User`s guide

E-Prime User’s Guide
Chapter 3: Critical Timing
Page 108
TimingParadigm2 Time Audit Data (ms)
Fixation Probe Mask
Trial OnsetDelay OnsetTime OnsetDelay OnsetTime OnsetDelay OnsetTime Fix-Probe Probe-Mask
1 12 4975 10 5474 10 5574 499 100
2 12 7768 10 8267 9 8366 499 99
3 17 11342 9 11840 10 11940 498 100
4 13 14317 10 14816 9 14915 499 99
5 17 17425 10 17924 10 18024 499 100
6 9 20334 10 20833 9 20932 499 99
7 17 23259 10 23758 10 23858 499 100
8 16 26434 10 26933 10 27033 499 100
9 13 29659 10 30158 9 30257 499 99
10 9 32634 10 33133 10 33233 499 100
11 8 35643 9 36141 10 36241 498 100
12 4 38618 10 39117 9 39216 499 99
13 6 41693 10 42192 10 42292 499 100
14 7 44735 10 45234 9 45333 499 99
15 6 47993 9 48491 10 48591 498 100
16 2 50968 10 51467 10 51567 499 100
Mean 10.5 9.81 9.63 498.81 99.63
S.D. 4.69 0.39 0.48 0.39 0.48
Note, the Fixation.OnsetDelay is random, whereas the Mask.OnsetDelay and Probe.OnsetDelay
are consistently near 10ms. This is due to the first display starting after the subject pushes the
spacebar to start the presentation. The time at which the subject presses the spacebar is random
with respect to the refresh rate, and that is reflected in the random delay of the fixation stimulus.
However, at the time of the Fixation presentation, the computer has synchronized with the vertical
blanking signal. We specified that the Duration of the fixation would end 10ms before the next
refresh interval (expected refresh at 499ms, set Duration to 489ms). Therefore, the expected
OnsetDelay of the Probe would be 10ms, and the mean duration observed was 9.81ms. The Fix-
Probe time (second to last column) is the duration that the Fixation was actually presented to the
subject. The display duration of 498.81ms is in close agreement with the 498.6ms expected
based on the refresh time calculations (e.g., 16.62ms/cycle * 30 cycles).
This experiment employed masking of the probe stimulus, and it is critical to check the timing of
the period between the Probe presentation and the Mask. Based on the refresh calculations, we
expected the refresh time to be 99.7ms. The observed time for Probe.OnsetTime to
Mask.OnsetTime was 99.63ms, again in close agreement to what was expected. Note also that
the Mask.OnsetDelay was 9.63ms; close to the 10ms expected.
Using the Time Audit features of E-Prime, users can verify and report the true timing of the
time critical segments of an experiment. In this case, it can be reported that the Fixation was
presented for 498.8ms (standard deviation 0.39ms), and the Probe for 99.6ms (SD 0.48ms). It is
important to report the actual times rather than the programmed times. Of course, if intended
times mismatch the actual times, check the specifications and the critical settings carefully.
Typical things that produce bad timing are miscalculation of the refresh rate, incorrect settings for
timing mode, insufficient PreRelease time, and having a computer that has other processes
running (see miss tick tests Appendix A).