SBE +
## EXB + MarriedStatus + FRE, data = Data.cu)
##
## Coefficients:
## (Intercept) PSQ IMA PHB PEV SIM
## Notloyal 4.262732 -1.497636 0.1014050 -0.5205148 0.7516796 0.09559579
## Loyal -19.560706 1.164437 0.9081724 0.3370220 0.3258676 0.51909665
## SBE EXB MarriedStatusMarried FRE2days/month-2days/week
## Notloyal 0.5057077 -1.603065 0.4625811 1.0924135
## Loyal -0.6033075 1.048026 0.9032797 -0.2057903
## FRE2days/year-1day/month FRE<2 days/year
## Notloyal 0.3096484 1.4882731
## Loyal -1.0558184 -0.9212638
##
## Std. Errors:
## (Intercept) PSQ IMA PHB PEV SIM
## Notloyal 2.214285 0.4218362 0.4896842 0.4779046 0.4000302 0.3692560
## Loyal 1.455165 0.1849965 0.1621868 0.1534888 0.1083540 0.1115148
## SBE EXB MarriedStatusMarried FRE2days/month-2days/week
## Notloyal 0.4206059 0.5627949 0.9369001 1.3018995
## Loyal 0.1366363 0.1846856 0.2262039 0.2616726
## FRE2days/year-1day/month FRE<2 days/year
## Notloyal 1.2187101 0.9795909
## Loyal 0.3499571 0.3610208
##
## Residual Deviance: 658.5319
## AIC: 706.5319MarriedStatus <- rep(c("Single", "Married"), each = 427) # Lap lai 427 lan = 854 qs , PEV gia tri 1-7 lap lai 122 lan = 854qs va bang voi Data.cu
PEV <- rep(c(1:7), 122)
Data.moi <- data.frame(MarriedStatus, PEV)
pp.PEV <- cbind(Data.moi, predict(mlm.chon1, newdata = Data.moi, type = "probs", se = TRUE))
by(pp.PEV[, 3:5], pp.PEV$MarriedStatus, colMeans)## pp.PEV$MarriedStatus: Married
## Normal Notloyal Loyal
## 0.330183769 0.007744623 0.662071608
## ------------------------------------------------------------
## pp.PEV$MarriedStatus: Single
## Normal Notloyal Loyal
## 0.56072795 0.02337454 0.41589751library(reshape)##
## Attaching package: 'reshape'## The following object is masked from 'package:dplyr':
##
## rename## The following objects are masked from 'package:tidyr':
##
## expand, smithslpp <- melt(pp.PEV, id.vars = c("MarriedStatus", "PEV"), value.name = "Prob")
library(ggplot2)
# ggplot(lpp, aes(x = PEV, y = lpp, colour = MarriedStatus)) + geom_line() + facet_grid(variable ~ ., scales = "free")