returning inline string in OSweb (vs. OpenSesame)

Comments

• DahmSF

  November 2019

  Hi S. Ross Marketing,

  The inline_javascript item works similarly to the regular inline_script item, except that it runs JavaScript instead of Python code. So you have to 'translate' your script.

  For further information see here.

  Cheers,

  Stephan
• srossmktg

  November 2019

  @DahmSF Aha, thanks.

  So basically, I need to get IT to hurry up setting up a an IRB data-compliant server so I can post the study online and then translate to js and then test it all?

  They just have to figure how to patch the XSS vulnerability on either JATOS or OS for me first...
• srossmktg

  November 2019

  Okay, I've got the study up on JATOS and I see the inline_javascript, but I'm unclear:

  a) how to get a random string validation code. At best, I've found this and input it in the input_javascript prepare tab

  //option 1 	
  var code = function () 
  {
   return Math.random().toString(36).substr(2);
  }
  

  b) I still have the inline_script in the prepare tab (see above). Do I still leave it, concurrent with the inline_javascript? Where/how do I translate the python into javascript?

  c) How do I modify my sketchpad then to output the code?

  ALTERNATELY

  Is there a simple means/procedure to get a random alphanumeric string code to output to participants in my sketchpad on OSWeb (per above)?
• srossmktg

  November 2019

  Still stuck.

  I've got this:

  define inline_script validation_code_script
  	set description "Executes Python code"
  	set _run ""
  	___prepare__
  	import random
  	import string
  	
  	def randomStringDigits(stringLength=6):
  		"""Generate a random string of letters and digits """
  		lettersAndDigits = string.ascii_letters + string.digits
  		return ''.join(random.choice(lettersAndDigits) for i in range(stringLength))
  	
  	# store identifier
  	var.set("identifier", randomStringDigits(8))
  	__end__
  

  this:

  define inline_javascript new_inline_javascript
  	set description "Executes JavaScript code (ECMA 5.1)"
  	set _run ""
  	___prepare__
  	//option 1 	
  	var identifier = function  () 
  	{
  	  return Math.random().toString(36).substring(2, 15) + Math.random().toString(36).substring(2, 15);
  	}
  	end
  

  and this:

  define sketchpad end_of_experiment_code
  	set duration keypress
  	set description "Displays stimuli"
  	draw textline center=1 color=white font_bold=no font_family=sans font_italic=no font_size=30 html=yes show_if=always text="Thank you for taking the time to participate in our study!<br><br>We appreciate your effort!<br><br>Please take note of the following validation code:" x=0 y=-128 z_index=0
  	draw textline center=1 color=white font_bold=no font_family=sans font_italic=no font_size=50 html=yes show_if=always text="[identifier]" x=0 y=75 z_index=0
  	draw textline center=1 color=white font_bold=no font_family=sans font_italic=no font_size=30 html=yes show_if=always text="and press any key to continue.<br />(This will return you to the Qualtrics survey instrument.<br>You will not be able to go return here.)" x=0 y=224 z_index=0
  

  Instead of getting this

  

  , all I want is for some unique alphanumeric identifier string to appear in OSweb (and log in the data) at the end of the RT study so participants can take the identifier string from OSweb and go enter it in Qualtrics (so I can append the RT data to the Qualtrics data).

  I understand there's something about translating from python in OpenSesame to javascript in OSweb, but this is pretty much my only painpoint left to launch my study and I don't have much in the way of programming skills. If anyone could help with some handholding, it would be extremely appreciated.
• srossmktg

  November 2019

  Replaced the inline_javascript function with a different one

  function randomStringDecimal() {
  var chars = [];
  var str = '';
  
  
  for( var i = 48; i <= 57; i++ ) {
  chars.push(String.fromCharCode( i ));
  }
  for( var i = 65; i <= 90; i++ ) {
  chars.push(String.fromCharCode( i ));
  }
  for( var i = 97; i <= 122; i++ ) {
  chars.push(String.fromCharCode( i ));
  }
  
  
  for ( var i = 0; i <=7; i++ ) {
  str += chars[Math.floor(Math.random()*chars.length)];
  }
  vars.identifier = str;
  }
  

  which removed the application error in OSweb. However, now nothing seems to be calling the function in OSweb and therefore the variable isn't being set nor displayed via the sketchpad.
• srossmktg

  November 2019

  FWIW to anyone looking at this in the future:

  What we did was add an inline_javascript_identifier item at the beginning. It incorporated the javascript:

  vars.identifier = Math.random().toString(36).substr(2, 9);
  // alert (vars.identifier);
  

  In the sketchpad where we wanted the code at the end of the study, we added the following:

  draw textline center=1 color=white font_bold=no font_family=sans font_italic=no font_size=50 html=yes show_if=always text="[identifier]" x=0 y=75 z_index=0
  

  We have a 9 character code (anticipating 300 participants and having timestamped data, the probability of two participants getting the same one was sufficient for us to not modify the JS).

  

  
  

  The study fails to run client-side, but the server-side application generates the code and writes it to the results.

  

  And now, participants can take the study online, get the code, it writes to results, they enter the code in Qualtrics, and we can use it to concatenate datasets!
• CSC

  September 2020

  Hi @srossmktg

  I have learned many tips in my study from your post: "returning inline string in OSweb (vs. OpenSesame)" in our study. Our study is the international study organized by Psychological Science Accelerator. You will see the details in this webpage: https://psysciacc.org/002-object-orientation/

  Before the global pandemic, we run this study in the lab room. Now we turned the script to OSWEB. Because this study is bundled with the other PSA study: https://psysciacc.org/003-gendered-prejudice/ We run this study on Qualtrics and this brought us a new challenge.

  Based on our registered plan, we run 002 and 003 studies in a randomized sequence. Some will run 002 then 003, and others will run 003 then 002. We have the problems how Qualtrics and JATOS exchange the participant's id and session id. Would you share your experience?
• eduard

  September 2020

  Hi,

  I might be a little off here, but maybe the JATOS crew can help here as well. Is that so, @kri ?

  Eduard

  
• kri

  September 2020

  Hi CSC, Hi Eduard

  it's always a bit difficult to arrive in a thread that was going on for a while: I'll try. I think CSC's questions is only loosely related to srossmktg original question. So I'll only answer this one.

  You seem to have 2 problems:

  1. Based on our registered plan, we run 002 and 003 studies in a randomized sequence. Some will run 002 then 003, and others will run 003 then 002.
  2. We have the problems how Qualtrics and JATOS exchange the participant's id and session id.

  1) In JATOS each of your experiments, 002 and 003, can be a component in a study. Additionally you would need a component before those that does the randomization. We have an example study "Randomize Tasks Between Workers" that does something similar and probably can be adapted to your use case.

  2) In JATOS's docs is a page about this: http://www.jatos.org/Use-Prolific.html. From Qualtrics you can pass on any parameter via URL query parameters. Then when you study run is finished you have to redirect back to Prolific with the link that Prolific gives you (again better explained in the docs).

  Hope this helped.

  Kristian