Trigger (stimulus) codes sent on response to EEG recording

[JDB]

I'm working on a project that involves interfacing an E-Prime
experiment with the ActiView/Biosemi interface for EEG recording.

I have this in the UserScript:
Const Sport as Integer = &H378

Every time an image appears, this code is sent immediately before:

InLine:
Intermediary.OnsetSignalEnabled = True
Intermediary.OnsetSignalPort = Sport
Intermediary.OnsetSignalData = c.GetAttrib("IntermediaryCode")

Intermediary.OffsetSignalData = 0
Intermediary.OffsetSignalPort = Sport
Intermediary.OffsetSignalEnabled = True

This works fine. The task then presents the participant with images,
to one of which they respond; the image is on-screen for 500ms,
followed by a fixation cross for another 500ms. They can respond at
any point, but the image & cross are on-screen for a set amount of
time. What I'd like is for the stimcode to be sent as soon as possible
after the response. I've been successful in sending a response code
after fixation cross goes away, using the following code:

InLine:
If TaskImage.RESP = "l" Then
resp = 55
ElseIf TaskImage.RESP = "s" Then
resp = 56
ElseIf TaskImage.RESP = "o" Then
resp = 57
ElseIf TaskImage.RESP = "w" Then
resp = 58
ElseIf TaskImage.RESP = "" Then
resp = 59
Else resp = 60
End If

FixationCross2.OnsetSignalEnabled = True
FixationCross2.OnsetSignalPort = Sport
FixationCross2.OnsetSignalData = resp

FixationCross2.OffsetSignalEnabled = True
FixationCross2.OffsetSignalPort = Sport
FixationCross2.OffsetSignalData = 0

However, this only works by sending itself at the beginning of a new
fixation cross. Ideally, subject responses would be sent WHEN they
respond. But I can't figure out how to do that. Is there some function
equivalent to "OnResponse"? It seems like it should be possible, but
I'm not sure how to do it.

Thanks for your help!

[David McFarlane]

If you use EP2.0.10.182 *Professional* or later then you are in luck
-- see the new Task Events feature documented at
http://www.pstnet.com/support/kb.asp?TopicID=4803 . (Remember, I
have uniformly advised everyone to spend the extra money for
Professional, no matter what PST advises.)

If you do not have EP2.0.10.182 Pro, then you might look at
http://groups.google.com/group/e-prime/browse_thread/thread/cfc3d0307d5c7fbd
and
http://groups.google.com/group/e-prime/browse_thread/thread/ef0e5a8d8b87aa0c
for hints. Or, of course, take this up with PST Web Support at
http://support.pstnet.com/e%2Dprime/support/login.asp , where they
strive to respond to all requests in 24-48 hours. And if you do get
an answer from PST staff, please extend the courtesy of posting their
reply back here for the sake of others.

BTW, although your particular inline code for the stimulus signal
works well enough (which is a virtue), it suffers in efficiency. I
have an entire lesson about handling these sorts of things in my
online course (although that lesson does not address your precise
response signal question).

-----
David McFarlane
E-Prime training
online: http://psychology.msu.edu/Workshops_Courses/eprime.aspx
Twitter: @EPrimeMaster (twitter.com/EPrimeMaster)

[JDB]

I'm glad to hear from someone else that the code appears to work. I do wish it were more efficient, but I don't have the budget to do the online course, for the moment. 

I've followed David's suggestion and upgraded to E-Prime 2.0.10 Professional (it was time), which does indeed have Task Events that seem to fit my needs. The problem is that the TaskImage.RESP is not a number, which is (I believe) what my external device needs—it's a letter. I don't think I can use an IF,Then statement in the parameters under Task Events. Am I missing something? 

One option I can think of is to send the accuracy (a 1 or 0) when the participant responds as a Task Event, and to send the numeric code for response at the Offset via InLine code (as before). 

Better, though, would be to now code using the code it generates from the Task Events. Is this possible? When I Generate the experiment, the relevant lines are:  

If ParallelPort.GetState() = ebStateOpen Then

TaskImage.Tasks.Add ParallelPort.CreateTask("TaskImage.Keyboard(1).Press", 0, WriteByteACC, , Long, CLogical(Yes))

End If 

(I've written out "WriteByteACC"—E-Studio writes it as "ebUCase_W, ebLCase_r", etc. I'm copying from another computer, but I'm also not sure whether the eb?Case letter-names are necessary.)

In any case, could I insert this as InLine code instead (before the TaskImage), and include the code I suggested above (*updated below)? Or should I instead look into the other links recommended (e.g., using the IsPending() function)? Any thoughts?

Barring suggestions, I will of course contact PST staff [and post their response here].

Thanks!

* Code to insert:

''''''''''''''''''''''''''''''''''''''''''''''''''''''''''''''''''''

' InLine - Code START

'''''''''''''''''''''''''''''''''''''''''''''''''''''''''''''''''''' 

If TaskImage.RESP = "l" Then 
        resp = 55 
    ElseIf TaskImage.RESP = "s" Then 
        resp = 56 
    ElseIf TaskImage.RESP = "o" Then 
        resp = 57 
    ElseIf TaskImage.RESP = "w" Then 
        resp = 58 
    ElseIf TaskImage.RESP = "" Then 
        resp = 59 
    Else resp = 60 
End If 

If ParallelPort.GetState() = ebStateOpen Then

TaskImage.Tasks.Add ParallelPort.CreateTask("TaskImage.Keyboard(1).Press", 0, WriteByte(custom), resp, Integer, CLogical(Yes))

End If 

'''''''''''''''''''''''''''''''''''''''''''''''''''''''''''''''''''' 

' InLine - Code END

'''''''''''''''''''''''''''''''''''''''''''''''''''''''''''''''''''' 

[David McFarlane]

Ooh, I did not think about Task Events having a
limited range of output options, thanks for
acting the pioneer and reporting back. I cannot
think of any way around that, so your best bet is
PST Web Support. We also use ActiView/Biosemi,
so I can confirm that for input it takes only
8-bit values, i.e., integers in the range 0-255 (aka &h00 to &hFF).

The bit of generated code that you supply does
not give us much hope either, it seems to confirm
that Task Events allows only a limited range of
fixed "Action" items. BTW, the "ebUCase_W" etc.
that you mention may be part of PST's effort to
reduce the use of quoted strings so as to avoid
"Script too big" problems, if you look in the KB
you can find some articles about this.

So offhand I agree that your best bet is your
plan to output a simple signal at the time of
response and then follow that with a later signal
for the coded response value. Just one hint, if
the signal on the output line is usually 0, then
the external equipment will not see another 0
output because there is no change on the output,
you have to output something different (I explain
this in my fMRI & EEG lesson).

BTW, I cannot resist a little programming style
note. The conventional way to indent your If-ElseIf-End If would be

If TaskImage.RESP = "l" Then
resp = 55
ElseIf TaskImage.RESP = "s" Then
resp = 56
ElseIf TaskImage.RESP = "o" Then
resp = 57
ElseIf TaskImage.RESP = "w" Then
resp = 58
ElseIf TaskImage.RESP = "" Then
resp = 59
Else
resp = 60
End If

At least you use ElseIf -- you wouldn't believe
how many people I see do this with an ugly and
inefficient sequence of plain If-Then-End If statements.

Regards,
-- David McFarlane

><http://www.pstnet.com/support/kb.asp?TopicID=4803>http://www.pstnet.com/support/kb.asp?TopicID=4803

>. (Remember, I
>have uniformly advised everyone to spend the extra money for
>Professional, no matter what PST advises.)
>
>If you do not have EP2.0.10.182 Pro, then you might look at

><http://groups.google.com/group/e-prime/browse_thread/thread/cfc3d0307d5c7fbd>http://groups.google.com/group/e-prime/browse_thread/thread/cfc3d0307d5c7fbd
>
>and
><http://groups.google.com/group/e-prime/browse_thread/thread/ef0e5a8d8b87aa0c>http://groups.google.com/group/e-prime/browse_thread/thread/ef0e5a8d8b87aa0c

>
>for hints. Or, of course, take this up with PST Web Support at

><http://support.pstnet.com/e%2Dprime/support/login.asp>http://support.pstnet.com/e%2Dprime/support/login.asp

>, where they
>strive to respond to all requests in 24-48 hours. And if you do get
>an answer from PST staff, please extend the courtesy of posting their
>reply back here for the sake of others.
>
>BTW, although your particular inline code for the stimulus signal
>works well enough (which is a virtue), it suffers in efficiency. I
>have an entire lesson about handling these sorts of things in my
>online course (although that lesson does not address your precise
>response signal question).
>
>-----
>David McFarlane
>E-Prime training
>online:

><http://psychology.msu.edu/Workshops_Courses/eprime.aspx>http://psychology.msu.edu/Workshops_Courses/eprime.aspx
>Twitter: @EPrimeMaster
>(<http://twitter.com/EPrimeMaster>twitter.com/EPrimeMaster)

[JDB]

Thanks, David. 

I've submitted a support request; I'll be sure to update here if anything useful comes of it. For now, I'll stick with the plan I began to outline:

1. Send a set code when the user responds (e.g., custom value = 54). 

2. At the subsequent slide, send the actual response as done before (using If/Then) statements. 

While not particularly elegant, this solution should allow such analyses as I'm interested in doing. (Performing conditional operations on EEG data will be possible, as will time-locking to participants' responses.) 

J

[JDB]

And follow-through from PST: 

At this time it is not possible to use Task Events as you describe, but we are considering whether it is possible to change this behavior in the future. I am glad you are able to use the other option for now. 

Well, glad we can do it in some way.

J