Cumulative distribution functions: An alternative approach to examine the triggering of prepared motor actions in the StartReact effect

There has been much debate concerning whether startling sensory stimuli can activate a fast‐neural pathway for movement triggering (StartReact) which is different from that of voluntary movements. Activity in sternocleidomastoid (SCM) electromyogram is suggested to indicate activation of this pathway. We evaluated whether SCM activity can accurately identify trials which may differ in their neurophysiological triggering and assessed the use of cumulative distribution functions (CDFs) of reaction time (RT) data to identify trials with the shortest RTs for analysis. Using recent data sets from the StartReact literature, we examined the relationship between RT and SCM activity. We categorised data into short/longer RT bins using CDFs and used linear mixed‐effects models to compare potential conclusions that can be drawn when categorising data on the basis of RT versus on the basis of SCM activity. The capacity of SCM to predict RT is task‐specific, making it an unreliable indicator of distinct neurophysiological mechanisms. Classification of trials using CDFs is capable of capturing potential task‐ or muscle‐related differences in triggering whilst avoiding the pitfalls of the traditional SCM activity‐based classification method. We conclude that SCM activity is not always evident on trials that show the early triggering of movements seen in the StartReact phenomenon. We further propose that a more comprehensive analysis of data may be achieved through the inclusion of CDF analyses. These findings have implications for future research investigating movement triggering as well as for potential therapeutic applications of StartReact. Our re‐analysis of recent StartReact literature shows there are task‐specific responses to startling stimuli which can influence the manifestation of, and the ability to detect StartReact on the basis of sternocleidomastoid (SCM) activity. When trials are categorised and analysed using cumulative distribution functions of reaction time data however, a more comprehensive examination of movement triggering may be obtained whilst avoiding the pitfalls of the traditional SCM activity‐based method.