keyboard presses delays

[David McFarlane]

Flavia,

>I am running a sentence processing experiment and I have to use the
>keyboard as input response device.
>
>I read on the manual that the error delay is estimated around 5 ms, is
>this correct?

I assume you are referring to Appendix A of the User's Guide, which 
reports keyboard latencies about 6.5 ms with a std dev of about 0.8 
ms.  Many years ago similar tests were published in Behavior Research 
Methods, and we ran our own and got latencies up to 15 ms on some 
systems, back then results were pretty variable, don't know how good 
they are now.  As I recall, the MEL folks (now PST) wrote a white 
paper about this explaining that you could make up the variance by 
adding a few extra trials.  That's OK unless you really need ms level 
accuracy for single trials (e.g., to control timing of subsequent stimuli).

>And also that, in order to minimize random delays, programs such as
>Antivirus and MS assistance should be disabled. Are there other
>programs that should be uninstalled or disabled while the experiment
>is running (in order to minimize random and routine delays)?

You can turn off the screen saver.  Also, I have heard that 
unplugging the network cable can help, though I have not tested that 
yet.  Make sure nothing is running in Task Scheduler.  In general, 
just disable *all* startup programs unless needed to run the 
experiment.  I like to use the startup manager in Spybot Search & 
Destroy, but other may like other startup managers.  Better yet, 
completely uninstall any programs not needed for running 
experiments.  (And if you get really serious you can also run 
services.msc to disable unnecessary services, but I can't tell you 
offhand which ones to disable there.)

And run the RefreshClockTest as in Appendix A to test your setup for 
clock errors.

-- David McFarlane, Professional Faultfinder


--~--~---------~--~----~------------~-------~--~----~
You received this message because you are subscribed to the Google Groups "E-Prime" group.
To post to this group, send email to e-prime at googlegroups.com
To unsubscribe from this group, send email to e-prime-unsubscribe at googlegroups.com
For more options, visit this group at http://groups.google.com/group/e-prime?hl=en
-~----------~----~----~----~------~----~------~--~---