User`s guide

E-Prime User’s Guide
Chapter 3: Critical Timing
Page 87
go, and unlikely to be touched by the operating system. The code for monitoring when to display
the stimulus is in a tight loop to keep execution short and keep the code in memory. If the
operating system takes control and steals time that is less than the PreRelease time, the tight
code segment of putting up the display will not be affected. Although the operating system can
still stop the tight code loop at presentation time, the probability of such an event is greatly
minimized by making the final critical actions brief. In general, a PreRelease value of 100-200ms
is sufficient for most presentation sequences.
While the PreRelease mechanism is a powerful and useful tool, there are a few caveats when
using PreRelease of which you should be aware. The first caveat is if you are using PreRelease
on an object that immediately precedes a FeedbackDisplay object, there is a small window of
time during which the subject could respond and incorrect feedback could be presented. This
may occur because the FeedbackDisplay object will read the response data at the beginning of
the PreRelease time of the previous object. For example, assume a probe stimulus is presented,
accepting response input for 1000ms, and specifying a PreRelease time of 100ms. The
FeedbackDisplay object would begin executing and preparing the display 900ms after the onset
of the object presenting the probe. At that time, the FeedbackDisplay object would check the
accuracy and response time of the last input, and choose a display message to show (e.g.,
Correct, Incorrect, No Response, etc). If the subject happened to respond any time between
901ms and 1000ms from the onset of the probe, the FeedbackDisplay would likely have already
prepared the feedback prior to the response being accepted. Thus, the resulting feedback
presented to the subject would indicate “No Response”, when in fact a response was received. It
is important to note that in this case, the response data would still be time stamped, scored and
logged correctly at the end of the trial, but because of the release of time to the FeedbackDisplay,
the display of the feedback to the subject could be incorrect. To avoid unwanted interactions with
the presentation of feedback to the subject, it is recommended that you set the PreRelease to
0ms on any object immediately preceding FeedbackDisplay objects. This will delay the
presentation of the feedback because setup time is required that is not absorbed by the previous
object. Typically, the time required for setup is less than 300ms, and often feedback is not
considered a time-critical event. Alternatively you could also adjust the Time Limit on the
response input so that the PreRelease period and the response period do not overlap.
Although it was stated in the previous example that the logging of the data would not be affected,
there are scenarios under which there can be unwanted interactions with PreRelease and data
logging. Specifically, using PreRelease can alter data logging if an active response ranges
through the end of a Procedure. If the current display has a response interval longer than the
duration of the trial (i.e., the Time Limit on the response mask is longer than the total duration of
the rest of the trial), or if the response data must be used in the script that may execute during the
PreRelease period, no PreRelease should be specified.
In the example below (see Figure 8), a 1 second duration Probe display has an extended
response period of 2 seconds (i.e., to allow response during the Probe+Mask period). In the
upper example (i.e., Log Post Response), the FeedbackDisplay is set up after a mask is
displayed and after the response interval has concluded (see “$” in Figure 8). In the Log Before
Response grouping (bottom of Figure 8), the Probe is displayed for 1 second before releasing to
the Feedback object. The Feedback object prepared the feedback and logged the data while the
response was still pending. If the response came in at 1.9 seconds post the onset of the Probe
display, the feedback would be inaccurate, and the response would not be logged. When using
extended input processing, the user must take care that the response processing is completed
before feedback processing and data logging for the trial begins.