search

kschuler / words-in-noise-EEG

An EEG task for Kelly Michaelis and Peter Turkeltaub. Learners try to determine words presented in a noise mask with either just auditory information or both auditory and visual information.
1 stars 2 forks source link

Pulling stimuli so that sets don't repeat within blocks #21

Open kcm1231 opened 7 years ago

kcm1231 commented 7 years ago

The E and H stimulus sets are balanced such that each set contains 100 trials of the same 4 answer choices, and the difference is that in the E condition, word 1 is the target and in the hard condition, word 2 is the target. We needed to have it set up this way so that the E and H sets are as comparable as possible. However, this means that in each block, you could get two consecutive trials with the same 4 answer choices. To fix this, I think we would need to either add some lines of code so that it makes sure not to put matched E and H stimulus sets in the same block, OR possibly create subfolders within the E and H folders and have the task draw from a new subfolder for each block. That way I could just make sure that the subfolders contain unique sets of stimuli.

NOTE - this will also be an issue with the Phoneme and ES conditions.

KS says: Yes, I see what you want to do here. This one requires a little more restructuring so I can’t give you a quick fix. But it can be done. The best way to do this is in the sub-folder way you mentioned above. I can, in principle, ask it not to choose a stimulus that contains an answer choice that was used already in the block. However, this could create problems toward the end of the set. I can explain in person on Friday some options.

kcm1231 commented 7 years ago

Almost done with the new stimulus folders. I'm making subfolders within each easy and hard folder. The subfolders are titled "block1". For each condition, blocks 1 and 2 are the baseline staircase, and blocks 3-12 are the main staircase. Each block folder has 10 stimuli in it. Let me know if this works or if I need to format differently!

kcm1231 commented 7 years ago

Do I need to make subfolders for the movie files too? Or will it pull the .wav file and then just find the corresponding .mov file?

I can easily make subfolders for the movies, but since there aren't easy/hard subfolders, I wasn't sure if block-specific subfolders were needed. If I do need them, I can make easy/hard parent folders and put them in there. I'll have to, since block1easy is diff stimuli than block1hard.

kschuler commented 7 years ago

Hey there - I wouldn't yet. The sub-folder architecture may not work out and may not be the simplest way to do this.

I don't really have any time to think about this between now and Monday... but after that I'll get back to you with some changes.

I actually bumped into Peter today so I asked him a few questions I have and now feel like I understand what you guys are doing a bit more.

On Fri, Feb 3, 2017 at 12:38 PM, kcm1231 notifications@github.com wrote:

Do I need to make subfolders for the movie files too? Or will it pull the .wav file and then just find the corresponding .mov file?

I can easily make subfolders for the movies, but since there aren't easy/hard subfolders, I wasn't sure if block-specific subfolders were needed. If I do need them, I can make easy/hard parent folders and put them in there. I'll have to, since block1easy is diff stimuli than block1hard.

— You are receiving this because you are subscribed to this thread. Reply to this email directly, view it on GitHub https://github.com/kschuler/words-in-noise-EEG/issues/21#issuecomment-277311784, or mute the thread https://github.com/notifications/unsubscribe-auth/AF0T7DVZnMkr-BZhCXvJ-HnRu1qBkZm9ks5rY2YLgaJpZM4LvKvk .

kcm1231 commented 7 years ago

Ok, sounds good, I'll hold off. Thank you for letting me know! and good luck!

kcm1231 commented 7 years ago

Hey there, I noticed that the way we have it set up now, in the Environmental and phoneme conditions, it's stopping the experiment after 70 and 60 trials, respectively. For these conditions, we have chunks of stimuli (30 and 40, respectively), that occur twice in the wordlist. I looked into it, and it seems like it's going through the wordlist and pulling randomly without replacement, but it's not treating the duplicated stimuli on the list as new items. This issue should go away when we change how the stimuli are pulled, but I wanted to let you know it was happening!