This study was conducted to measure the reliability of fatigue measures in an intermittent overhead work task. Fatigue measures included several EMG based parameters and subjective discomfort ratings through use of the Borg CR-10 scale. This study was part of a larger existing study that simulates overhead work in an automobile manufacturing plant. Ten participants used a drill tool to perform an overhead tapping task for one hour at a height relative to individual anthropometry.

Reliability indexes, including Intraclass Correlation Coefficients, Standard Errors of Measurement, and Coefficients of Variation were determined for each fatigue measure for each of three shoulder muscles (anterior deltoid, middle deltoid, and trapezius). High reliability implies repeatable results, and precise and credible methods. Conversely, measurement error and subject variability can lead to low reliability of measures.

The results indicated that ratings of perceived discomfort (RPD) parameters (slope and final rating) showed relatively high reliability. Intercepts for mean power frequency (MnPF), median power frequency (MdPF), and root means square (RMS) also showed very high reliability. Actual slopes for MnPF, MdPF, and RMS showed low reliability overall, and normalizing slopes did not necessarily improve reliability. Taking the absolute value of slopes led to a noticeable increase in reliability. RPD slope did not correlate with any of the EMG slopes.

The high reliability of RPD parameters allows for its inexpensive application to the industrial setting for similar overhead tasks. The reliability of EMG intercepts implies consistent methods; however the reliability of overall EMG trends is suspect if the slope is not reliable. Some EMG slope parameters show promise; however, more research is needed to determine if these parameters are reliable for complex tasks.