Eric,
Depends on what you mean by "parallel tasks". Some folks apply that
term when subjects perform trials of two different tasks in random
succession, especially when one of the tasks requires some
memorization or vigilance (e.g., n-back trials interrupted with
addition trials, where one must keep in mind the prior n n-back
stimuli while still performing the addition trials). This usage
contrasts with concurrent asynchronous performance of two tasks
(e.g., typing a short essay while images flash on the border of the
screen, and the subject must press a foot pedal whenever a particular
image appears).
For the first kind of dual task in E-Prime you would want a TrialList
that uses levels (rows) with two different Procedures. The trick
then is getting the randomization to work as you like. If you are
not too fussy then you can just have the E-Prime List randomize all
the A-Procedure & B-Procedure trials for you, but if you need to add
constraints to the randomization then things get considerably tricky
(see, e.g.
http://www.pstnet.com/forum/Topic2186-5-1.aspx or
http://support.pstnet.com/forum/Topic3166-5-1.aspx ).
For the second kind of dual task, I once did the example above all in
C under DOS. I did that by contructing a randomized "time schedule"
of all timed events before the start of the session, and programming
an event loop which handled every key & foot press and looked at the
time schedule to update the screen images as specified. In principle
you could do much the same thing in E-Prime, perhaps in part by
building a custom TrialList in inline script much as shown in PST's
downloadable NoRepeat.es example. Although, just thinking of my own
example, it turns out that that one would be very easy to do in
E-Prime with proper application of Extended Input (see Appedix C of
the User's Guide) and the EchoClient (see the online E-Basic Help).
Don't know that any of this helps you, but there you go.
-- David McFarlane, Professional Faultfinder