[open] Qualtrics Survey Integration

Comments

• Edwin

  November 2014 edited 8:36PM

  Hi,

  You can't directly integrate it, but you can open it in a browser from OpenSesame. Put the following code in the Run phase of an inline_script:

  import webbrowser
  
  # change the URL to your survey's URL
  webbrowser.open("http://www.qualtrics.com/", autoraise=True)
  

  The most convenient place for this, I think, would be the end of your actual experiment (to prevent potential screw-ups when the browser is closed and the experiment needs to be raised to the foreground again). But this is up to you, and very much open for testing.

  Good luck!

  Edwin
• Anna13

  November 2014 edited 8:36PM

  Thank you. It is working. I put inline-script at the end of the experiment. but I was hoping to detect the period of time the subject spends on the survey to match it to his/her physiological data I'm recording. I don't have python programming language experience. When I did what you suggested, it ended the experiment and the pop up message appeared to save data file right before displaying the survey.
  Is there anyway to make the survey part of the whole experience. I mean once the subject finishes answering the survey, he gets back to the experiment in Opensesame!

  Also, Is there a way to make opensesame open a program in my laptop once the experiment start. I'm considering using emwave desktop relief system for measuring heart rate variability. The only concern I have is to synchronize the timing, I want to detect subject'sHRV during stimuli exposure. So, I was wondering if opensesame can open emwave program at the beginning of the experiment and add triggers to HRV measurements!
• Edwin

  November 2014 edited 8:36PM

  To return to the experiment afterwards, could you try appending a sketchpad item with a Duration of keypress to the experiment? (Place it directly after the inline_script)

  There isn't likely a good way of matching physiological data and the questionnaire, unless the website that collects the questionnaire provides you with timestamps of when participants answered each question. You can adopt the inline_script to save the exact starting time of the questionnaire, by inserting the following:

  import time
  import webbrowser
  
  # get the current time
  ct = time.localtime()
  date = time.strftime(u"%4.f-%2.f-%2.f" % (ct.tm_year, ct.tm_mon, ct.tm_mday))
  time = time.strftime(u"%2.f:%2.f:%2.f" % (ct.tm_hour, ct.tm_min, ct.tm_sec))
  
  # log the current time
  exp.log("survey start,%s,%s" % (date,time))
  
  # change the URL to your survey's URL
  webbrowser.open("http://www.qualtrics.com/", autoraise=True)
  

  BUT! Normally, when recording physiological data, you can send triggers to whatever recording equipment you use. This is usually done via the serial or parallel port (or a USB connection emulating a serial port). You can do so by using either the parallel or the serial trigger plug-in, which you can insert in the experiment directly before the inline_script.

  On your second question: Is your laptop a different computer than the computer you use to run the experiment? If it isn't, you can use the os module to make system calls, which can start up an executable:

  import os
  
  os.system('start "go" "C:\path_to_program\program.exe"')
  

  If you laptop is a different computer than the one you use to run the experiment, things get more complicated, as you will have to implement a way to make them communicate to eachother. If, as you say, you are a programming novice, this is probably not something you would want to attempt if you can find an easier way to get around it. It will require quite a bit of learning and messing around with scripting, and is therefore quite time consuming. If you do want to learn, we're happy to give you some pointers, of course!
• MichaelPi

  July 2015 edited 8:36PM

  I know it's an old thread, but here is how I've dealt with integrating Qualtrics as a step in an Opensesame experiment, using NirSoft's super-versatile Nircmd:

  PREPARE:

  NirPath = exp.get_file("nircmdc.exe")
  

  RUN:

  import subprocess
  from time import sleep
  subprocess.Popen('"C:\\Program Files\\Internet Explorer\\iexplore.exe" -k http://www.qualtrics.com')
  sleep(3)
  subprocess.Popen(NirPath + ' win activate title "Expyriment Backend"')
  subprocess.Popen(NirPath + ' win activate title "qualtrics"')
  

  I'm using a folder as filepool, and I've place nircmdC.exe from NirSoft in it.

  My experiment will run on several Windows machines, and I've opted to use IE's Kiosk mode to run the QT survey, with an instruction at the end to press Alt-F4. Not very elegant, but it works.

  Nirsoft's tool is used to restore the focus after the Ss close IE. It's an indirect route (activate OS, activate IE, let/(hope) windows 'revert' to OS when IE closes), but it seems to be working in internal testing so far. The study is not yet 'in production' so if anyone has constructive comments they will be greatly appreciated.

  The study can be seen at github (still a work in progress):
• eduard

  July 2015 edited 8:36PM

  Hi Michael,

  Thanks for sharing!

  Eduard

  
• MichaelPi

  August 2015 edited 8:36PM

  Back again with a related issue..
  I had to change a few things because the method I published wasn't working (can't remember how exactly it didn't work; so many test-runs so many ways to malfunction...).
  This is what I've got so far:

  PREPARE:

  import subprocess
  from time import sleep
  from multiprocessing import Process
  
  def processExists(processname):
          # taken from Ewerybody at http://stackoverflow.com/questions/7787120/python-check-if-a-process-is-running-or-not
      tlcall = 'TASKLIST', '/FI', 'imagename eq %s' % processname
      # shell=True hides the shell window, stdout to PIPE enables
      # communicate() to get the tasklist command result
      tlproc = subprocess.Popen(tlcall, shell=True, stdout=subprocess.PIPE)
      # trimming it to the actual lines with information
      tlout = tlproc.communicate()[0].strip().split('\r\n')
      # if TASKLIST returns single line without processname: it's not running
      if len(tlout) > 1 and processname in tlout[-1]:
          print('process "%s" is running!' % processname)
          return True
      else:
          print(tlout[0])
          print('process "%s" is NOT running!' % processname)
          return False
  
  def qualIE():
      subprocess.Popen('"C:\\Program Files\\Internet Explorer\\iexplore.exe" -k https://www.qualtrics.com/')
      print('qualIE ran')
      # just for debug, count repeats of processExists
      existReps = 0
      while(processExists('iexplore.exe')):
          print('existReps:')
          print(existReps)
          existReps = existReps + 1
          sleep(1)
  
  p = Process(target=qualIE)
  

  RUN:

  p.run()
  

  The end goal here is launch the Qualtrics survey (or any other external EXE for that matter), then set the focus to that EXE using Nircmd* (probably redundant, but just to make sure nothing slips by). Then, run a While loop to figure out when the EXE is no longer running, and set the focus to the backend, to continue the experiment.

  This seems to run as expected when running in windowed mode (Ctrl+W), but hangs on 'Preparing Experiment' when running in full-screen (Ctrl+R). I've tried with both the Legacy and the Xpyriment backends, with OpenGL turned on and off, as I thought that is the main difference between windowed and fullscreen modes (clearly there are other differences, but I thought this might be the crux of the matter in this case), but to no avail.

  What am I doing wrong here?
  I've posted the relevant code at
  https://github.com/pinusm/Opensesame_qualtrics_hang

  • I've left out the Nircmd related code in the above passage, because it does not seem to be involved with unexpected behavior I'm experiencing. For completeness, this is the function definition:

  NirPath = exp.get_file("nircmdc.exe")
  def front(windowsname):
      subprocess.Popen(NirPath + ' win activate title "%s"' % windowsname)
      print('Window "%s" is in front!' % windowsname)
  

  and these are the calls:

  # PREPARE: in def qualIE(), after 'subprocess.Popen('"C:\\Program Files\\Internet....'
  front('qualtrics')
  
  # RUN, after 'p.run()'
  front('Backend')
  
• MichaelPi

  August 2015 edited 8:36PM

  A bit more info, as it might be needed:
  I've been able to reproduce this behavior on Windows 10 and Windows 7, using Opensesame 2.9.6, 2.9.7pre5, and 3.0.0a8-py2.7.
  As mentioned, this only happens when running in fullscreen, not windowed. I've test both 'plain' run in a window mode (Ctrl-W, as mentioned), and quickrun mode (Ctrl-Shift-W).
  IE version was 11, in both testing environments.
  Michael
• sebastiaan

  August 2015 edited 8:36PM

  Hi Michael,

  What you're trying to do is pretty unusual, and I doubt someone will be able to help you with this. It's a matter of tinkering until it works, I suppose.

  But you mention that it does work in window mode, right? So what you could try to do is just use a window that covers the full screen. You can do this by:

  • setting the back-end to legacy
  • making sure that the resolution of your experiment matches the display resolution
  • under back-end settings, set window position to 0,0 and
  • draw window frame to 'no'

  Cheers!
  Sebastiaan

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

  August 2015 edited 8:36PM

  Hi Sebastiaan,
  Thank you very much for your suggestion! It appears to be working on my development machine, and I'll be sure to test it on the actual machine that will be running experiment and report back shortly!!

  I'd also like to apologize for filing an uninformative bug report. I thought the info provided here would be enough, and I must admit, I'm not very experienced with filing bug reports. Was there a guidelines document somewhere and I've missed it?

  The reason I'm inclined to think this is an Opensesame issue is that the behavior in fullscreen and windowed mode are different, and that fullscreen mode hangs while Opensesame is preparing the experiment, and that this happens also in a striped-down version of the experiment (https://github.com/pinusm/Opensesame_qualtrics_hang). While I don't have the proper training to know if this means something isn't working in Opensesame or the backend or someother background process, I think this means the problem is not with my part of the code (Again, I'm not very experienced; Please correct me if I'm wrong).

  I do not believe this is an unusual experimental design, as it applies to all external processes in general; it was IE and Qualtrics in my case, but IMHO a solid and robust way to integrate and run external process, and detect and regain focus once they've finished running, could be beneficial in many scenarios. That being said, I realize this might be a platform specific issue, and might not be what you'd like to invest your resources into.. I thank you for efforts anyhow, and highly appreciate your work and product (!!), and I'll try to contribute whenever I can...
  Michael