Theoretical and empirical efficiency of sampling strategies for estimating upper arm elevation

Area of Science:

• Occupational health
• Ergonomics
• Industrial hygiene

Background:

• Accurate exposure assessment is vital in occupational health.
• Traditional sampling efficiency models often rely on simplifying assumptions.
• Understanding data structure, like autocorrelation, is key to improving sampling strategies.

Purpose of the Study:

• To evaluate the statistical efficiency of various sampling strategies for upper arm elevation data.
• To compare theoretical predictions of sampling efficiency with empirical results using bootstrap simulations.
• To identify optimal sample sizes and allocation methods within and across measurement days.

Main Methods:

• Investigated 65 sampling strategies using minute-by-minute upper arm elevation data from house painters, car mechanics, and machinists.
• Collected data over four working days, varying total sample times and block lengths.
• Employed both standard theoretical models and nonparametric bootstrapping for efficiency assessment.

Main Results:

• Found violations of independence and homoscedasticity assumptions in theoretical models due to within-day autocorrelation.
• Wider distribution of measurements across time and days, particularly using 1-min blocks at fixed intervals over 4 days, enhanced sampling efficiency.
• Theoretical estimates overestimated efficiency, especially for larger block sizes and shorter total sample times.

Conclusions:

• Sampling efficiency for autocorrelated exposure data improves by widely distributing measurements across time and days, preferably with fixed intervals.
• Collecting more data than theoretically suggested is recommended for achieving desired precision.
• Sampling larger proportions of the working day mitigates autocorrelation effects and improves exposure estimates.