Stratified statistical analysis for effectiveness evaluation of frontline worker safety intervention: Case study of construction steel fabrication

•Stratified analysis is presented to evaluate effectiveness of safety intervention.•Measure for quality of hazard identification reports was found to be critical.•Handwriting of hazard analysis forms was effective for less experienced foremen.•Spending sufficient time discussing safety issues was ef...

Full description

Saved in:
Bibliographic Details
Published in:Safety science 2019-06, Vol.115, p.89-102
Main Authors: Chang, Jihun, Han, SangUk, AbouRizk, Simaan M., Kanerva, Jim
Format: Article
Language:English
Subjects:
Case studies
Confounding
Effect modification
Fabrication
Frontline safety
Handwriting
Inspection
Intervention
Intervention effectiveness evaluation
Occupational accidents
Occupational safety
Poisson density functions
Poisson regression
Regression analysis
Regression models
Safety
Safety management
Safety programs
Statistical analysis
Steel construction
Steel industry
Structural steels
Supervisors
Citations: Items that this one cites
Items that cite this one
Online Access:Get full text
Tags: Add Tag
No Tags, Be the first to tag this record!
Description
Summary:•Stratified analysis is presented to evaluate effectiveness of safety intervention.•Measure for quality of hazard identification reports was found to be critical.•Handwriting of hazard analysis forms was effective for less experienced foremen.•Spending sufficient time discussing safety issues was effective for junior groups. Frontline employees play a key role in implementing and maintaining onsite safety programs and techniques. Evaluations of the effectiveness of ongoing safety interventions, however, have relied on the subjective judgment of supervisors. Through a case study, this research aims to explore practical measures to assess the effectiveness of safety interventions (i.e., pre-task planning and worksite inspection) implemented at the frontline level. With the data collected from construction steel fabrication shops, a Poisson regression model integrated with stratification analysis is proposed to identify the relationships between performance measures and incident records; consider confounders and effect modifiers such as age, experience, and task types; and adjust the influence of the third variables on the regression analysis. The results show that content coverage rates, longhand description, and safety communication times are statistically related to incident reduction (p-value
ISSN:0925-7535
1879-1042
DOI:10.1016/j.ssci.2019.01.030