A second crude way of measuring response times is to calculate the time differences of when materials are accessed on the Web server. Scientific LogAnalyzer (Reips & Stieger, 2004; see Using Web-Based Methods: An Example, this chapter) includes a routine to calculate these times from servers' log files.
So, is there any way to accurately measure reaction times via the Internet? There is: Eichstaedt (2001) developed a Java-based method for very accurate response time measurements. A clever combination of applets ensures continuous synchronization and calibration of timing between server and client, which minimizes timing inaccuracies produced by the Internet.
Was this article helpful?