+- FindingFive Discussion Forum (http://discuss.findingfive.com)
+-- Forum: Researchers (http://discuss.findingfive.com/forumdisplay.php?fid=3)
+--- Forum: Experimental Design (http://discuss.findingfive.com/forumdisplay.php?fid=4)
+--- Thread: Comprehension questions after response following specific stimuli (/showthread.php?tid=53)
Comprehension questions after response following specific stimuli - atfrederiksen - 02-03-2021
I'm working on a forced choice experiment, where participants are presented with a sentence including a pronoun and then choose which of two names they think the pronoun referred to. That all works with my current code. However, I also want to follow some of the trials with a comprehension question (also a forced choice response) after participants have made their response. The problem I'm having is that the comprehension questions have to occur after specific items. I tried to use catch-trials, but that doesn't seem like the right approach, given that each comprehension questions can only follow one specific trial.
I also tried to set up as a separate trial template containing the trials to be followed by comprehension questions plus the comprehension questions themselves. By keeping the stimulus_pattern in a fixed order, I can make sure the appropriate comprehension question follows its intended stimulus + response. But then I could not figure our how to create a block consisting of N regular trials followed by exactly two of these separate trials, i.e. 1) a stimulus followed by its forced choice response, followed by 2) a comprehension question asking about something in the preceding stimulus sentence, followed by a forced choice response).
Does anyone know a way to do this?
RE: Comprehension questions after response following specific stimuli - Lucy - FF Researcher Support - 02-03-2021
Hello! Have you looked at trial yoking? This tutorial goes over the implementation: https://news.findingfive.com/2019/07/17/tutorial-yoked-trials/
Please let us know if that doesn't work for your design.
RE: Comprehension questions after response following specific stimuli - atfrederiksen - 02-04-2021
(02-03-2021, 12:53 PM)Lucy - FF Researcher Support Wrote: Hello! Have you looked at trial yoking? This tutorial goes over the implementation: https://news.findingfive.com/2019/07/17/tutorial-yoked-trials/
Please let us know if that doesn't work for your design.
Thanks! Yoked trials should work well for pairing comprehension questions to specific stimuli. What I still cannot figure out is how to mix in such yoked trials with non-yoked trials, that is, stimuli that are not followed by comprehension questions. For my study, only 25% of trials are followed by a comprehension question.
RE: Comprehension questions after response following specific stimuli - Ting - 02-04-2021
There are potentially two ways to approach this 25% requirement:
If there's no strict need for randomization of stimuli, then perhaps the easiest way is to use the catch trial feature (https://help.findingfive.com/api/procedures.html#catch_trials). The downsize is that it will introduce a pattern: a comprehension question pops up every X stimuli.
Alternatively, a hacky way to do it is to tweak the trial template that generates your comprehension question trials so that 75% of them actually a blank trial (or a sentence saying "this screen is intentionally left blank" or something like that). Then when you yoked everything together, it'll seem like only 25% of them have a comprehension question.
RE: Comprehension questions after response following specific stimuli - atfrederiksen - 02-04-2021
(02-04-2021, 10:23 AM)Ting Wrote: There are potentially two ways to approach this 25% requirement:
If there's no strict need for randomization of stimuli, then perhaps the easiest way is to use the catch trial feature (https://help.findingfive.com/api/procedures.html#catch_trials). The downsize is that it will introduce a pattern: a comprehension question pops up every X stimuli.
Alternatively, a hacky way to do it is to tweak the trial template that generates your comprehension question trials so that 75% of them actually a blank trial (or a sentence saying "this screen is intentionally left blank" or something like that). Then when you yoked everything together, it'll seem like only 25% of them have a comprehension question.
Thanks! The first option won't work in this case, because of the predictability of the questions. The second option seems to works - I created blank text stimuli and paired them with background_audio responses set to duration: 0.01. And if anyone has a more elegant solution than what I did here, I'd love to hear it! Thanks again.