• Journal of the Korean Society of Safety
• /
• v.27 no.2
• /
• pp.111-116
• /
• 2012

This study deals with the traffic accidents of circular intersections in Korea. The purposes are to comparatively analyze the characteristics by operational type, and to develop the models using the data of 82 intersections. In pursuing the above, this study gives particular emphasis to modeling such the accidents as the roundabout and rotary in urban area. The main results analyzed are as follows. First, the null hypotheses that the number of accidents are the same in both the urban and rural intersections, and roundabout and rotary in urban area, were analyzed to be rejected. Second, 3 accident models were developed, which were all statistically significant. The independent variables used in the above models are the ADT, number of approach lane, bus stop, parking facilities, and others. This study could be expected to give some implications to the traffic safety policy decision-making.

• Journal of Korean Society of Transportation
• /
• v.32 no.1
• /
• pp.73-81
• /
• 2014

This study dealt with the impacts of geometric structure on traffic accidents occurring on the interstates. There are standard values for the case of geometric structure which are recommended in the design guideline/policy; however, in the previous models, geometric variables were adapted as integrated ones as opposed to mixed ones in the real world so that derived models had a weakness to reflect the real. Therefore, using subdivided geometric variables, this study tried to derive the model which reflects the real world. In addition, by calculating elasticity, each variables' effect to the accidents are estimated. Hopefully, this study will help to establish the future guideline/policy of geometrics considering traffic safety.

• The Korean Journal of Applied Statistics
• /
• v.31 no.4
• /
• pp.475-484
• /
• 2018

Gamma generalized linear models are useful for non-negative and skewed responses. However, these models have received less attention than Poisson and binomial generalized linear models. In particular, hypothesis testing for the significance of regression coefficients has not been thoroughly studied. In this paper we assess the performance of various test statistics for gamma generalized linear models based on numerical studies. Our results show that the likelihood ratio test and F-type test are generally recommended and that the partial deviance test should be avoided in practice.

• Journal of Information Technology Services
• /
• v.10 no.4
• /
• pp.191-204
• /
• 2011

Recently, many software companies are trying to improve the software quality and project outcome with more costs and efforts in development time. In the software convergence and integration environments, it is required efforts to gain high quality of software. In other words, it is required to utilize software engineering knowledge and technology for higher software quality and better software project productivity. The Software development productivity can be varied by software process capability according to building a framework for software development, selection and use of appropriate technology, human resource management. Software process capability will influence software project outcome which is the general opinion. This study provides empirical evidence about software engineering efforts and investment approach to lead software project performance. We measured the software engineering efforts by SW engineering level and analyzed the corelation between software engineering level and schedule deviation. And, we verified that this performance is affected by the size of software company. As a result, software process capability is important to build a infrastructure and develop systematically software project. The higher software engineering level can lead to improved software project performance.

• Journal of the Korean Society of Safety
• /
• v.32 no.6
• /
• pp.131-136
• /
• 2017

This study deals with the small-size roundabout accidents. The purpose of this study is to analyze the characteristics of small-size roundabout accidents from developing various types of accident models, and to discuss the improvement countermeasures by vehicle type. The geometric characteristics of 36 roundabouts are surveyed, and the traffic accident data from 2008 to 2014 are collected and classified as those of car, truck and motor cycle. In this study, dependent variable is the number of accident and independent variables are such 15 variables as geometry and traffic volume. The main results are as follows. First, the null hypotheses that the size of roundabout and type of vehicle are not related to traffic accident are rejected. Second, 8 count data models which are all statistically significant are developed. Third, the number of circulatory roadway lane and sidewalk are selected as common variables of roundabout size. Finally, the number of entry and circulatory roadway lane are selected as common variables of vehicle type.

• Journal of the Korean Society of Safety
• /
• v.32 no.3
• /
• pp.123-129
• /
• 2017

This study deals with the traffic accident of truck at circular intersection. The purpose of this study is to develop the truck accident models based on type of accident and conflict. In pursuing the above, the study gives particular attentions to selecting the appropriate models among Poisson and Negative binomial models using statistical program LIMDEP 8.0. The traffic accident data from 2007 to 2014 are collected from TAAS data set of Road Traffic Authority. Such the dependent variable as number of truck accidents and the 24 independent variables as geometry, traffic volume and others are used. The main results are as follows. First, 5 Poisson models (${\rho}^2$ of 0.164~0.351) which are all statistically significant are selected. Second, the common variable based on type of accident and conflict is analyzed to be truck apron width. The specific variables are, however, evaluated to splitter island, area of splitter island, speed limit sign, truck apron, number approach road, circular intersection sign, speed hump and traffic volume. Finally, widening the truck apron width and improving the above specific variables are analyzed to be important for truck accident reduction at circular intersections.

• Journal of the Korean Society of Safety
• /
• v.33 no.3
• /
• pp.77-82
• /
• 2018

This paper aims to analyze the accident based on changing patterns of traffic culture index. For this purpose, this paper particularly focuses on classifying the traffic culture patterns and developing the traffic accidents using panel count data model. The main results are as follows. First, the traffic culture patterns are divided into 'increasing', 'decreasing' and 'other' patterns. The null hypotheses that the number of accident are the same over patterns are rejected. Second, 4 fixed effect negative binomial models which are all statistically significant are developed. Third, the regions with 'increasing' pattern are analyzed to be mostly the counties, and to demand the traffic law enforcement. Fourth, the regions with 'decreasing' pattern are evaluated to be mainly the districts and to require such the traffic culture as turn signal usage. Finally, the regions with 'other' pattern are analyzed to be mostly the cities and to ask for enhancing the level of traffic culture.

• International Journal of Highway Engineering
• /
• v.17 no.6
• /
• pp.65-73
• /
• 2015

PURPOSES : This study tries to develop the accident models of 4-legged signalized intersections in Busan Metropolitan city with random parameter in count model to understanding the factors mainly influencing on accident frequencies. METHODS : To develop the traffic accidents modeling, this study uses RP(random parameter) negative binomial model which enables to take account of heterogeneity in data. By using RP model, each intersection's specific geometry characteristics were considered. RESULTS : By comparing the both FP(fixed parameter) and RP modeling, it was confirmed the RP model has a little higher explanation power than the FP model. Out of 17 statistically significant variables, 4 variables including traffic volumes on minor roads, pedestrian crossing on major roads, and distance of pedestrian crossing on major/minor roads are derived as having random parameters. In addition, the marginal effect and elasticity of variables are analyzed to understand the variables'impact on the likelihood of accident occurrences. CONCLUSIONS : This study shows that the uses of RP is better fitted to the accident data since each observations'specific characteristics could be considered. Thus, the methods which could consider the heterogeneity of data is recommended to analyze the relationship between accidents and affecting factors(for example, traffic safety facilities or geometrics in signalized 4-legged intersections).

• International Journal of Highway Engineering
• /
• v.13 no.1
• /
• pp.13-19
• /
• 2011

The purpose of this study is to analyze the characteristics of traffic accidents and to develop the models by day and night-time in the case of the arterial link sections. In pursuing the above, this study uses the 224 accident data occurred at the 24 arterial link sections in Cheongju. The main results analyzed are as follows. First, it was analyzed that the number of accidents during day was more than night, but the accidents rate during night was higher than day. Second, four models which were all statistically significant were developed. Finally, the differences between the day and night models were comparatively analyzed using independent variables.

• Communications for Statistical Applications and Methods
• /
• v.25 no.2
• /
• pp.173-184
• /
• 2018

In medical or public health research, it is common to encounter clustered or longitudinal count data that exhibit excess zeros. For example, health care utilization data often have a multi-modal distribution with excess zeroes as well as a multilevel structure where patients are nested within physicians and hospitals. To analyze this type of data, zero-inflated count models with mixed effects have been developed where a count response variable is assumed to be distributed as a mixture of a Poisson or negative binomial and a distribution with a point mass of zeros that include random effects. However, no study has considered a situation where data are also censored due to the finite nature of the observation period or follow-up. In this paper, we present a weighted version of zero-inflated Poisson model with random effects accounting for variable individual follow-up times. We suggested two different types of weight function. The performance of the proposed model is evaluated and compared to a standard zero-inflated mixed model through simulation studies. This approach is then applied to Medicaid data analysis.