Reliability of reaction time measurements in brain-damaged patients

The higher sensitivity of chronometric analysis leads to an increase of their use in the evaluation of brain-damaged patients. However, patients have usually higher reaction time (RT) variability, suggesting a possible decrease in the reliability of RT measurement. This reliability has never been evaluated in patients. This study assessed the reliability of RT estimates in 2 populations of brain-damaged patients and controls, using the Kendall coefficient of concordance W. Concordance coefficients were computed for simple detection and choice tests. W values were good and significant. This study suggests that RT measured in simple detection and choice tasks represent a reliable measurement and supports the idea that the higher sensitivity associated with the use of chronometric analysis is not obtained at the cost of a lower reliability. Moreover, these results provide information for the elaboration of RT tests.