Millisecond Forums

Hi,

I am trying to program a version of the self-ordered pointing task (SOPT; Petrides and Milner, 1982).  In this task the participant is presented with a 4 x 3 array of 12 different pictures on screen. Their task is to pick a different picture on each of 12 trials.  It is a memory task and the number of repeated choices is the dependent variable.  Participants are not given feedback about whether their responses are repeated or not.

My first problem with this task is in recording correct (choosing a new picture) or incorrect (choosing a previously chosen picture) responses based on a participant's previous choices.  The problem is that i cannot do it, but it is not fundamental to examining the data because there is a record of the pictures chosen and I can look at that for repetitions.  However, that's not very neat and I'd rather Inquisit recorded correct or incorrect choices.  Any ideas?

The second and third problems are linked to the first in that they seem to require a record of a participant's previous choices.  There are two additional rules for the SOPT:  Rule 1 is that you cannot pick from the same position more than twice in a row.  Rule 2 is that you cannot pick stimuli in an alphabetical order (e.g. boat, cat, dog etc).  Participants are not informed of these rules until they break them. 

I wonder whether Inquisit 2.0 would support a program with these rules or whether I need to upgrade to version 3.0 or even just resort to programming in visual basic?  At the moment I am considering multiple conditional branching to include all possible choices, but that seems labourious, untidy and may not even work.  Any suggestions would be gratefully received :)

Gordon

Hi Gordon,

hmm, honestly I can't see any way to pull this off with Inquisit 2. This version simply lacks many of the features you'd need to make it work (e.g. conditional logic). I am pretty sure that it can be done with Inquisit 3, though. Although V3 certainly doesn't provide everything you'd expect from a full-fledged programming language (VB, C++, etc.), it does offer an amount of programmability that's sufficient to build most tasks. You might just put the 30-day trial period to good use and see if you can get the SOPT going in V3. There also are a bunch of sample scripts available at millisecond.com that might offer some hints / ideas for implementing the SOPT as they make extensive use of the programmability features and bear some other similarities to the SOPT (e.g. SymmSpan, Information Sampling Task).

Coding the task natively in VB is the safest bet of course, and if you're a proficient coder this just might be the fastest option, too.

Best wishes from a fellow Inquisit user,

~Dave

Thanks for the advice Dave.  Your conclusion was the one we'd arrived at in our lab with a couple of only semi-proficient Inquisit 2 users.  I will look at the scripts for version 3 and see what can be done.  The advantage of its use is that Inquisit is the common language in our lab.  Thank you again for the reply.

~Gordon

You're welcome. One question: Would you keep me posted on your progress? The SOPT seems like an interesting task and I'd really like to see it once it's finished.

Thanks and good luck!

~Dave

Hi Gordon,

Do you happen to have a detailed description of this task that you could share? This might be something we would program to be included in our task library.

Thanks,

Sean

The Petrides and Milner (1982) study used paper stimuli, and a version has been programmed into a task battery (for purchase) by Peterson et al (Peterson, J.B., Pihl, R.O., Higgins, D., Lee, A., 2002. NeuroCognitive Battery. Version 2. 0. ExamCorp, Deerfield Beach).  This version was described by Thush et al (2008; Addictive Behaviors, 32, 1367-1383.)

It's actually a pretty straightforward task. Participants are presented with a 4 x 3 matrix of 12 pictures (I am using black and white clip art drawings of everyday objects e.g. a book, a drink, a bus).  No sizes of pictures are specified in any of the citations above so I've just made them uniform across pictures and big enough that people can see what they are.  There are 3 blocks of 12 trials each.  On each trial participants are instructed to select one picture that they have not previously selected.  The idea is to pick a new picture on each trial and so the task is used as a test of memory.  The number of correct responses (unpicked items) across all three blocks is the dependent variable.

Other than the two rules I mentioned in my previous post there is nothing more to it.  I have programmed a version using Inquisit 2.0 but have had to modify it slightly to be explicit about the rules, but participants can still break these unless an experimenter watches them.  It's not ideal but it might meet our needs.  I've not tried anything yet in Inquisit 3.0 but mean to have a look later today or tomorrow.

If you need more information then I'll be happy to supply it if I can.

Good luck

Thanks much for the info, Gordon, that's extremely helpful.

I will definitely put this on our list for our own battery of tasks. If you are willing to share your current script, that would definitely speed things up. You can either reply to this thread with the script attached, or email it to support at millisecond dot com.

Thanks,
Sean

Sorry, I've just now read this.  I'll send you the script and I'll even throw in the stimuli I downloaded from the microsoft clip art.

:edited for spelling: