[solved] Need help fixing ST-IAT problem

Comments

• eduard

  June 2015 edited 1:38AM

  Hi,

  In order to help you, I would need more information, i.e. how did you implement the loop, the keyboard_response, the timeout and such. The problem can have many causes, it's hard to spot it at this point.

  Thanks,

  Eduard

  
• hughturpin

  June 2015 edited 1:38AM

  Hi Eduard,

  Each loop has 24 cycles which run one time each in a random order.

  The loops consist of a sequence containing: 1) a fixation dot (duration 500 millisecs); 2) a sketchpad which displays the 24 cycles (these are JPEG images); 3) a keyboard response (correct response and allowed response are both entered as [Response], which is E or I depending on the cycle; timeout is set to infinite); 4) the logger.

  Loops are separated by a sketchpad item with instructions for the next test followed by a keyboard response to begin (space bar).

  Here's a link to an image:

  

  Many thanks for any advice you are able to give me with this,

  Hugh
• hughturpin

  June 2015 edited 1:38AM

  Can anyone help with this? Deadlines are looming...
• eduard

  June 2015 edited 1:38AM

  Hi,

  So far I don't see an obvious mistake here. Do your initial fixation screen and waiting-for-response screen look identical? In other words, is it possible that the problem of not waiting in the very end of a trial happens always, but only after the last trial one can see it? Does the logfile look normal and if so, do your responses match your expectations? Finally, what happens if you tick the "Flash pending key presses at sequence start" box?

  Sorry, but this issue is a bit unclear to me. Would you mind uploading the experiment, so that I can have a look on it directly?

  Thanks,

  Eduard

  
• hughturpin

  June 2015 edited 1:38AM

  Hi Eduard,

  No Problem. The log file seems normal and ticking the 'flush pending key presses' button hasn't made any difference to the problem. The fixation dot and response-waiting screens are different, and the problem definitely only occurs on the last trial in each loop. In all the others it waits for input. It's very strange.

  I've uploaded the ST-IAT (two versions - both have the problem). To run the experiment you'll need the stimuli files for the file pool. There are dozens of them and Dropbox won't let me share folders without an email address. If you need to run the experiment rather than just look at it in Opensesame is there a way you'd like me to send you the stimuli folder?

  https://dl.dropboxusercontent.com/u/49609310/STIAT1.opensesame
  https://dl.dropboxusercontent.com/u/49609310/STIAT2.opensesame

  Thanks again for taking a look at this,

  Hugh
• eduard

  June 2015 edited 1:38AM

  I'm not sure whether this is good or bad news, but you're experiment is running just fine for me. So, also on the last trial nothing will happen until the response key is pressed. After that however, it will go right away to the next loop. If you're just checking quickly whether the experiment works by pressing any key, it might appear as if the last element is skipped. But that's not what happens in your case?
  Maybe it's a version thingy. Which one do you use? I have have OpenSesame 2.9.6 running on Ubuntu.

  Eduard

  
• hughturpin

  June 2015 edited 1:38AM

  Thanks for looking at it. That is really strange. I'm running 2.9.6 as well, but on horrible Windows 8. Maybe I should try it on a computer with a different operating system.

  I've run through it thoroughly several times now, behaving as if I were a participant, and the last trial in each loop is definitely skipping the input on my computer. By the way, what did you use to populate the file pool? I'm using a load of JPEGs. Could it be related to that somehow? Just grasping at straws here...
• eduard

  June 2015 edited 1:38AM

  I was using 2 JPG images of my own. One I used for with the sequence and the other for the instruction slides.

  Could you maybe try to insert another element at the end of your sequence, e.g. another empty sketchpad presented for 50ms or something similar. In this way, you could force your last trials to be skipped at a later point in the sequence, so that responding won't be affected.

  Btw., when you say that the logging file looks fine, does it also include the last trial? Does it appear just without response, or is it not present at all?

  
• hughturpin

  June 2015 edited 1:38AM

  Hi Eduard - I don't know why, but it is working now. I uninstalled and re-installed Opensesame and opened the file again and now it runs properly. No idea why that worked, but I'm very relieved.

  Many thanks for all your help!
• eduard

  June 2015 edited 1:38AM

  Good to hear.

  Good luck with your data. acquisition

  Eduard