Reaction-dependent ISI

I have come across a certain problem which I cannot solve on my own.

I would like to create a task where participants are presented two kinds of shapes (i.e. primary-task stimuli) and are supposed to press certain keys depending on the type of stimuli they see (e.g. ‘a’ key for circle and ‘l’ key for square). However in 25% of trials after the presentation of the primary-task stimulus an auditory stimulus (stop-signal) appears and the participants are supposed to withhold their reaction. And here comes the problematic issue: the interval between the primary-task stimulus and the stop-signal is initially set at 250 msec and should be adjusted continuously with a staircase tracking procedure: when inhibition is successful, the interval increases by 50 msec; when inhibition is unsuccessful, the interval decreases by 50 msec.

Is there any possibility to create such a conditon (i.e. the variable onset of an auditory stimulus depending on participant’s input in randomized 25% of trials) and if so, how can I do it? I’d be happy with any help.

This task, if at all possible in SuperLab 4.5 would take very long to set up because you would need to have one block of trials for ever 50 ms threshold, resulting on a lot of duplication. In SuperLab 5, this task would be much easier to implement. You would use the Trial Levels feature, in SuperLab 5 this feature adds a new level selection model called ‘manual.’ You can then create a rule (at the trial level) that would cause SuperLab to select from the next or previous trial level depending on inhibition.

Thank you for your answer (and for the new version of SuperLab).:slight_smile:

problematic issue


I think I have followed your advice but I guess I must have missed something. It appears to me that the rules and trial levels I have created are dead and are not used during the experiment. I’ve attached the experiment, perhaps I didn’t check or uncheck a certain option.

Is there a possibility that you took a look at those rules and trial levels? The SL4 manual unfortunately doesn’t give me any answers to my problem :frowning:

Best regards

circle_square.tgz (10.6 KB)

Here is a modified version of the experiment. This version demonstrates how to do the staircase design in SL5.

This modified version also uses stimulus lists, which is an important time-saving feature.
To run the modified demo, choose “demo-only group” as the Participant Group in the Run dialog.

The demo block uses only one trial: “shapes with sound.” This trial uses a stimulus list containing 5 circles and 5 squares.

For each shape in the list, notice that the same series of five events will happen:

  • cue
  • show one shape
  • pre-sound delay
  • sound
  • post-sound gather response

Also note that the only places where a participant will have the opportunity to press a key is during:

  • pre-sound delay
  • post-sound gather response

Other events provide no opportunity for a response, because they are set to end immediately (see the “Input” tab for each event).

  • On the feedback tab for "pre-sound delay" and "post-sound gather response" you can observe how the experiment is set-up to track whether inhibition was successful.
  • The three [B]Rules[/B] attached to trial "shapes with sound" accomplish the rest of the work.

Note: the two Rules for manipulating trial levels also currently causes a test text event to appear during the experiment. Once you are satisfied with how the experiment is working, then you can remove these test events.

new_circle_square.sl5 (13.4 KB)

Thank you, Monika!

The experiment you have attached seems to be a totally new one rather than a modification of the old one :slight_smile: I am not quite sure if I get a few things. In the experiment the limits of increase and decrease are 50 msec (min) and 1250 msec (max). If the signal delay is supposed to increase/decrease by 50 msec why are the levels in trial levels increasing/ decreasing set by 200 msec? Was it just an example? Or have I missed something? :slight_smile:

The time limit was changed for our testing purposes. The 50 ms increments that you used in your experiment was so subtle it was not obvious whether it was working right without looking in the datafile.

It’s important to note when you edit the 50 ms time limits, that you must edit both trial variables.

  • The first trial variable is "pre-sound interval" which is used by event "pre-sound delay."
  • The second trial variable, "total response window minus pre-sound" is used by event "post-sound gather response."

Changing the increment will involve changing both of those variables. A crucial piece of the design is that for each level in the two variables:

level 1 of one variable + level 1 of the other = 2000 ms
level 2 of one variable + level 2 of the other = 2000 ms
level 3 of one variable + level 3 of the other = 2000 ms
… and so on …

Anyone can edit our demo to choose a new increment size and a new total (instead of the 2000 ms total). However, both variables must be edited correctly to preserve the nature of the design.

I see, thank you very much! :slight_smile: