I am using a version of the emotional stroop task with voicerecord and have two problems.
1. It seems as if the latencies are incorrect. Our pilot data indicates that participants are responding to words as quickly as 30 to 40ms.This seems a little too fast for vocal responses. I also ran the original version from the task library and was getting similarly fast RTs. Do you think this is a problem with the script or is it could it be something else like the microphone or speech recognition software?
2. When I attemted to run the 'analyze recorded responses' tool I keep the following error
'Speech Error -2147201023: Speech Error. Line 427, File win\SpeechEngine.cpp'
I've even install the previous speech recognition engine (although probably not necessary since I'm using a windows 7 machine), but I am still getting the same error. Any advice would be greatly appreciated.
It could be any of those things. You may want to test the script on a different machine and follow the below advice from the "How to Setup and Use Setup Speech Recognition" documentation topic.
"Before speech recognition will work, you must run the Microphone Wizard to adjust microphone input to appropriate levels. To run the Microphone Wizard, go to the Tools menu in Inquisit, select Speech Recognition, and then select Microphone Wizard."
When I attemted to run the 'analyze recorded responses' tool I keep the following error
Note that Windows 7 is not among the supported operating systems for Inquisit 2.0. The last supported Windows version is Vista. In addition, those tools may not work on 64-bit versions.
Thanks for the quick response! I had already gone through those steps and still was getting incorrect latencies and errors with the "analyze recorded responses". I've looked back through the forum and it seems as if others have had a similar problem with the analyze recorded responses. Also, I'm running inquisit 4.0, is there a reason you mentioned inquisit 2.0 in your last response? Again, thanks for any advice.
You posted in the Inquisit 2.0 forum.
Ah, so sorry, I didn't even realize that.
No problem, but hence my mentioning Inquisit 2.0. What happens when you supply a list of valid responses in the 'Analyze Recorded Responses' tool prior to hitting the 'Analyze' button?
The analyze recorded response does seem to fair a little better when I include the list of valid responses (i.e., blue, red, green, yellow), but I still get the same error message. Also I believe I've solved the problem with the RT. I simply had to add /responsetime=0 (which didn't seem to be in the original script) to get it to function properly. However there are a few lingering issues.
1. The analyze recorded response is listing the responses completely backwards. In other words, it lists the last trial first and the first trial last, this makes it difficult to re-format this data file for later analyses.
2. Occasionally the entire program crashes. I've run the monkey with no problem, but when I have people run through the task it will still crash from time to time. Not sure what's going on here.
As always, any advice is greatly appreciated.
I might misunderstand, but why would this affect reformatting for analysis. If the order is indeed backwards as you say, wouldn't a simple sorting operation suffice?
Neither am I and I doubt we can figure this out right here and now. I would recommend submitting the script along with as much info about your system (hard- and software specs) and the observed conditions of the crash to inquisit<at>millisecond.com so the developers can take a closer look.