[open] Experiment does not save anymore

Comments

• sebastiaan

  November 2013 edited 10:11PM

  Hi Marianne,

  If variables are missing from the logfile, you can add them explicitly by disabling 'Automatically detect and log all variables' in the logger, and add the missing variables through the 'Add custom variable' button.

  If this doesn't solve things, could you describe your problem in more detail so we don't have deduce the problem from inspecting your experiment and logfiles!

  Cheers,
  Sebastiaan

  
  
  Check out SigmundAI.eu for our OpenSesame AI assistant!
  
• MLittel

  November 2013 edited December 2013

  I'm sorry, i'll try my best to explain in more detail.

  We present subjects with eye-movements (smooth-pursuit plugin) and beeps at the same time. There are high and low beeps and participants have to respond as fast as possible to the beeps by pressing "z" or "/" on the keyboard. Simultaneously they have to follow the movement of the dot with their eyes. After that they have to recall a memory and make the same eye movements while thinking of that memory. Afterwards they score the memory on 2 VAS (slider plugin).
  This loop is repeated for eye-movements of different frequencies (0.4, 0.6, 0.8, 1.0, 1.2 Hz and no eye-movements (blank)). These loops are presented randomly.

  A loop of one certain frequency (smoothpursuit+beeps only) looks like this:

  -LoopDualTask
  
   -SequenceDualTask
  
   -Parallel
  
     -Smooth_pursuit (e.g., 0.4 Hz, for 170000 ms)
  
     -LoopBeeps (2 beeps, repeated 40 times)
  
     -SequenceBeeps
  
      -Advanced_delay (duration 2600 ms, jitter 400 ms)
  
      -Sampler (sound file [beep] defined in "Loopbeeps", duration = 0) 
  
      -Synth (length=0, duration = 0)
  
      -Keyboard_response ([CR] defined in "Loopbeeps", allowed responses z,/, Timeout=2000 ms, flush pending keypresses)
  
      -logger
  

  I want opensesame to record the reaction time to each beep, and whether the response is correct (high or low tone). When a button is pressed, the next beep must be presented with an interval of 2600 ms +/- 400 ms. The experiment has to wait with presenting a new tone until a button is pressed in response to last tone OR until 2000 ms are over (Timeout = 2000 ms).

  So: beep - wait - buttonpress - 2600+/-400ms - beep

  Or: beep - wait - 2000 ms timeout - 2600+/-400ms - beep

  Did I do this correctly? (I guess not)

  Right now the logger does not log all responses. For most beeps the output file says response = none (despite the fact that I really pressed the buttons). Furthermore the experiment always seems to wait for 2000 ms until presenting a new tone, even when a button was pressed.

  Furthermore the logger does not log which eye_movement frequency was presented. I declared a new variable in "LoopDualTask", i.e., "EyeMovement_frequency" with the correct frequency underneath. But the logger does not record this; displays NA instead.

  I want to note that for random presentation of the different loops of different frequencies, I had to use an inline script from this website:

    import random
  
    sequence_list = ['04sequence', '06sequence', '08sequence', '10sequence', '12sequence', 'NoEM_sequence']
  
    random.shuffle(sequence_list)
  
      exp.items[sequence].prepare()
  
      exp.items[sequence].run()
  

  ...and I had to create a dummy list with all the frequency loops. This dummy list is set to "run if = never". I'm not sure whether this interferes with logging, but after I inserted this script the experiment did not log everything anymore.

  Before, the experiment DID log all responses, DID log EyeMovement_frequency (and I believe it did not wait till 2000 ms before continuing to the next beep, but I'm not sure). Of course, before the frequency loops were not presented at random.
  I'm a total noob and can't figure out what went wrong.

  I have another question about this study design. It is not very elegant to present the smooth_pursuit for 170.000 ms, which is a guestimation of how long it would take for a participant to respond to all beeps. It would be nice if the eye movements stopped after the participant responds to the last beep. How can i do that?
  So how can you make to operations in a Parallel stop at the same time?

  Thank you in advance
• sebastiaan

  December 2013 edited 10:11PM

  Hi Marianne,

  Well that's quite a lot that you want to do at the same time! I think you'll be much better of using an inline_script than trying to build this using the GUI. It may be possible, but it will end up being a complicated mess (looking at the structure you describe above, that may have already happened!).

  Below you see a script that does the following, which, from what I understand, is about what you need:

  • Show an oscillating smooth pursuit target
  • Play 2 beeps, and record the response time after each one.
  • End after second response

  This is script is a bit tricky, because it requires some bookkeeping about where you are in the trial. But if you carefully walk through the logic of the loop and carefully read the comments, you should be able to figure out how it works.

  import math
  from openexp.canvas import canvas
  from openexp.keyboard import keyboard
  from openexp.synth import synth
  
  amp = 200. # Vibration amplitude (pixels)
  freq = .5 # Vibration frequency (hz)
  T = 1000. / freq # Vibration duration is derived from frequency
  dur1 = 1000 # Duration before first beep
  dur2 = 1000 # Duration before second beep, relative to the first response
  
  # Create canvas, keyboard, and synth (beep) objects
  my_canvas = canvas(exp)
  my_keyboard = keyboard(exp, timeout=0)
  my_synth1 = synth(exp, freq='a1')
  my_synth2 = synth(exp, freq='c1')
  
  wait_for_resp1 = False
  wait_for_resp2 = False
  
  t0 = self.time()
  while True:
      # Draw the smooth pursuit target
      t = self.time() - t0
      dx = amp * math.sin(((t % T)/T) * 2 * math.pi)
      my_canvas.clear()
      my_canvas.fixdot(x=my_canvas.xcenter() + dx)
      my_canvas.show()
  
      # If its time for the first beep, and it hasn't already been played, then
      # play it and start waiting for the first response.
      if dur1 != None and t >= dur1:
          my_synth1.play()
          wait_for_resp1 = True
          dur1 = None
  
      # If its time for the second beep and the first beep has already been
      # played and we are not waiting for the first response, then play it and
      # start waiting for the second response.
      if dur1 == None and not wait_for_resp1 and dur2 != None and t >= dur2:      
          my_synth2.play()        
          wait_for_resp2 = True
          dur2 = None
  
      # If we are waiting for the first response, log it, and adjust the duration
      # for the second beep based on the response time.
      if wait_for_resp1:      
          resp, t_resp = my_keyboard.get_key()
          if resp != None:
              exp.set('response_1', resp)
              exp.set('response_time_1', t_resp - t0)
              wait_for_resp1 = False
              dur2 += t_resp - t0
  
      # If we are waiting for the second response, log it, and break the loop.
      # Done!
      if wait_for_resp2:      
          resp, t_resp = my_keyboard.get_key()
          if resp != None:
              exp.set('response_2', resp)
              exp.set('response_time_2', t_resp - t0)
              break
  
      # Sleep a bit to give the computer time to do other things, such as playing
      # the beep in the background.
      self.sleep(10)
  

  Good luck!

  Cheers,
  Sebastiaan

  
  
  Check out SigmundAI.eu for our OpenSesame AI assistant!