e-prime task runs for variable durations

[David McFarlane]

Jim,

It's your program.  I have used Cumulative timing mode on very 
complex tasks that span many minutes in an fMRI scanner, with exactly 
0 ms error on the final .TargetOnsetTime, which is the operative 
measure here.  And if you understand the simple arithmetic underlying 
Cumulative mode (much simpler than the PST documentation makes it 
out), then you will know that it *has* to be exact, anything else 
would violate the laws of arithmetic!

So I say this again Jim -- Stop working on your actual experiment 
programs and start working on *lots* of *small* demo programs until 
you work out the fundamental concepts of Cumulative vs. Event, 
PreRelease, Terminate vs (none), and Duration vs. Time Limit inside & 
out, forward & backwards, sideways & up & down!  And I do not mean a 
"paired [sic] down" version of your experiment program, that is not 
fair.  I mean start with a completely blank slate.  Add one List, one 
Procedure, and one TextDisplay.  Set it to log *everything*.  Run 
that with every combination of parameters that you can think of 
(preferably in a systematic way) and examine the changing 
relationships between *all* the logged timing measures in the 
resulting .edat files until you think you know how everything 
works.  Then try to think up more tests that will prove you wrong, 
and run those tests.  (Need I say that I am just asking you to apply 
the same scientific process to your research tools as you do to your 
research itself?)  Then add one or two more TextDisplay objects to 
the Procedure and continue your battery of tests.  Etc., etc.  And 
after all that, still do *not* go back to your existing broken 
program.  Instead, start with another blank slate and rebuild your 
experiment program from scratch, this time doing a test run after 
*each* and *every* change.  No, I am *not* kidding.  Do it this way 
and you will finally creep up on the factor that breaks your existing 
program, and only then will you be close to the understanding you 
need.  For the final stage (and I have neglected to stress this 
earlier) you must also study the code automatically generated by 
E-Studio.  That will reveal much of the inner workings.  Only after 
you have done the exercises described above *combined with* studying 
and understanding the code generated by E-Studio will you be in a 
position to tame your programs as desired.

Over thirty years in this business has taught me the hard way, again 
and again, that I will not find the solution to even the "simplest" 
problem until I do the hard work of tearing each system down to its 
roots and understanding it to the very core.  Not what we wish, but 
there it is.  The sooner you learn that lesson the sooner you will 
progress in this endeavor.

-- David McFarlane, Professional Faultfinder
"You got to test that piece of software, You got to test it for yourself,
No one else can test it for you, You got to test it for yourself."
(Apologies to the Fairfield Four)


>I'm developing a basic CPT task to use in conjunction with fMRI. The
>task displays a letter for 250ms, a blank screen for 1 second, then
>scores responses using inline code. Rinse and repeat.
>
>One thing that I've noticed with e-prime is that the duration of a
>task changes across trials, even when all objects are set to
>cumulative (and objects are set to run for a specified length of time
>as opposed to terminating with user input). For example, I ran a task
>this morning 4 times, 2 with 100ms prerelease and 2 without
>prerelease. The difference between the longest and shortest trials
>accross task versions was 5 seconds; the difference between the
>average duration for the 2 task versions was 2 seconds. This is
>problematic for me since most of the time I am using e-prime in
>conjunction with fMRI or physio data collection. I can't get a handle
>on how long I should set my scanner program for if the e-prime program
>has a variable duration.
>
>Is this variable duration to be expected or something problematic with
>my task? If it is to be expected, how do folks synch scanner/physio
>programs with e-prime?
>
>Jim

-- 
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.