GO-noGO inhibition task. Log file RTs all the same. Correct responses appear undefined.

nomi

subject-5.xlsx

Hi all,

I would appreciate some help with my very first experiment in Open Sesame. A go-noGo, SART task.

Participants have to press SPACE to all numbers appear on screen (1-9) but inhibit their response to number "3".

The experiment seems to run with no problems but I am not sure what is wrong with the log file.

Question 1. I seem to be doing something wrong as there is no variability to Reaction Times (RTs). They are all 1000ms (as specified in Timeout keyboard response).

Question 2. At the same time the correct responses seem undefined. As far as I know:

After a correct response, the correct variable is automatically set to 1; after an incorrect response , correct is set to 0; if no correct response is specified, correct is set to 'undefined'. I have defined my variables like this:

What I get in the log file is correct responses ("1") to only the noGo trials, even after I have wrongly pressed the SPACE bar while I should not.

You can find attached one of the log files.

Any ideas please?

eduard

Hi,

Question 1

Can you share your experiment?

Question 2

You should not specify correct in the loop table, but correct_response

Edaurd

nomi

Hi Edaurd,

Many thanks for your prompt response.

Question 1 - same RTs to all responses 1000ms, as specified inTimeout in keyboard response.

Please find what I have created below. I am brand new here, so hopefully it is something very easy that I just cannot realize.

My first SART_probs with log.osexp

Question 2

Eduard: You should not specify correct in the loop table, but correct_response

Nomi: I have changed it to correct_responseon both the loop table and the keyboard response but still the same :-( . I also tried to leave the correct response empty on keyboard response.

eduard

Hi,

The problem is that you have 1150 ms of visual information during which participants can respond before the actual keyboard_response item starts running. So probably what happens is that the response all occur during the fixation dot mask or the blank 500 ms part, so that they are not registered by the keyboard response item.

To fix it, you basically have to make sure your keyboard response item starts earlier. Or you tell participants to respond later (probably easiest with a response cue, like disappearing fixdot, or so).

Normally, I would recommend the first option, but in your case that is a little tricky, because I suppose both number and mask are essential, so that 650 ms of will have to pass, before the screen can turn empty. Right?

The easiest solution would require Python scripting. If you're up for that, I can give you pointers. Another option is to use co-routines. It seems your situation is the perfect use case for co-routines (I think it makes the most sense to try that first). Finally, you can also try to a more hacky approach, which also does the trick. See attachment:

My first SART_probs with log.osexp

Btw, this problem is also the reason for the correct variable not to work properly (if you have no response collected, OS can't tell you whether it is correct or not).

Hope this clears things up.

Eduard

nomi

Hi Eduard,

I will read it thoroughly and give it a try or more! Hopefully, I will get back with good news!

Thanks!

nomi

Hello back,

I have tried the co-routines as you proposed.

My first SART_probs with log_try co-routines.osexp

It collects the answers but still RTs do not look quite right. They are actually quite fast even when I leave the number without pressing anything (keyboard duration in the co-routine is 1150ms).

I tried to find things online with regards to co-routines but still I cannot understand what is wrong.

Having checked your hack that you sent in a previous message I am afraid I don't fully understand how it works (the logic behind it). I can also see that the logger collects the RTs (I think!) but I am not sure what is going on with the correct_response columns (two keyboards = two correct_response cols).

Thanks again,

Nomi.

eduard

Hi Nomi,

You sure? I tried your experiment and I get reasonable response times (in ms)

As for my code, what it does it basically splits the response collection into two phases. One in response to the mask, and one in response to the empty screen. In order to get accurate response times, I added the inline_scripts that take the time before and after the stimulus/response. As for the correct_response, that shouldnt matter (having 2 keyboard_response items), because they have the same information in it.

nomi

OK, I will have a closer look. I so appreciate it! I need to make it more complicated in the near future so I will most probably get back :-) with a membership as well, as soon as I get a paypal!

nomi

I just wanted to say that it worked!

Thanks Eduard!

PS: I have posted more problems I encountered on my way to make it more complicated :-)

eduard

Great :)