Applying Classification Trees to Analyze Electrical Contractors' Accidents

Pouya Gholizadeh, Behzad Esmaeili

Research output: Chapter in Book/Report/Conference proceedingConference contribution

1 Citation (Scopus)

Abstract

When considering the impermanent environment of construction projects for electrical contractors, it is not always easy to discover the root cause of an accident. In addition, although previous studies have indicated that most accidents can be attributed to spatial and temporal interactions of multiple features, a limited number of studies have empirically explored these relationships. One data-mining technique that can be used to address these limitations and investigate the causal relationship among events is a classification and regression tree (CART). The current study applied CART analysis to explore the relationship between major characteristics of a construction project and/or work attributes, and injury outcomes. CART is an appropriate statistical analytical tool for analyzing data with non-linear relationships, a high-order of interactions, and a large number of missing values. To analyze a reliable and representative database of accidents involving electricians, content analysis was conducted on 320 accident reports obtained from the Occupational Safety and Health Administration database to identify attributes that led to accidents. All accidents occurred between 2009 and 2012. The outcomes of accidents (fatality versus no-fatality) were considered as dependent variables, and features such as project cost, project type, project end-use, and task attributes that cause accidents were considered as independent variables. The findings describe how specific characteristics of a project (e.g. cost) or a construction task (e.g. working near wiring) can be used to predict the probability of fatality for an electrician on a jobsite. The results of the study also can be used by safety managers to revise safety practices and training programs.

Original languageEnglish (US)
Title of host publicationConstruction Research Congress 2016: Old and New Construction Technologies Converge in Historic San Juan - Proceedings of the 2016 Construction Research Congress, CRC 2016
PublisherAmerican Society of Civil Engineers (ASCE)
Pages2699-2708
Number of pages10
ISBN (Electronic)9780784479827
DOIs
StatePublished - 2016
EventConstruction Research Congress 2016: Old and New Construction Technologies Converge in Historic San Juan, CRC 2016 - San Juan, Puerto Rico
Duration: May 31 2016Jun 2 2016

Other

OtherConstruction Research Congress 2016: Old and New Construction Technologies Converge in Historic San Juan, CRC 2016
CountryPuerto Rico
CitySan Juan
Period5/31/166/2/16

Fingerprint

Contractors
Accidents
Electric wiring
Data mining
Costs
Managers
Health

ASJC Scopus subject areas

  • Civil and Structural Engineering
  • Building and Construction

Cite this

Gholizadeh, P., & Esmaeili, B. (2016). Applying Classification Trees to Analyze Electrical Contractors' Accidents. In Construction Research Congress 2016: Old and New Construction Technologies Converge in Historic San Juan - Proceedings of the 2016 Construction Research Congress, CRC 2016 (pp. 2699-2708). American Society of Civil Engineers (ASCE). https://doi.org/10.1061/9780784479827.269

Applying Classification Trees to Analyze Electrical Contractors' Accidents. / Gholizadeh, Pouya; Esmaeili, Behzad.

Construction Research Congress 2016: Old and New Construction Technologies Converge in Historic San Juan - Proceedings of the 2016 Construction Research Congress, CRC 2016. American Society of Civil Engineers (ASCE), 2016. p. 2699-2708.

Research output: Chapter in Book/Report/Conference proceedingConference contribution

Gholizadeh, P & Esmaeili, B 2016, Applying Classification Trees to Analyze Electrical Contractors' Accidents. in Construction Research Congress 2016: Old and New Construction Technologies Converge in Historic San Juan - Proceedings of the 2016 Construction Research Congress, CRC 2016. American Society of Civil Engineers (ASCE), pp. 2699-2708, Construction Research Congress 2016: Old and New Construction Technologies Converge in Historic San Juan, CRC 2016, San Juan, Puerto Rico, 5/31/16. https://doi.org/10.1061/9780784479827.269
Gholizadeh P, Esmaeili B. Applying Classification Trees to Analyze Electrical Contractors' Accidents. In Construction Research Congress 2016: Old and New Construction Technologies Converge in Historic San Juan - Proceedings of the 2016 Construction Research Congress, CRC 2016. American Society of Civil Engineers (ASCE). 2016. p. 2699-2708 https://doi.org/10.1061/9780784479827.269
Gholizadeh, Pouya ; Esmaeili, Behzad. / Applying Classification Trees to Analyze Electrical Contractors' Accidents. Construction Research Congress 2016: Old and New Construction Technologies Converge in Historic San Juan - Proceedings of the 2016 Construction Research Congress, CRC 2016. American Society of Civil Engineers (ASCE), 2016. pp. 2699-2708
@inproceedings{2e0796fddd44438cb4499a991b31005a,
title = "Applying Classification Trees to Analyze Electrical Contractors' Accidents",
abstract = "When considering the impermanent environment of construction projects for electrical contractors, it is not always easy to discover the root cause of an accident. In addition, although previous studies have indicated that most accidents can be attributed to spatial and temporal interactions of multiple features, a limited number of studies have empirically explored these relationships. One data-mining technique that can be used to address these limitations and investigate the causal relationship among events is a classification and regression tree (CART). The current study applied CART analysis to explore the relationship between major characteristics of a construction project and/or work attributes, and injury outcomes. CART is an appropriate statistical analytical tool for analyzing data with non-linear relationships, a high-order of interactions, and a large number of missing values. To analyze a reliable and representative database of accidents involving electricians, content analysis was conducted on 320 accident reports obtained from the Occupational Safety and Health Administration database to identify attributes that led to accidents. All accidents occurred between 2009 and 2012. The outcomes of accidents (fatality versus no-fatality) were considered as dependent variables, and features such as project cost, project type, project end-use, and task attributes that cause accidents were considered as independent variables. The findings describe how specific characteristics of a project (e.g. cost) or a construction task (e.g. working near wiring) can be used to predict the probability of fatality for an electrician on a jobsite. The results of the study also can be used by safety managers to revise safety practices and training programs.",
author = "Pouya Gholizadeh and Behzad Esmaeili",
year = "2016",
doi = "10.1061/9780784479827.269",
language = "English (US)",
pages = "2699--2708",
booktitle = "Construction Research Congress 2016: Old and New Construction Technologies Converge in Historic San Juan - Proceedings of the 2016 Construction Research Congress, CRC 2016",
publisher = "American Society of Civil Engineers (ASCE)",

}

TY - GEN

T1 - Applying Classification Trees to Analyze Electrical Contractors' Accidents

AU - Gholizadeh, Pouya

AU - Esmaeili, Behzad

PY - 2016

Y1 - 2016

N2 - When considering the impermanent environment of construction projects for electrical contractors, it is not always easy to discover the root cause of an accident. In addition, although previous studies have indicated that most accidents can be attributed to spatial and temporal interactions of multiple features, a limited number of studies have empirically explored these relationships. One data-mining technique that can be used to address these limitations and investigate the causal relationship among events is a classification and regression tree (CART). The current study applied CART analysis to explore the relationship between major characteristics of a construction project and/or work attributes, and injury outcomes. CART is an appropriate statistical analytical tool for analyzing data with non-linear relationships, a high-order of interactions, and a large number of missing values. To analyze a reliable and representative database of accidents involving electricians, content analysis was conducted on 320 accident reports obtained from the Occupational Safety and Health Administration database to identify attributes that led to accidents. All accidents occurred between 2009 and 2012. The outcomes of accidents (fatality versus no-fatality) were considered as dependent variables, and features such as project cost, project type, project end-use, and task attributes that cause accidents were considered as independent variables. The findings describe how specific characteristics of a project (e.g. cost) or a construction task (e.g. working near wiring) can be used to predict the probability of fatality for an electrician on a jobsite. The results of the study also can be used by safety managers to revise safety practices and training programs.

AB - When considering the impermanent environment of construction projects for electrical contractors, it is not always easy to discover the root cause of an accident. In addition, although previous studies have indicated that most accidents can be attributed to spatial and temporal interactions of multiple features, a limited number of studies have empirically explored these relationships. One data-mining technique that can be used to address these limitations and investigate the causal relationship among events is a classification and regression tree (CART). The current study applied CART analysis to explore the relationship between major characteristics of a construction project and/or work attributes, and injury outcomes. CART is an appropriate statistical analytical tool for analyzing data with non-linear relationships, a high-order of interactions, and a large number of missing values. To analyze a reliable and representative database of accidents involving electricians, content analysis was conducted on 320 accident reports obtained from the Occupational Safety and Health Administration database to identify attributes that led to accidents. All accidents occurred between 2009 and 2012. The outcomes of accidents (fatality versus no-fatality) were considered as dependent variables, and features such as project cost, project type, project end-use, and task attributes that cause accidents were considered as independent variables. The findings describe how specific characteristics of a project (e.g. cost) or a construction task (e.g. working near wiring) can be used to predict the probability of fatality for an electrician on a jobsite. The results of the study also can be used by safety managers to revise safety practices and training programs.

UR - http://www.scopus.com/inward/record.url?scp=84976428008&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=84976428008&partnerID=8YFLogxK

U2 - 10.1061/9780784479827.269

DO - 10.1061/9780784479827.269

M3 - Conference contribution

SP - 2699

EP - 2708

BT - Construction Research Congress 2016: Old and New Construction Technologies Converge in Historic San Juan - Proceedings of the 2016 Construction Research Congress, CRC 2016

PB - American Society of Civil Engineers (ASCE)

ER -