[Weighting methodology for the national EVREST survey data].

Area of Science:

• Occupational health
• Survey methodology
• Statistical analysis

Background:

• The Evolution and Relations in Health at Work (EVREST) survey collects national data on workplace health.
• Accurate weighting is crucial for generalizing survey findings to the broader population.
• Reference data availability can impact survey estimates.

Purpose of the Study:

• To detail the weighting methodology for the EVREST survey.
• To assess the impact of weighting on survey estimates.
• To evaluate the effect of a two-year data lag on reference data.

Main Methods:

• The study utilized data from 26,227 employees surveyed in 2013-2014.
• A two-step weighting process involved accounting for participation probability and calibrating sample margins.
• Reference data from 2012 and 2014 Declarations of Social Data (DADS) were used.

Main Results:

• Weighting reduced estimate differences to within +/- 2.0% for 90% of variables (2014 DADS).
• Differences were within +/- 2.0% for 83% of variables using 2012 DADS.
• The two-year gap in reference data had a minimal impact on estimates.

Conclusions:

• A robust weighting methodology was established for the EVREST survey.
• This methodology ensures the representativeness of survey results.
• The findings support the reliability of EVREST data despite potential reference data lags.