Statistical Modelling for Job Satisfaction of Maharashtra State Road Transport Corporation Key Members
Avinash S. Jagtap, Nilambari A. Jagtap, Pooja S. ZanjurneJob satisfaction, Health problem, Stepwise regression14-19DecemberMobility is one of the characteristics of human beings; movement is the feature of human nature which simply means mere movement of persons from one place to another for getting comfort journey. Bus conductors and Drivers of MSRTC are serving to society. Moreover they can be considered as face of MSRTC as they are having direct contact with passengers. In this paper we study various factors affecting on the job satisfaction of conductors and drivers. Bus conductors and drivers are randomly selected as a sample. Their job satisfaction is analyzed through structured questionnaire. The first part of the questionnaire is focused on the personal information like age, education, number of family members, earning members etc. and second part is focused on the official information of the respondents like number of charge sheet, presentism and abstentism. The statistical analysis is done by using the statistical tools like Chi square test, linear discriminant analysis, multiple logistic regression. It is clear from this study that the working environmental factor affects the job satisfaction. The accuracy of fitted model is 67%. Linear Discriminant model has 96% accuracy.
Introduction: Traffic staff is one of the important factors for operating functions. Traffic staff includes drivers, conductors, traffic controllers, line checking staff, traffic inspectors and other traffic staff. The drivers and conductors are appointed with their services based on the sanctioned schedules, their work norms are set according to the provisions of Motor Transport Workers Act 1961 (MTW Act) which prescribes a daily steering duty of 8 hours and spread over duty of 12 hours.
Maharashtra State Road Transport Corporation(MSRTC) is a leading passenger road transport organization in India having fleet of over 18,000 buses and operating over 1,05,000 bus trips daily covering 58 lacks of Kms. MSRTC is aimed to provide punctual, safe, comfortable and economical services to the passengers. It spread in all over Maharashtra nearly 88% villages. MSRTC not only working in the area of providing transport facility to passengers but also it provide service carriage of parcels currier and Allied material by using the carriage of buses with consideration of given information MSRTC obviously need competent, motivated, trained ,satisfied and alert Human Resource for running one of the biggest public transport organizations of India. The following are two personnel who constitute the subject matter of study. As we know the conductors and drivers are Crew members (Conductors and Drivers) of MSRTC.
A bus conductor has to deal with thousands of passengers every day. Buses are two most important forms of public transport the other being trains. The burden of dealing with increasing number of commuters is very heavy. He has to behave well with the passenger and see to it that everyone is issued a right ticket, and thus increase passenger satisfaction, and in the process earn maximum revenue for the organization.
Job satisfaction is most important thing in any job. If the key members satisfy with their job then it tends to high quality of work they do. But they do not satisfy with their job it may be a serious issue because the key members have direct contact with public so that if they are unsatisfied with their job their behaviour is rood about passenger and this is one kind of the loss to MSRTC so that the checking crew member satisfaction with their job is important thing. With the help of this paper we will find which factor are strongly affecting on job satisfaction of key members. Also we find the association between variables.
Material and Methods:
The present study is undertaken on the employees (Drivers and conductors) of selected MIDC ST Depot, Baramati, Dist. Pune of MSRTC. The study is mainly based on primary data collected from the respondent the drivers, conductors, driver cum conductors of MSRTC by way of structured questionnaire, to be precise, pre-coded. This interview schedule has been organised in two parts. The first part of the questionnaire is focused on the personal information like Age, Education, No. of family Members, Earning Members etc. and second part is focused on the official information of the respondents like number of charge sheet, presentism and abstentism. Here we use the variable like Ts/TTS/DW is the permanent, temporary, and contract basis.
Multiple logistic regression: This is one type of generalized linear model (GLM). Multiple Logistic Regression is used when response variable is binary. It is utilized for predicting the outcome of a categorical variable based on one or more independent variables. Specific form of logistic regression model
Π(x)=exp^(g(x))/[1+exp^g(x)]
Generalized linear model: In a GLM the response variable distribution must only be a member of the exponential family which includes the normal, Poisson, binomial, exponential and gamma distribution as members. Distribution that the members of the exponential family have the general form
g(x)=β0+β1X1+…………….+βp Xp
We are interested in determining the future event, our response is binary for that purpose we use the GLM after getting the significant variable from cluster analysis we able to fit generalized linear model for our data. From the fitted model we can predict the future event.
Stepwise regression: Stepwise regression includes regression models in which the choice of predictive variables is carried out by an automatic procedure. Forward Selection, which involves starting with no variables in the model, testing the addition of each variable using a chosen model comparison criterion, adding the variable (if any) that improves the model the most and repeating this process until none improves the model. AIC –Akien’s Information Criteria: The general for calculating AIC is
AIC= -2*ln (likelihood) +2*k
AIC can also be calculated using residual sum of squares from regression,
AIC=n*ln (RSS/n) +2*k
Where, n is the number of data points, RSS is the residual sum of squares, AIC requires a bias-adjustment small sample size. From the stepwise regression we get the significant variables for regression models for our data. The model which has smallest value of AIC as compare to the other models it gives the best regression model for our data.
Confusion Matrix: A Confusion Matrix is a visual performance assessment of a classification algorithm in the form of a table layout or matrix. Each column of the matrix represents predicted classifications and each row represents actual defined classifications. This representation is a useful way to help evaluate a classifier model. A well behaved model should produce a balanced matrix and have consisted percent correctness numbers for accuracy, recall, precision and F measure. Accuracy= (True positive + false positive)/ (total)
Precision=True positives / (true positives + false positives).
To determine the accuracy of the fitted model we use the confusion matrix for the response variable.
Linear Discriminant Analysis(LDA): Linear Discriminant analysis and the related fisher’s linear discriminant are methods used in statistics, pattern reorganization and machine learning to find a linear combination of features which characterizes or separate two or more class of objects or events.LDA is a closely related to analysis of variance and regression analysis, which also attempt to express one dependent variable as a linear combination of other features or measurements.
Data Analysis: The analysis includes chi-square test of independence to examine the nature of relationship between variables. The collected data will be analysed by using Multiple Logistic Regression Analysis.
Abbreviations : Faci: Facilities, San: Sanitary, Clean : Cleanness, Spac: Space, Rest : Rest house facility, Guest: Guest house facility, df : Degrees of freedom, Dist :Distance, SY: Service years, Pay: Payment, EM: Earning members , HP: Health problem, Ts: Time skill, famliym: Family members, JS : Job satisfaction, TPM : Transition probability matrix.
Association between job satisfaction and various environmental factors
Logistic regression
Model 1:
The regression equation is,
Job satisfaction ~Age+Space+Distance+Serviceyears+Payment+Earningmembers+Healthproblem+Timeskill+ status+Guest house
Start: AIC=130.5
Coefficients:
(Intercept) Age spac Dist
-1.169e+00 2.422e-02 1.408e+00 1.175e-03
SY familym Pay EM
-4.508e-02 -1.361e-02 3.914e-05 3.110e-01
HP Ts status Guest
-9.731e-01 -5.486e-01 -7.272e-01 4.700e-01
Degrees of Freedom: 91 Total (i.e. Null); 80 Residual
Null Deviance: 126
Residual Deviance: 106.5 AIC: 130.5
>summary(model)
Call:
glm(formula = JS ~ Age + spac + Dist + SY + familym + Pay + EM +
HP + Ts + status + Guest, family = "binomial", data = mydata)
Model 2:
> X=step(model)
Start: AIC=130.46
JS ~ Age + spac + Dist + SY + familym + Pay + EM + HP + Ts +
status + Guest
Model 3:
Step: AIC=126.67
JS ~ spac + Dist + SY + Pay + EM + HP + Ts + status +
Guest
Model 4:
Step: AIC=124.96
JS ~ spac + Dist + SY + Pay + EM + HP + Ts + Guest
Model 5:
Step: AIC=123.37
JS ~ spac + SY + Pay + EM + HP + Ts + Guest
Model 6:
Step: AIC=121.93
JS ~ spac + SY + EM + HP + Ts + Guest
Model 7:
Step: AIC=120.01
JS ~ spac + EM + HP + Ts + Guest
Model 8:
Step: AIC=118.56
JS ~ spac + HP + Ts + Guest
Model 9:
JS ~ spac + HP + Ts
Deviance Residuals:
Min 1Q Median 3Q Max
-1.4745 -0.8385 -0.5272 0.9069 2.0210
Degrees of Freedom: 91 Total (i.e. Null); 88 Residual
Null Deviance: 126
Residual Deviance: 109.5 AIC: 117.5
Pridictor
Estimate
Std.Error
Zvalue
Pr(>|Z|)
Intercept
-0.08131
0.65692
-0.124
0.9015
Spac
1.5404
0.47891
3.216
0.0013
HP
-1.0388
0.65307
-1.591
0.1117
Ts
-0.78322
0.49801
-1.573