help needed with the OS output

mfp49

Hi
I employed the Flanker task included with four blocks as follow:
Flankers to the left and target to the left
Flankers to the right and target to the right
Flankers to the left and target to the right
Flankers to the right and target to the right
in addition I have the trials in each of the blocks divided into nose and no noise.
I have added separate loggers for each block .
The output seems to give me a cumulative value of the Reaction Times. Is there anything wrong with the output or the program. Can I reach to the RTs for each trial by calculating difference between each RT for its previous one?
The output file can be downloaded from the following link:

https://file.town/download/i3awzftkpb380grukhnvw75mc

Thanks in advance,
Masoud

eduard

Hi Masoud,

If you want to know whether anything is wrong with your experiment, you have to upload the file so that we can check (and give more info on how it is supposed to work), otherwise it is impossible for us to say what is wrong (in case something is wrong).

I had a look at your file and all the independent variables seem to be there and just fine (but again, to be sure you have to give more detail). The opensesame output includes a response time for every item a keypress is required, i.e. all the instruction slides, actual trials, etc. In your data file there are multiple columns that could potentially be the important one. However, none of them seem to make much sense numerically. That is, I can't find a column in which the numbers look like response times usually look. So probably, something is wrong there. To be sure, please let us know which variable you used to store the response time.

Or again, share the experiment with us.

Thanks,

Eduard

mfp49

Thanks Eduard.
Please may I have your Email address then I can send the program file to you?

Peace,
Masoud

mfp49

Dear Eduad
Hi
Please see the program file in the following link:

https://file.town/download/kct55hnpdeduytqaun3tn9wj4

Masoud

eduard

Hi Masoud,

Attached the updated experiment.

The main problems were that you call each separate loop in the prepare phase of the inline_script and that you haven't use a keyboard_response item at all, to collect the responses for each trial. Furthermore, your experiment is a little convoluted with different items and sequences. I am not 100% sure what it is that you are investigating, but I simplified it quite a bit. (There is still potential for much more simplification, but I wasn't entirely sure whether I can make more changes without affecting your research question, so I didn't).

In any case, the logfile should write proper response times now (in the column response_time), so you should be able to at least look at the data.

Good luck.

Eduard

mfp49

Dear Eduard
Thanks very much for your time and attention.
For your information, the data are collected now. Is there any chance to have the response times extracted from the outputs from the older version? I have attached a sample output.

Peace,
Masoud

eduard

Hi Masoud,

I don't know how accurate it is, but you can take the difference of the variables time_new_logger_X and the corresponding time_target_X. As you have different target sketchpads and loggers per block, you have to select the correct ones.

Does that make sense?
Eduard

mfp49

Dear Eduard
Hi
Still no luck! I have tried the difference between each of the time_target S and the relevant time_logger. The difference is the same number and is about 500.
Any other advice is really appreciated.
Masoud

eduard

Hi Masoud,

Maybe, you could take then the time of the keypress on the target vs. the onset of the target, but I don't have much hope that this will do it. Even if it gave sort of reasonable values, it will most likely have quite a bit of variability in it, which would make it impossible to interpret them in a meaningful way. I'm afraid there is not much that can be done. It's a little late now, but generally, you should always test your experiment on small pilot sample of 2-3 people to see, whether you can analyse whatever you are interested in and only if you can make the pipeline work, test a full batch of participants.

Sorry,

Eduard