How long is the average keystroke delay?

Question

I am currently helping someone experiment with reaction times. For this experiment, the response time on the keyboard is measured. For this experiment, it may be important to know how many errors can be entered due to the delay between keystrokes and processing in the software.

Here are some factors that I have discovered using Google already:

• The USB bus is rated for 125 Hz minimum and 1000 Hz maximum (depending on the settings, see this link ).
• On Windows, additional keyboard buffers can be added that can delay keystrokes further, but I do not know about their logic.

Unfortunately, it is not possible to control the logic of a low level experiment. The experiment is written in E-Prime , a software that is often used for this type of experiment. However, the company offering E-Prime also offers optional equipment that they advertise for accurate reaction times. Therefore, they seem to be aware of this effect (but do not report how great it is).

Unfortunately, you need to use a standard keyboard, so I need to provide ways to reduce latency.

Answer 1

any latency from keystrokes can be attributed to the debounce routine (I usually use 30 ms for security), and not to the processing algorithms themselves (unless you evaluate only the first keystroke).

Answer 2

If you are using an experiment in which milliseconds of time are important, you can use http://www.blackboxtoolkit.com/ to find the sources of errors.

Your needs also depend on the nature of your research. I conducted RT experiments in Eprime using the keyboard. Since any error should be consistent on average among the participants, for some designs this is not a big problem. If you need to synchronize data, but with something else (for example, eye tracking or EEG) or want to draw conclusions about RT, where a certain value is important, then the serial release of E-Primes (or another brand, although I had compatibility problems into the past with other brands and eprime) is a must.