Work

Beyond aggregate injury declines: Illustrating a minimum reporting profile with paired employee data

Year: 2026
Type: article
Abstract: Introduction: Aggregate injury metrics are incomplete and can support different interpretations. Even when an overall decline is accurate, it does not show which employees changed or whether the account changes when the same responses are summarized differently. This short communication presents a minimum reporting profile and demonstrates it with a worked example. Method: The example uses a historical panel of paired self-reports from 71 manufacturing employees surveyed twice about injuries during the preceding four weeks. Employees rated 10 injury types using ordered frequency categories. I summarized the same responses as any reporting, the number of types reported, and a score combining breadth with ordinal frequency. For the latter two indicators, I separated decreases from increases and distinguished employees who stopped reporting injuries from those who continued reporting at lower levels. Results: In this panel, any reporting fell from 73.2% to 46.5%. For the frequency-category score, transitions to no reporting accounted for 47.7% of all decreases, while lower scores among continuing reporters accounted for 52.3%. Small increases partly offset these decreases. For the injury-type count, transitions to no reporting accounted for 64.3% of all decreases. Practical Applications: The worked example shows that evaluations using paired injury reports should report the aggregate change, movements between no and some reporting, decreases and increases among employees reporting at both times, and sensitivity to indicator choice. The profile clarifies how an aggregate result arose. It does not convert self-reports into verified incident counts, explain why change occurred, or support generalization beyond this panel.
Author: Nick Turner
Language: English
Subfield: Safety Research
Open Access status: hybrid
Estimated APC paid: $3,220
OpenAlex ID: W7233605325
FWCI:
Cites: 14
Cited by:
Related to: