The keyboards are (still) all right in response time experiments.

Abstract

Response times (RTs) are a ubiquitous variable for assessing cognitive and motor processes. However, variability introduced by keyboards, especially in online experiments, has raised concerns among behavioral researchers. Here, we evaluate the impact of keyboard delays on RT measurements using linear mixed-effects models and grouped data t-tests through a series of simulations. The results showed that the impact of keyboard delays on statistical power is minimal in most cases. Keyboard-induced variability does not inflate type I error rates and has a negligible impact on power, except in rare scenarios of RT distribution shifts or in studies focused on individual differences with low signal-to-noise ratios. Thus, commercially available keyboards remain suitable for most RT experiments, including those conducted online.