What is Reaction Time Data
Reaction Times (RTs) are a way to measure someone's response to a stimulus. Examples of tasks measuring RTs are the Color-Shape Switch task or the Flanker task. The duration of the response (also called response latency) is then often taken to be an index of an executive function (e.g., in the case of the Flanker task, inhibition).
Collecting RTs of older adults
When collecting RTs of older adults, there are several practical considerations to keep in mind. These include:
Include practice blocks. RTs are typically collected using computers. Younger adults (especially those studying at universities where they are asked to participate in behavioral tasks, anyway) are typically very well-versed in doing these tasks, however older adults display more variation with regards to the degree of experience using computers. Therefore, it is good practice to have one or more practice blocks (i.e., blocks where the participant practices the task without their responses counting towards the percentage of correct trials). It is not uncommon to have a requirement that a participant needs to have an accuracy score of 80% or higher before they can do the actual task. When one foregoes these practice trials, the task itself may be too difficult for a participant.
Increase the maximum allowed response times. Sometimes, a researcher may want to implement a maximum allowed RT for tasks (e.g. if no response is given after 5000ms, abort the trial). Compared to younger adults, the RTs of older adults are generally somewhat slower, but there is also much more variability in the duration of these RTs. Therefore, it is good practice to pilot test your particular task, to see what the maximum allowed RT should be set to. As a rule of thumb, one could consider increasing maximum allowed RTs by 30% if one's participants are aged between 65-76, and by 40% if they are 75 and older. This is based on data from Hultsch, MacDonald, and Dixon (2002), who studied RT variability on a number of tasks. We checked how long RTs were for those who took 2 standard deviations longer than the average participant in all groups. In theory, approximately 97.5% of responses on a task should be quicker than this. We then calculated how much longer these RTs were in the older group compared to the younger group.
Always pilot new tasks. Be aware, though, that even between tasks there is a large degree of variability. The increases in Hultsch and colleagues’ study ranged from 5% to 41% depending on the task and age group. Therefore, it is good practice to always pilot one's tests, or to forego a maximum allowed response time duration.
References
Further reading