Problems with Just Noticeable Difference expt

Hi all,

I’m new to SuperLab (4.0.7b) and am trying to set up my first experiment: a just noticeable difference study using audio files.

The current setup is shown in the image below. I have also uploaded the .sl4 file, minus audio files of course.

The experiment currently works like this:


  • The Instructions block begins - participants are told "You will hear voice pairs - please hit 'Y' if you can hear a difference between them, or 'N' if you can't".
  • The Task block begins. An 'A' is printed on screen, then the first audio file in list A is played, then a 'B' is printed on screen and the first audio file in list B is played.
  • The Response event occurs and participants hit Y or N.
  • Steps 2 and 3 are repeated until every pair of files from lists A and B have been played and responded to.
  • The End block occurs, telling participants that the study is over. [/LIST]

    I am having trouble setting up three things:


  • I would like the audio pairs to be played in random. At the moment the audio file pairs are played in order according to their position in the stimulus lists. I can't randomise the trial because I need to preserve the event structure "Play A, Play B, Response".
  • I would like the voice pairs to be played three times before the participant responds, instead of just one. So it should be ABABAB-Response, rather than AB-Response as it is currently.
  • I would like the Y and N keys to be the only ones that 'work'. At present, if the participant accidentally hits another key the experiment continues. This isn't a major problem, but I'd like to be able to minimise the number of incomprehensible responses. [/LIST]

    jnd.sl4 (8.25 KB)

  • Actually, you’re pretty close already.

    1. Randomize the trials in the block. This won’t randomize which two items get paired together. It will randomize the trials that get created as a process of using the stimulus lists. Select Experiment menu->Expand Events That Use Lists… to see what happens. This will create a COPY of your experiment just as it would be at runtime.

    2. At the moment, your two stimulus lists have the same files in them, which isn’t what you want. These two lists represent your pairs. The good thing here is that if you use the same list more than once in the same trial, it references the same stimulus. Therefore, all you have to do to get your “ABABAB” is copy and paste the four AB events you’ve already created.

    3. Currently, the only way to limit input to the “Y” and “N” keys is to set them both as the correct response and to set the event to move forward only on a correct response.

    Thanks, Hank

    I don’t have a copy of the software on my home machine, but I will put your suggestions in place on Monday.

    The two stimulus lists do in fact have different files in them, they just have very similar names (the only difference is a minus symbol at the start of the names for one set). I did previously try randomising the events in the trial and everything got jumbled up, but you’re saying that randomising the trials in the block will do it? I should have tried that, really…

    Anyway, thanks a lot for the reply and have a good weekend.

    ohhhhh… the minus symbol. I even went through both lists multiple times and thought I was verifying every one of the characters. My mind skipped the “punctuation.”

    Everyone tries randomizing the events in the trial as a method of randomizing stimulus lists. I even tried putting a tool tip on it explaining that it won’t do that. Sometimes, I wonder if Hisham’s method of putting the blue text in would actually make a difference here.


    Sorry to be a pain. I followed your advice but I still can’t get the experiment to run random trials (question 1 in my first post). The first voice-pair that gets played always comprises the first stimulus in lists A and B, the second pair the second stimuli, and so on.

    I’ve re-uploaded the .sl4 file - I hope you don’t mind taking another look at that. I can’t see any other settings I can change to get it to work…


    jnd.sl4 (10.8 KB)

    Are you certain? You have it set up correctly. I plugged in my own sound files at my end, and it was randomized. I ran through it about a half-dozen times to make certain, and the order of the pairs does change. The pairs themselves don’t, but the order does.

    I’m really sorry to have bothered you - you’re right. I had run through the experiment three times and the order was the same each time. Must have been by chance. The odds of that must be quite high, even with just three pairs…

    Thanks for getting back to me and for taking the effort to set up the expt with your own audio.