Reliability of force production based on the sense of effort at different force levels

Force production under sense of effort conditions is essential for motor control and clinical assessment; however, its reliability across different force levels remains unclear. In addition, the lack of standardized measurement methods has limited the accumulation of consistent findings. Therefore, establishing reliable measurement conditions is necessary for both research and practical applications. This study examined the test-retest reliability of grip-force production based on the sense of effort at three force levels: 25%, 50%, and 75% of maximum voluntary contraction (MVC). Thirty-three healthy young men performed grip-force tasks across two sessions separated by 1 week without external feedback. Relative reliability was assessed using intraclass correlation coefficients (ICCs), and absolute reliability was evaluated using the standard error of measurement (SEM) and the 95% minimal detectable change (MDC). The highest reliability was observed at 25% MVC across both outcome measures. At higher force levels, absolute reliability varied depending on the error index, with lower reliability at 50% MVC for absolute error and at 75% MVC for signed error. These findings indicate that reliability depends on the aspect of error being measured and that increasing force levels alter the structure of error rather than simply increasing its magnitude. Furthermore, systematic bias at low force levels suggests that high reliability does not necessarily reflect close correspondence to the prescribed force level. Overall, low-force conditions provide more reliable assessments, highlighting the importance of considering both the magnitude and direction of error.