Intermixed trial types
Hello,
I'm trying to make an experiment block that intermixes two different trial types. The experiment looks at semantic priming within an auditory lexical decision task. The key part of the experiment presents two lexical decision trials sequentially, with these trial contingencies based on a relevant/irrelevant semantic prime. The participant makes a response to both the prime and the target.
However, I would like my nonwords to be programmed as single trials without contingencies, so that real words and nonwords aren't always presented in pairs.
I thought I would do this by setting up a column in the block to determine trial type (paired, single), then use a 'run if' statement in the sequence to state that prime audio and prime response should only be played for paired trials. That didn't seem to work, so I tried embedding the different paired/single trial sequences into another 'parent' sequence and putting the 'run if' in the parent sequence (as was suggested elsewhere on this forum). That doesn't seem to work either.
On both occasions I get a message 'exception message: ` openexp._sampler.legacy.init() the file '.wav' does not exist. I have checked that all my wav files are loaded correctly, so I think it's because it's still trying to play the prime item when there's nothing there.
I would be super grateful for any suggestions on how I can make this work!
Many thanks,
Emma
Comments
Hi Emma,
I can't judge if the experiment is set up correctly: but the error is very straightforward: somehow the sampler is looking for a file without a name: namely the file: ".wav" . So all you'd have to do is to make sure that the variable with which you call the sound files is correctly defined and it might work (at least for this problem). Good luck,
Roelof
Thanks for your reply Roelof! Yes agreed. I've checked through all of those, and can only see that the problem must be in the 'run if' being set up properly - it tries to read/play a "prime" audio when there isn't one there, even though I've tried to ask it not to.
I've tried to say only run a prime audio and prime response collection if the trial type is labelled as X, but no such luck. Should this work in principle?
Emma
This method might work, but the problem might be with the order in which elements are prepared/executed. I think if you are using sampler items, these are prepared before the 'run if' statement is evalutated. One (unelegant) way around this is by adding a filler sound file to the file pool and have the sampler prepare this file, just to avoid the error. Another way would be to use an inline script and prepare the sampler item in the run phase, which is not executed if 'run if' is evaluated to false.
But these are just guesses, since I still don't have a clear idea of how your experiment is exactly set up, it would be easier to provide exact insight in the possible problem if we had a look at the experiment. Hope the above suggestions might work though, good luck
Roelof
Ahh that makes sense, I didn't understand enough about how the elements are prepared. The unelegant way works! And perhaps I might find some time to have more of an attempt with the script.
Thanks muchly for your help!
Emma