“…There is a prior study [48] that found a higher throughput compared to these prior experiments when using gaze; however, it can be challenging to make comparisons across studies and this work did not have any comparison to common modalities, which we utilize in our study. Unlike what we provide in our work, many prior studies, including these three discussed here [18,39,48] do not measure or report eye tracking calibration statistics in their reports; at best they report the manufacturer's stated eye tracking quality, but even that is rare. In cases where the eye tracker performance is poor, but not reported on a participant level, it is impossible to know whether poor interaction efficiency is due in earnest to the experimental manipulation or simply the result of tracking errors.…”