HYPOTHESIS TESTING OF ROAD TRAFFIC ACCIDENT DATA IN INDIA

International Journal of Civil Engineering and Technology (IJCIET) Volume 8, Issue 6, June 2017, pp. 430 435, Article ID: IJCIET_08_06_047 Available online at http://www.iaeme.com/ijciet/issues.asp?jtype=ijciet&vtype=8&itype=6 ISSN Print: 0976-6308 and ISSN Online: 0976-6316 IAEME Publication Scopus Indexed HYPOTHESIS TESTING OF ROAD TRAFFIC ACCIDENT DATA IN INDIA Pooja Sikdar Research Scholar, Kalinga University, Raipur, Chhattisgarh Prof. Ahsan Rabbani Assistant Professor, Kalinga University, Raipur, Chhattisgarh Prof. N.K. Dhapekar Assistant Professor, Kruti Institute of Technology and Engineering, Raipur, Chhattisgarh Dr. Govardhan Bhatt Professor, National Institute of Technology, Raipur, Chhattisgarh ABSTRACT Analysis of factors affecting road accidents is of utmost importance to reduce frequency of occurrence. This paper highlights hypothesis of data of road accidents of various parts of India. ANOVA is applied to carry hypothesis of data and to conclude the factors which promotes accidents. Preventive measures are been justified to reduce the frequency of road accidents. Research paper includes hypothesis and study of relationship between types of lanes, total no. of injuries, accidents, persons killed, types of vehicles, awareness of drivers and accidents, types of highways (National highway, State highway and Other roads) and no.of accidents, persons killed or injured. Z test is also applied to check the relationship between no. of accidents in National highway and State highway with Other roads. Keywords: Accidents, Hypothesis, Anova, Z Test. Cite this Article: Pooja Sikdar, Prof. Ahsan Rabbani, Prof. N.K. Dhapekar and Dr. Govardhan Bhatt. Hypothesis Testing of Road Traffic Accident Data in India. International Journal of Civil Engineering and Technology, 8(6), 2017, pp. 430 435. http://www.iaeme.com/ijciet/issues.asp?jtype=ijciet&vtype=8&itype=6 1. INTRODUCTION Accidents are unwanted suddenly happening incidence which results in loss of lives, loss of income, property and some times life time damages in body of human and vehicles [2]. The highest number of accidents were found in India as per WHO and mostly M2W riders and trucks have more number of accidents rate in year 2011[5]. Increasing fatalities rate require more attention to be controlled. Various investigations are carried out by other countries to find the accident data. India has adopted the technique of investigation done by police http://www.iaeme.com/ijciet/index.asp 430 editor@iaeme.com

Pooja Sikdar, Prof. Ahsan Rabbani, Prof. N.K. Dhapekar and Dr. Govardhan Bhatt S. no. authorities to find out the reason behind accidents. By comparing the fatalities rate year wise it is found that India has largest number of data as compared to other countries[6]. Some of the major reasons found for accidents are fault of driver, driver of other vehicle, pedestrian, passenger, cyclist, road condition, motor vehicle, weather condition and poor light etc; but some times barrier of roads are also being responsible for accidents such as median, road side barriers etc. Various safety measures such as speed breakers and crash cushions are made to make traffic calm and reduce their speeds, by government the new strategy was planned for road safety which is also called 4E s of safety in road viz. education, enforcement, engineering, environment and emergency care[7]. In some survey it is seen that the reasons behind accidents are high speed and wrong side overtaking and tailgating[8]. Accidents are also occurred due to the mental aggressiveness and it was more at the age of 19-25 years old and it is more in females as compared to males, it is also more in peoples having higher degree like post graduates and mostly in students according to profession wise, and specially those who have less driving experience. Some of the major reasons are found for accidents by reviewing various literature's viz over speeding, drunken drivers, driver s distraction, red light jumping and avoiding safety materials such as helmets and seat belts etc and some special type of reason was also found out that condition of road was so bad that it plays role in accidents after it being renewed by government, it doesn't last for six months[9]. After study the statistic of accidents it is important to aware people for taking precautions and safety measures to mitigate or avoid accidents to a certain extent. And this awareness can make by organizing camp for road safety measures and accidents reasons awareness. 2. METHODOLOGY AND CALCULATIONS ANOVA technique is applied in hypothesis as this technique helps in performing this test in one go and therefore it is considered to be important technique of analysis. This technique is possible to draw inferences weather the sample have been drawn from population having the same mean. It has been applied in the field of highway engineering. Hypothesis through Z test is also carried out and is tested on the basis of information obtained from the data. Carriageway width Table 1 Accidents classification in Bhopal, Lane wise. Fatal Grievous Injury Total no. of accidents Total no. of person s killed Total no. of person s injured 1. Single Lane 3926 3653 28768 4453 34633 2. Two Lane 1555 1356 12086 1864 13348 3. Three Lane 168 191 1138 168 1477 4. Four Lane 212 222 1860 185 1596 Null hypothesis is accident classification (lane wise) which can be attributed to chance ie, atleast two means are equal [1]. Applying One-Way Anova- N=20, K=4. Correction factor=(σt ij ) 2 /N, T ij =entries in each row and column, Correction factor=636857694.1. SST(sum of squares total)=σ(t 2 ij )-correction factor, SST=1780448441. SSB(sum of squares between group column)={(σt i ) 2 /N i }-correction factor, SSB=688982097.9. SSE(sum of squares with in group error) =SST-SSB, SSE=1091466343. MSB(mean square between group column)=ssb/k-1, MSB=229660699.3. MSE(mean square with in group error)=ssw/n-k, MSE=68216646.44. Test statistic =F K-1 N-K =MSB/MSE, F 16 3 =3.366637196. F 3 16 =3.24@α=5%. As F statistic (F 3 16 ) is greater than the tabulated value of 3.24 at 5 % level of significance so, null hypothesis is rejected. http://www.iaeme.com/ijciet/index.asp 431 editor@iaeme.com

Hypothesis Testing of Road Traffic Accident Data in India Parameters Table 2 Share of different vehicles primarily responsible in road accidents, fatal accidents, persons killed and persons injured. Two wheelers Auto Rickshaws Cars, jeeps and taxis Buses Trucks, Tempos, Tractors and other articulated vehicles Other motor vehicles Other vehicles/ Objects Accidents 26.3 6.4 22.2 8.6 21.1 9.9 5.5 Fatal accidents 22.2 4.2 18.4 8.8 26.8 11.9 7.7 Persons killed 21.4 4.3 18.5 9.1 27.2 11.3 8.2 Persons injured 24.6 7.6 23.2 11 19.8 9.1 4.7 Null hypothesis is accidents due to various types of vehicles can be attributed to chance [4]. The data of 2013 is tabulated above which is hypothesized by One-Way Anova technique. N=28, K=4. Correction factor=(σt ij ) 2 /N, T ij =entries in each row and column, Correction factor=5714.285714. SST(sum of squares total)=σ(t ij 2 )-correction factor, SST=1659.634286. SSB(sum of squares between group column)={(σt i ) 2 /N i }-correction factor, SSB=2.8571 x 10-7. SSE(sum of squares with in group error)=sst-ssb, SSE=1659.634286. MSB(mean square between group column)=ssb/k-1, MSB=9.523666667 x 10-8. MSE(mean square with in group error)=ssw/n-k, MSE=69.15142858. Test statistic =F N-K K-1 =MSB/MSE, F 24 3 =1.377219077 x 10-9. F 24 3 =4.72@α=1%; F 24 3 =3.01@α=5%. As F statistic (F 24 3 ) is smaller than the tabulated value at 1% and 5 % level of significance so, null hypothesis is accepted. Table 3 Awareness of drivers on different safety dimensions as mentioned in the content of the questionnaire. S. No Contents Drivers were Aware Drivers were Not aware 1. Awareness about vehicle registration documents and type of vehicle 66 34 to be used 2. Awareness about maintaining minimum distance between two cars 52 48 while driving 3. General Awareness about parking place 39 61 4. Awareness about drinking driving 69 31 5. Awareness about Seat belt usage 89 11 6. Right of way of the vehicle while driving on hill roads 89 11 7. Awareness about White continuous or Broken lines 27 73 8. Awareness about by left lane on road 52 48 9. Awareness about Starting the vehicle on road during high traffic 66 34 volume 10. Safety Awareness for pedestrians during rainy seasons 52 48 11. Awareness about right way of Using Horn 58 42 12. Awareness about right way of Emergency stopping 40 60 13. Awareness about controlling speed while descending 82 18 14. Awareness about Use of Hazard Lights 57 43 15. Awareness about use of mobile 77 23 16. Road Sign Major Road Ahead 57 43 17. Road Sign about Steep decent 84 16 18. Road Sign about Give Way 61 39 19. Road Sign about Vehicle Prohibited on Both Direction 74 26 20. Road Sign about No Crossing 68 32 Null hypothesis is safety dimensions can be attributed to chance [3]. Applying Two Way Anova- N=40, R=20, C=2 Correction factor=(σt ij ) 2 /N, T ij =entries in each row and column, http://www.iaeme.com/ijciet/index.asp 432 editor@iaeme.com

Pooja Sikdar, Prof. Ahsan Rabbani, Prof. N.K. Dhapekar and Dr. Govardhan Bhatt Correction factor=100000. SST(sum of squares total)=σ(t ij 2 )-correction factor, SST=17538. SSR(sum of squares between rows)={(σt i ) 2 /N i }-correction factor, SSR=0. SSC(sum of squares between columns)={(σt j ) 2 /N j }-correction factor, SSC=6708.1. SSE(sum of squares residual error)= SST-(SSR+SSC), SSE=10829.9. MSC(mean square between columns)=ssc/c-1, MSC=6708.1. MSR(mean square between rows)=ssr/r-1, MSR=0. MSE(mean square residual error)=sse/(c-1)x(r-1), MSE=569.9947368. Test statistic - F Column (C-1)x(R-1) C-1 =MSC/MSE, F Row (C-1)x(R-1) R-1 =MSR/MSE. F Column 19 1 = 11.76870516, F Row 19 19 =0. F Column 19 1 =4.38@α=5%, F Column 19 1 =8.18@α=1%. F Row 19 19 =2.19@α=5%, F Row 19 19 =3.23@α=1%. As F statistic (F Column 19 1 ) is greater than the tabulated value at 1% and 5 % level of significance so, null hypothesis is rejected. And F statistic (F Row 19 19 ) is smaller than the tabulated value at 1% and 5 % level of significance so, null hypothesis is accepted. Table 4 Number of accidents, persons killed and injured as per road classification. Road Classification National State Other Highways Highways Roads No. of Accidents 1,36,786 1,24,358 2,25,332 No. of Persons Killed 45,612 40,768 51,192 No. of Persons Injured 1,43,107 1,36,790 2,14,996 Null hypothesis is number of fatalities can be attributed to chance [4].The data of 2013 is tabulated above which is hypothesized by Two-Way Anova technique. N=9, R=3, C=3. Correction factor=(σt ij ) 2 /N, T ij =entries in each row and column, Correction factor=1.391143291 x 10 11. SST(sum of squares total)=σ(t 2 ij )-correction factor, SST=3.761300588 x 10 10. SSR(sum of squares between rows)={(σt i ) 2 /N i }-correction factor, SSR=2.77203493 x 10 10. SSC(sum of squares between columns)={(σt j ) 2 /N j }-correction factor, SSC=7118566394. SSE(sum of squares residual error)= SST-(SSR+SSC), SSE=2774090186. MSC(mean square between columns)=ssc/c-1, MSC=3559283197. MSR(mean square between rows)=ssr/r-1, MSR=1.386017465 x 10 10. MSE(mean square C- residual error)=sse/(c-1)x(r-1), MSE=693522546.5. Test statistic- F Column (C-1)x(R-1) 1 =MSC/MSE, F R-1 Row (C-1)x(R-1) =MSR/MSE. F 2 Column 4 =5.132180943, F 2 Row 4 = 19.98518249. F Column 4 2 =6.94@α=5%, F Column 4 2 =18.20@α=1%. F Row 4 2 =6.94@α=5%, F Row 4 2 =18.20@α=1%. As F statistic (F Column 4 2 ) is smaller than the tabulated value at 1% and 5 % level of significance so, null hypothesis is accepted. And F statistic (F Row 4 2 ) is greater than the tabulated value at 1% and 5 % level of significance so, null hypothesis is rejected. In the year 2014 number of road accidents occurring in NH are 137903 and on SH are 123408. It is likely that the road accidents happening on other roads are 228089[4]. As per Z-test- x =Σx/2; x =130655.5, μh=228089. No. of accidents (x) ( x- x ) 2 National Highway 137903 52526256.25 State Highway 123408 52526256.25 Σ 105052512.5 σ = [{1/(n-1)} x Σ( x- x ) 2 ], n=no. of entries, σ =10249.51279. Z= ( x-μh)/σ, Z= - 9.50615917. Normal distribution of Z test lie between ± 2.58. Calculated value is beyond the zone of acceptance. So null hypothesis is rejected with 1%, 5% and 10% level of significance. In the year 2015 number of accidents occurring in NH are 142268 and on SH are 120518. It is likely that the road accidents happening on other roads are 238637[4].As per Z-test- x =Σx/2; x =131393, μh=238637. http://www.iaeme.com/ijciet/index.asp 433 editor@iaeme.com

Hypothesis Testing of Road Traffic Accident Data in India No. of accidents (x) ( x- x ) 2 National Highway 142268 118265625 State Highway 120518 118265625 Σ 236531250 σ = [{1/(n-1)} x Σ( x- x ) 2 ], n=no. of entries, σ =15379.57249. Z= ( x-μh)/σ, Z= - 6.973145715. Normal distribution of Z test lie between ± 2.58. Calculated value is beyond the zone of acceptance. So null hypothesis is rejected with 1%, 5% and 10% level of significance. 3. CONCLUSIONS There is no relationship between the types of lanes and total number of injuries, accidents and persons killed. Traffic rules should be implemented stringently to avoid road accidents. Number of accidents, persons injured and killed depends upon various types of vehicles like two wheeler, auto rickshaws, cars, jeeps, buses, trucks, tractors, other articulated vehicles, motor vehicles and objects etc. Mix drinking and driving may prove to be dangerous. Use of mobiles and disobey of traffic rules and signs affects the number of accidents as far as the types of vehicles are concerned. It is well justified that road safety awareness programs will reduce number of accidents. Number of persons killed and injured depends on the type of highway i.e. national highway, state highway and other roads. Number of accidents occurring in year 2014 and 2015 happening on other roads may likely to be different than two lakhs twenty eight thousand and two lakhs thirty eight thousand respectively. REFERENCES [1] Sharma N et al; (2014).Road Safety and Accident Prevention-An Indian Perspective. International Journal of Innovative Engineering Research (IJIER). Volume 1 Issue 1, July 2014. [2] Agbeboh G. U. et al; (2013).Empirical analysis of road traffic accidents: A case study of Kogi State, North-Central Nigeria. International Journal of Physical Sciences. Vol. 8(40), pp. 1923-1933, 30 October, 2013 DOI: 10.5897/IJPS2013.3978 ISSN 1992-1950 2013 Academic Journals. [3] Neelima Chakrabarty et al; (2013).A Survey on Awareness of Traffic Safety among Drivers in Delhi, India. The SIJ Transactions on Industrial, Financial & Business Management (IFBM). Vol. 1, No. 2, May-June 2013. ISSN: 2321 242X 2013. [4] www.morth.nic.in [5] Swastik Narayan et al; Characteristics of Fatal Road Traffic Accidents on Indian Highways. JP Research India Pvt Ltd. [6] Kaushal D Shah et al; A Comparative Study of Accident Safety on Haryana Roads. IOSR Journal of Mechanical and Civil Engineering (IOSR-JMCE). e-issn: 2278-1684, p-issn: 2320-334X. PP 13-16. [7] G.Udayakumar et al; (2014).Devising a New Technique to Reduce Highway Barrier Accidents. IJRET: International Journal of Research in Engineering and Technology. eissn: 2319-1163. pissn: 2321-7308. Volume: 03 Special Issue: 07 May-2014. [8] Neelima Chakrabarty et al; (2013).Aggressive Driving Case Studies and Mitigations in India. International Journal of Scientific and Research Publications, Volume 3, Issue 2, February 2013 1 ISSN 2250-3153. [9] Geethabai et al; (2016).A Literature Study On Road Accidents Statistics And Reasoning. (IJITR) International Journal of Innovative Technology and Research. Volume No.4, Issue No.6, October November 2016, 4979-4984. http://www.iaeme.com/ijciet/index.asp 434 editor@iaeme.com

Pooja Sikdar, Prof. Ahsan Rabbani, Prof. N.K. Dhapekar and Dr. Govardhan Bhatt [10] N.K.Dhapekar (2016).SHM of ordinary portland cement concrete structures using XRD. International Journal of Applied Engineering Research. [11] N.K.Dhapekar (2015).Study of phase composition of Ordinary Portland Cement concrete using X-Ray diffraction. International Journal of Scientific and Engineering Research. [12] N.K.Dhapekar (2014).Structural health monitoring of concrete structures evaluating elastic constants and stress strain parameters by X-ray diffraction technique. International journal of civil engineering and technology(ijciet). [13] Ahsan Rabbani et al; (2017).Hypothesis of data of road accidents in India-Review. International Journal of Civil Engineering and Technology (IJCIET), Volume 8, Issue 6, June 2017. [14] Ahsan Rabbani et al; (2017).Study of rigid pavements-review. International Journal of Civil Engineering and Technology (IJCIET), Volume 8, Issue 6, June 2017. [15] B.Naga Kiran, Dr. N. Kumara Swamy and Dr. C. Sashidhar, Prediction of Road Accident Modelling for Indian National Highways International Journal of Civil Engineering and Technology (IJCIET) Volume 8, Issue 1, January 2017, pp. 789 802 http://www.iaeme.com/ijciet/index.asp 435 editor@iaeme.com