Empirisoft Support

    Welcome to Empirisoft Support
Results 1 to 5 of 5

Thread: Ramdomization Issues

  1. #1
    Join Date
    Aug 2007
    Posts
    4

    Ramdomization Issues

    Hi. I am creating an IAT and have gotten a bit stumped on randomization.

    For example, on a block with flower and good on the right and insect and bad on the left, I have 5 words per stimulus category in my text files. I would like to run 30 trials in the block in random order except that I would like the trials to alternate between either flower/insect and good/bad. So the trials would be random, but alternate between presenting a target and then adjective stimulus.

    However, I haven't been able to figure out how to randomize this way. In looking at the sample IAT input file, I see that they are randomized in this way, except that they are running 8 trials with an equal number of trials per category. With my setup of 30 trials, this doesn't work because some of the target and adjective stim lists automatically get more trials because of the way they must be listed. Is there a way to do what I'm trying to do?

  2. #2
    Join Date
    Nov 2005
    Posts
    3,328
    So does 5 divide equally into the number of times you call each stimulus within the block?

  3. #3
    Join Date
    Aug 2007
    Posts
    4
    I don't think so. I have 4 files of stimulus lists that I'm calling from for one block (target1, target2, adject1, adject2). There are 5 stimulus words in each of those lists. I would like to do 30 random trials but that alternate between listing one of the target words and one of the adjective words.

    For example, the way the stimuli are listed in the example IAT are:
    target1
    adject1
    target1
    adject1
    target2
    adject2
    target2
    adject2

    So basically, what I want to set up is exactly like the sample IAT, except that each stimulus file has 5 words instead of 4 and I want to have 30 random trials instead of 8. So the set-up of the sample IAT produces the effect I want, but because they are only calling for 8 trials, the 4 categories of stimulus equally divide between the trials so that each stimulus list is called on 2 times.

    With my setup of 30 trials this does not work to list them this way because if I just list 30 trials some lists would have to necessarily be used more than others. I thought that one potential fix could be to list 40 trials, but put a skip after the 30th trial to go to the next block (that way they’re evenly divided, but still random). However, I read in the manual that to do this, the trial on which you want to place the skip (so for me, trial 30) would have to be fixed. Is there any way to have all 30 trials random instead of having to fix one of them in place? Or is there another method besides skipping that would set up my block this way?
    Last edited by jarvis24; 08-22-2007 at 03:23 PM. Reason: clarification

  4. #4
    Join Date
    Nov 2005
    Posts
    3,328
    Might be easier if I could see your input file and stimlists. Could you reduce a sample version to just 1 block so that I can see exactly what you've got? Basically, whatever the bare minimum number of trials is that's required to illustrate the problem.

  5. #5
    Join Date
    Aug 2007
    Posts
    4
    I emailed you a zip folder with the input and stim files.

Similar Threads

  1. Recoding issues with response options
    By jarvis24 in forum MediaLab Older Versions: Suggestions
    Replies: 3
    Last Post: 01-23-2007, 02:44 PM

Posting Permissions

  • You may not post new threads
  • You may not post replies
  • You may not post attachments
  • You may not edit your posts
  •