[open] Displaying Error Feedback with 'Swipe Stimuli' on Android

Comments

• eduard

  November 2015 edited 4:39PM

  Hey,

  I don't really understand the problem. When exactly do you want to present the feedback? During the movement, or after it is done? From what I get so far, your situation is not much different than in a regular experiment.

  Would you mind helping me out here?

  Eduard

  
• CharWes

  December 2015 edited 4:39PM

  Hi Eduard,

  I would like the feedback to be presented if the image is swiped in the wrong direction (there is a sketchpad item before each image telling the participant which way to swipe), which I thought could be done by importing another sketchpad item to the script when it's the wrong way?

  Also, there needs to be feedback if the response time is too short or too long, however I'm not entirely sure how to do this.

  So basically, for the former, the feedback would be after the movement is done (i.e. if the response is incorrect) and the latter would be before the movement if it is too slow and after if it was too quick, if that makes any more sense?

  Thanks!
• eduard

  December 2015 edited December 2015

  Hi,

  Assuming that you have a variable that codes the correct response, this code here (adapted from the one you posted) should do what you want.

  from openexp.mouse import mouse
  from openexp.canvas import canvas
  
  my_mouse = mouse(exp)
  my_canvas = canvas(exp)
  feedback_cv = canvas(exp) # new
  
  correct_response = exp.get('correct_response') # new
  path = exp.get_file(u'NEM4.png')
  upper = exp.get("height") - 200
  lower = 200
  min_rt = 100 # new
  max_rt = 2000 # new
  my_canvas.image(path)
  trialstart = my_canvas.show()
  while True: # new
      button, position, movstart = my_mouse.get_click(timeout=None)
      exp.set ("position_NEM4", position)
      if movstart - trialstart < min_rt: # new
          feedback = "Too quick!" # new
          break # new
      elif movstart - trialstart > max_rt: # new
          feedback = "Too slow!" # new
          break # new
      while True:
          pos, time = my_mouse.get_pos()
          my_canvas.clear()
          my_canvas.image(path, x=pos[0], y=pos[1])
          timestamp = my_canvas.show()
          if pos[1] < lower:
              response = "top"
              responded = True # new
              break
          if pos[1] > upper:
              response = "bottom"
              responded = True # new
              break
      break
  
  if responded: # new
      if correct_response == response: # new
           feedback = "Correct!" # new
      else: # new
           feedback = "Wrong!" # new
  feedback_cv.text(feedback) # new
  feedback_cv.show() # new 
  self.sleep(1000) # new
  # calculate the movement time
  movtime = timestamp - movstart
  # calculate the RT
  resptime = movstart - trialstart
  exp.set("response_NEM4", response)  
  exp.set("movetime_NEM4", movtime)
  exp.set("reponse_time_NEM4", resptime)
  
  # note: I did not define a correct response here. 
  # This should have happened earlier
  

  As you can see I added another while loop to make it work. Can you try the code and let me know whether it does what it supposed to do?

  Thanks,

  Eduard

  
• CharWes

  December 2015 edited 4:39PM

  Whereabouts should I be defining the correct response? Is it in this bit:

  correct_response = exp.get('correct_response') # new 
  

  or

  if responded: # new
      if correct_response == response: # new
           feedback = "Correct!" # new
  

  The correct response for example would be 'bottom'.

  Thanks for your help.
• eduard

  December 2015 edited 4:39PM

  Rather the first than the second, in best case even earlier. Like you said, a sketchpad informs the subjects which direction is the correct one. I suppose this direction is not constant but can change over trials. Therefore, you would have to have a variable, whose value is updated on every trial and defines the direction. This variable (if it has the values bottom and top) you can use as correct response variable.

  The typical approach is to define for every unique trial the respective correct response in the loop table. If you did so, you can retrieve those values with the command exp.get(variable).

  Does this make sense?

  Eduard