Thread Rating:
  • 0 Vote(s) - 0 Average
  • 1
  • 2
  • 3
  • 4
  • 5
"continue" button and trial duration
#1
Hello! :-) 
I have a question about the "continue" button. I see on the study specification grammar website that FF will automatically determine when a trial gets a "continue" button based on whether the trial duration is set to 0sec or more. i.e a study that has a duration of more than 0 won't (necessarily) get a "continue" button, but one with no duration or defined responses will. 
In my study I have trial templates with three stimuli all of which need to have different durations. The first needs to be 2000ms, the second 1000ms, and the third 500ms. None of the stimuli can have responses or buttons/button presses because they're meant to serve the same function roughly as fixations. In blocks they're yoked together with corresponding trial templates of three critical stimuli; essentially, one trial, composed of two trial templates, would be 2000ms pause >> critical stimulus >> 1000ms pause >> critical stimulus >> 500ms pause >> critical stimulus. 
I've found that I can avoid the stimuli requiring responses or button presses when I define the trial duration. However, if I keep the trial duration, then each stimulus can't have an individual duration of its own (or it can, but it would disappear before the whole trial template moves on leaving a blank screen). I have also tried to set each individual stimulus duration to its own respective duration. However, when I go to comment out the whole trial duration, the "continue" button automatically appears again. 
I'm wondering if there is a way to keep my code so that each stimulus can appear for its respective duration time (2000ms, 1000ms, and 500ms) without defining a trial duration and without the "continue" button automatically appearing? 
Thank you so much! 
Grace
Reply
#2
Hi Grace,

Thanks for taking the time to explain your study design so clearly. May I ask a somewhat redundant question to make sure I understand the conundrum you are facing: the 2000ms, 1000ms, 500ms pauses are generated from the same trial template, correct? Because of that, it forces you to set the longest duration, 2000ms, as the trial duration for the template. When that happens, for the pauses that last only 1000ms and 500ms, it would leave a blank screen before the trial moves onto the critical stimulus.

Did I get this right? So you are essentially looking for a mechanism to let trials advance after a duration that is conditioned on how long the pauses should be?
Reply
#3
(07-10-2021, 08:49 PM)Ting Wrote: Hi Grace,

Thanks for taking the time to explain your study design so clearly. May I ask a somewhat redundant question to make sure I understand the conundrum you are facing: the 2000ms, 1000ms, 500ms pauses are generated from the same trial template, correct? Because of that, it forces you to set the longest duration, 2000ms, as the trial duration for the template. When that happens, for the pauses that last only 1000ms and 500ms, it would leave a blank screen before the trial moves onto the critical stimulus.

Did I get this right? So you are essentially looking for a mechanism to let trials advance after a duration that is conditioned on how long the pauses should be?

Hi Ting, 

Yes! That is exactly the conundrum that I am facing right now. I should also clarify that the pauses are in one trial template (which is where my current conundrum arises), and the critical stimuli are in another (this trial template runs smoothly), and these are the two trial templates that are yoked together in an alternating fashion to form one trial.

Thank you! 

Grace
Reply
#4
Happy to help! We are brainstorming a solution for you, hang tight!
Reply
#5
Hi Grace,

We came up with a hacky solution for you but then realized that it might just make sense to create a new feature (a very minor tweak), which would make your design trivial to implement and would be generalizable enough to benefit other researchers as well. 

In short, we have extended the "barrier" property (check out this tutorial if this is the first time you've encountered this concept) to text stimuli as well. When a stimulus's barrier property is set to true, the rest of the content on the same trial waits for the stimulus to finish. Once the stimulus finishes, the trial proceeds. So, in your case, you'd want the text stimulus to have a duration of X seconds, and to act as a barrier. Then, once the barrier is released, the trial can auto advance to the next critical stimulus trial. 
Here's what you need to do:

  1. Starting tomorrow (7/13/2021), go to our testing site (https://testing.findingfive.com) to test this feature. This feature isn't available on the production servers yet.
  2. Add back the stimulus-level durations to the individual "pause" text stimuli, like "duration": 2, etc.
  3. Add a new "barrier": true property (no quotes on the word true) to the "pause" stimuli as well.
  4. Add "auto_advance": true (no quotes either) to the trial template holding the pause stimuli.
  5. Remove any trial-level duration from the trial template.
In other words, you are telling FF that each of those pause stimuli has its own duration and should act as barriers on their trials (for X seconds as defined by the duration). Once the barrier effect is released, the corresponding trials should advance on their own. 

I think this will produce the effect you are hoping for? Let us know if this works as expected. Once you give us the thumbs up, we'll move the feature to the production server.

Cheers!
Reply
#6
Hi Ting,

Thank you SO much! And my apologies for the delay in response!

I just tried it out on both my longer experiment and another sort of "tester" experiment. My longer experiment has image stimuli for the pauses and I first tried it there, and found the "continue" button still appeared. I tried it on text stimuli too as in my tester experiment, and I found that the pauses had disappeared altogether. I double checked the flow of everything between my trial templates and in blocks, and I found that when I removed the "auto_advance" on the trial with the pauses, they appear again.

I can try again later today and look a little bit longer at my code and read up more on the barrier feature.

Thank you again so much! :-)

Grace
Reply
#7
You are most welcome, Grace.

I just tested the feature and it works for me. Make sure you followed the steps in my previous exactly - setting both the duration and the barrier properties for the stimuli, and setting auto_advance in the trial templates, on the testing server (https://testing.findingfive.com). These features will work on the testing server only but not the production server yet.
Reply
#8
(07-15-2021, 09:15 AM)Ting Wrote: You are most welcome, Grace.

I just tested the feature and it works for me. Make sure you followed the steps in my previous exactly - setting both the duration and the barrier properties for the stimuli, and setting auto_advance in the trial templates, on the testing server (https://testing.findingfive.com). These features will work on the testing server only but not the production server yet.

Hi Ting, 

Just to clarify, is the testing server the same place that I built my experiment? When I click the link, it takes me to the finding five homepage. Perhaps last night it didn't work for me because I was testing it in the wrong place?
Reply
#9
The testing server looks identical to the production server, but in the URL address bar it will say testing.findingfive.com instead of www.findingfive.com. You'll see all your experiments there too, but they may behave differently on the testing server than the production server.

Give it a try again? If you can't get it to work, I'll ask Rachel to take a look at your study.
Reply
#10
(07-16-2021, 09:07 AM)Ting Wrote: The testing server looks identical to the production server, but in the URL address bar it will say testing.findingfive.com instead of www.findingfive.com. You'll see all your experiments there too, but they may behave differently on the testing server than the production server.

Give it a try again? If you can't get it to work, I'll ask Rachel to take a look at your study.

Hi Ting, 

It worked! I tried it with both text and images and it worked for both. This is amazing. Thank you again for all of your help!
Reply


Forum Jump:


Users browsing this thread: 1 Guest(s)