Publisher
University of Tennessee at Chattanooga
Place of Publication
Chattanooga (Tenn.)
Abstract
Workplace safety is not a new field of study, but the application of data analytics as a predictive tool is relatively new to the area. Beyond the obvious harm caused to workers, fatal and nonfatal injuries accrue a significant financial burden nationwide. Previous safety measures have focused largely on reactive tools. While these have done a significant job reducing nonfatal workplace injuries, they have done little to assuage fatal ones. Predictive analysis may be the next step in reducing workplace injury. By mining and processing safety-related data, predictive analysis can inform employers of high risk situations before they happen. By taking this preventative approach, employers have the chance to stop workplace incidents before their employees suffer harm. While these analyses have generally used national-level data sources and shown great success at reducing injury in some industries, such as construction, there has been little work done either in the manufacturing sector or at the company level. Using company-level data collected from a chemical manufacturing company, this team will create a predictive model for workplace incidents. Data collected from the 2017 and 2018 years will be used to train and validate a prediction model. That model will then be tested in 2019 data to determine its predictive efficiency.
Date
October 2020
Subject
Industrial and organizational psychology
Document Type
posters
Language
English
Rights
http://rightsstatements.org/vocab/InC/1.0/
License
http://creativecommons.org/licenses/by/4.0/
Included in
Generating a predictive model for injury rates for a chemical manufacturing company
Workplace safety is not a new field of study, but the application of data analytics as a predictive tool is relatively new to the area. Beyond the obvious harm caused to workers, fatal and nonfatal injuries accrue a significant financial burden nationwide. Previous safety measures have focused largely on reactive tools. While these have done a significant job reducing nonfatal workplace injuries, they have done little to assuage fatal ones. Predictive analysis may be the next step in reducing workplace injury. By mining and processing safety-related data, predictive analysis can inform employers of high risk situations before they happen. By taking this preventative approach, employers have the chance to stop workplace incidents before their employees suffer harm. While these analyses have generally used national-level data sources and shown great success at reducing injury in some industries, such as construction, there has been little work done either in the manufacturing sector or at the company level. Using company-level data collected from a chemical manufacturing company, this team will create a predictive model for workplace incidents. Data collected from the 2017 and 2018 years will be used to train and validate a prediction model. That model will then be tested in 2019 data to determine its predictive efficiency.
Department
University of Tennessee at Chattanooga. Dept. of Psychology