Hi,
i have one question about the hardware requirements. Recently one of my friends warned me of using LCD monitor and laptop. It is said that LCD monitors and laptops will cause considerable errors to the measurement of reaction time because of the problem with the slow refresh rate. So i wonder whether Inquisit also has the same requirement. To be honest, using CRT is now not always acessible in my school, because my school panel nearly changed all the CRT Monitors to LCD last year.
LCDs should be okay as long as your research doesn't require very short and exact stimulus onset and offset times (e.g. subliminal priming). This thread has some more info: http://www.millisecond.com/community/forums/p/1222/3649.aspx#3649.
~Dave
"To understand recursion, you must first understand recursion." - Unknown Zen Master
Thanks, Dave. I also have antoher question that was also warned by my friend. It is said that many services and programs in Windows will also compete the resources with the Inquisit and thus affect the presentation of the stimulis, causing inconsistency of the measurement of RT. So i am really depressed by his remarks. It seems nearly impossible to shut up every running services and programs. The systemic errors seem unavoidable. The only thing i can do is just try to optimalize the system while running Inquisit through some ways like msconfig. Now i am running the implicit priming experiment. I don't know the systemic errors will really lead to the fake experimental results.
Don't freak out because of your colleague's remarks. Things aren't that bad, really. (1) Inquisit usually does a really good job to effectively monopolize the hardware at runtime, even on systems that aren't tuned optimally. (2) While measurement error is unavoidable -- this applies to any method, computer-based or not -- you'd only run into problems if said error varied systematically with experimental conditions (thus creating a confound). In within-subjects-designs that usually won't happen, error will be distributed randomly and equally across conditions and -- if necessary -- can be further reduced by increasing the number of measurements (trials). (3) If you're still worried, get your institution to buy a Blackbox toolkit or have your tech shop build something similar to specifically check a given system's timing capabilites.
Calm down, all is well.
Thanks very much, Dave. I think you are right. Now i am going on.