Publisher
University of Tennessee at Chattanooga
Place of Publication
Chattanooga (Tenn.)
Abstract
Organizations have been able to make decisions that increase success and solve practical problems using data analytics and evidence-based management (Barends et. al, 2014). Analysis of injury-related data can reveal environmental workplace weaknesses and safety policy areas for improvement (Illinois Department of Public Health, 2023; Centers for Disease Control and Prevention [CDC], 2015). However, proper variable development, harmonized measurement and data collection are a critical first step to determine the depth of analyses available for analytics. The data analytics readiness tool (DART) was developed to assist organizations in better understanding their current measurement capabilities, so that more advanced analytic techniques can be utilized to maximize organizational safety (Ezerins et al., 2022). Hinson et al., (2021) made use of similar data inventories and data analytic strategies, therefore, our study aims to replicate and build upon these methodologies. We are currently implementing the DART system within a large oil refinery in the southern United States to assess the potential predictive validity between variables of interest and incident probabilities. The DART will be used to sort variables into three broad categories as follows: Culture, prevention, and production. Culture is a measure of the organization's safety culture, prevention includes behaviors and tasks intended to reduce injuries (safety audits, inspections, hazard identifications, and safety observations), and production includes total hours worked, number of employees on site, amount of output, and even day of the week. The results of the DART will provide a roadmap for a) what data to retrieve that has high enough quality for initial analysis, leading to b) initial hypotheses, data cleaning and analysis, and c) direction for improvements to improve the quality of other variables of interest. Usage of the DART allows data to be obtained, organized and aggregated in a timely manner. Additionally, this tool can help identify the required variables to strengthen the analysis. This is especially critical, as it can also highlight variables that organizations may not be collecting information on to assess what may be causing the outcome. Through use of the DART in this proposed study, future data analyses can be completed successfully as certain criteria must have been met to ensure that there is adequate coverage of variables in order to predict the validity between the aforementioned variables of interest and incident probabilities (Compagnone & Ludwig).
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
Assessing Data Analytics Readiness as a first step to Predict Incident Probability: A replication
Organizations have been able to make decisions that increase success and solve practical problems using data analytics and evidence-based management (Barends et. al, 2014). Analysis of injury-related data can reveal environmental workplace weaknesses and safety policy areas for improvement (Illinois Department of Public Health, 2023; Centers for Disease Control and Prevention [CDC], 2015). However, proper variable development, harmonized measurement and data collection are a critical first step to determine the depth of analyses available for analytics. The data analytics readiness tool (DART) was developed to assist organizations in better understanding their current measurement capabilities, so that more advanced analytic techniques can be utilized to maximize organizational safety (Ezerins et al., 2022). Hinson et al., (2021) made use of similar data inventories and data analytic strategies, therefore, our study aims to replicate and build upon these methodologies. We are currently implementing the DART system within a large oil refinery in the southern United States to assess the potential predictive validity between variables of interest and incident probabilities. The DART will be used to sort variables into three broad categories as follows: Culture, prevention, and production. Culture is a measure of the organization's safety culture, prevention includes behaviors and tasks intended to reduce injuries (safety audits, inspections, hazard identifications, and safety observations), and production includes total hours worked, number of employees on site, amount of output, and even day of the week. The results of the DART will provide a roadmap for a) what data to retrieve that has high enough quality for initial analysis, leading to b) initial hypotheses, data cleaning and analysis, and c) direction for improvements to improve the quality of other variables of interest. Usage of the DART allows data to be obtained, organized and aggregated in a timely manner. Additionally, this tool can help identify the required variables to strengthen the analysis. This is especially critical, as it can also highlight variables that organizations may not be collecting information on to assess what may be causing the outcome. Through use of the DART in this proposed study, future data analyses can be completed successfully as certain criteria must have been met to ensure that there is adequate coverage of variables in order to predict the validity between the aforementioned variables of interest and incident probabilities (Compagnone & Ludwig).
Department
University of Tennessee at Chattanooga. Dept. of Psychology