Can a body’s reaction time be unner 100 milliseconds?

Woods and colleagues uised unpredictable stimulus timin, practice trials and a response window tae define whit events countit. Thir choices are pairt o measurin a response tae the signal, athin ower-early or mistimed response. The paper anentit its results for equipment delays an aw, showin why a software timestamp canna jist be equatit wi a bodily event. Woods et al., 2015

Reaction and prediction are different behaviours

Gin the neist event is predictable, ye can get yer muivement redd up afore it happens. Sae a response near the expected time can be gey close tae the event athoot stertin efter ye detect it. That isnae the same quaistion as simple reaction time tae a new, unpredictable cue. Task instructions and stimulus schedules are needfu tae mak sense o an unco sma interval. Woods et al., 2015 Woods et al., 2015

A choice response adds pickin an action. The classic leeteratur on response alternatives shaw that task structure affects latency; a minimum claimed for a simple sicht gemme shouldnae be applied tae a gemme that needs decisions atween several actions. Hick, 1952

Why ae spectacular score is puir evidence

The fastest observation is an ootlier at the tail. Repeatin a task mony times gies mair chances for an unusual observation, an anticipatory response or a measurement mistak. Research on reaction-time ootliers explains why researchers set oot hou extreme responses are dealt wi, insteid o treatin ilka displayed nummer as equally informatin. Ratcliff, 1993

Device precision disnae validate a bodily response

Browsers can record fine-grained timestamps, but timestamp resolution isnae the same as accuracy aboot the physical start o a stimulus and input. Platform comparisons document baith offsets and variation. A score wi several decimal places isnae necessarily a mair accurate estimate o yer response tae the sicht cue. Anwyl-Irvine et al., 2020 Bridges et al., 2020

Nor is it valid tae uise an equipment correction fae a different set-up tae rescue or disprove ae particular result. Calibration applies tae the apparatus that wis measurt. Athoot it, the direction and size o aa the timin errors in yer ain set-up are uncertain. Woods et al., 2015 Anwyl-Irvine et al., 2020

Whit tae dae wi an unco fast result

Keep the hale run and its rules in mind: wis there ony ower-early responses, were the intervals predictable, and wis the nummer a minimum or an average? Compare the same statistic ower a fixed nummer o attempts. Dinna increase the nummer o tries and read the new minimum as gin the protocol hadnae changed. Research on trial counts shaw that measurement properties depend on the data gathered and the quaistion speirt. Miller, 2023 Ratcliff, 1993

Ae score on its lane tells ye little aboot a steady personal trait. Reliability maun be shown, especially whan a task is uised tae rank fowk. The richt conclusion is that an unusual event happened in this gemme; confirm that an exceptional human response happened needs a controlled measurement. Hedge et al., 2017

There is nae grund here for cryin somebody dishonest, diagnosin a problem or settin a universal hard floor. The important difference is atween a displayed interval and evidence that a new sensory event actually set aff the response. Woods et al., 2015 Bridges et al., 2020