Set the timing to record RT?

[Martin Yang]

Dear all,

I'm still a rookie in E-Prime and designing an ABX discrimination
task. I got a question about the timing to record RT. Does anyone know
how to set the timing to record RT in E-Studio 2.0? I looked up the
user's guide and it says that RT is always recorded from the onset of
the response logging event. However, this is not the way I want. I
need to record RT right AFTER subject has heard the audio stimuli. I
tried two different methods, but none of them gave me accurate RTs.

The first thing I tried was making the sound object itself as the
response logging event, but in that case, RT would be measured from
the onset of the event, which was BEFORE the whole sound file has been
played to the subject and the RT would include the duration of the
sound "file", which was way longer than subject's actual response.

The second thing I tried was adding a "Wait" object which records RT
AFTER the sound object. The duration of the sound object was set to
1500ms, but given that the length of sound files were all different,
the wait object didn't immediately start right after the sound "file"
was played. Rather, the wait object started after the whole 1500ms of
sound object had elapsed. Now the problem is that subject may already
start responding before the whole 1500ms has elapsed, so some parts of
RT will be missing since subject starts responding before the wait
object which records RT starts. So, does anyone have any idea to
record RT right after the sound "file" has been played to the
subject?

Thank you in advance for your great help!!

-Martin

[Peter Quain]

maybe something like this would work (or maybe not..): use 'dummy'
object (a text object, duration = 0, wait for response = forever;
response action = jump to label) just before sound object, to collect
RT. There is a sample experiment on e-prime samples which illustrates
this, maybe someone remembers its name? Is it CollectInputDuringInLine?

Log onset time of dummy. Log offset time of sound object. Compute
adjustRT = sound offset - dummy onset. After RT is collected, compute
realRT = dummy.RT - adjustRT . Set realRT as attribute.

[Peter Quain]

You say you're new to e-prime. Using dummy object, something like
this in an inline before the end of a trial would log true RT from
sound offset:

dumOnset = c.GetAttrib("dummy.OnsetTime")
dumRT = c.GetAttrib("dummy.RT")
soundOffset = c.GetAttrib("soundObject.OffsetTime")
adjustRT = soundOffset - dumOnset
realRT = dumRT - adjustRT
c.SetAttrib "ReactionTime", realRT

[David McFarlane]

Martin,

What if you set the End Sound Action of your SoundOut object to
Terminate? Wouldn't that end the SoundOut as soon as your sound file
ended (before the 1500 ms Duration of the SoundOut object) and then
go on to collect your RT from there?

-- David McFarlane, Professional Faultfinder