Showing specific Keys in data output

Tpatrick

Hi,

My experiment currently asks participants to press one of two keys (L; E) during an experimental trial. Both responses are correct/incorrect because its a perceptual task. I don't want to designate one response as right vs. wrong, what I want is the number of times each key was pressed and I can't seem to find a way to identify which key was pressed on which trial in the excel output.

Any help is appreciated!

eduard

Hi Patrick,

Please share your experiment, and one of the output files. If you have a keyboard response item and a logger in your experiment (used correctly), you should see which button was pressed in the column response with one value per row/trial.

Eduard

Tpatrick

subject-5557575.csv

subject-99999.csv

Tpatrick

I am having a hard time copying my experiment. It keeps giving me an error code 413. Ill keep trying. In the meantime my experiment has a New Keyboard Response after the stimulus with L;E chosen as allowed responses. Nothing listed as correct responses. Then the next thing is a Data logger with the box checked for all variables. In the output as youll see, I don't see what keys were pressed.

Tpatrick

Fab

Hi @Tpatrick,

The error code you get trying to upload your task is most probaby because it is too large to be uploaded. Could you upload a stripped down verison of it?

Best,

Fabrice.

Tpatrick

RichExperimentabbreviated.osexp

Fab

Hi @Tpatrick,

Thanks for uploading a stripped down version of your task. I've just taken a look and spotted some issues:

(1) This does not relate directly to the problem you highlighted but I should mention it: I infer that your complete task includes lots of loops with duplicated trial sequences and objects. If the structure od your task is the same in all loops and only the contents of the loop change, you should avoid creating lots of copies of the same objects and sequences and, instead, used linked copies. Also note that using lots of data logs may create problems too. So, I'd recommend using lnked copies of objects instead of creating 7 or more versions. It also has the advanage of reducing the size of the experiment and the code, nd helps the task run smoother.

(2) One reason why you response collection is not working is because your keyboard even has a duration of 0 ms. It is therefore not possible for the task to register the participant's response because there is not time for that. The keyboard event stops right when it starts. Therefore, the response is always gonna be "none". If you insert some Javascript inline (you could do this in Python too of course) to output the response to the console, you'll see that it outpus "None", which shows you that no response has been collected. For the response to collected, the duration of the keyboard object should be set to "infinite" (such that the keyboard object keeps running untul a response has been registered). But wait... See the next point.

(3) You may ask: But how can it be that no response was registered if I actually pressed the key and the task reacted? Well, that is because what your task does is to collect two responses. You set the duration of the media object to "keypress", which means that a response is taken to stop the video. When that happens, your task moves to the keyboatfr object, but since you set its duration to 0, the response is never collected. The result is that the "response" variable (which is the same as the response_new_keyboard_response_7 variable) is going to be "None" and show empy in your data log. Note that the response variable always contains the latest response registered, but that you can access each response through the direct variable relating to specific objects. So, the variable response_New_Media_plyer_mpy_1 will contain the response to the media object, the variable response_new_keyboard_response_7 will contain the response to the keyboatf object). Now, since you're collecting a response to the media object, there is actually no need to run a keyboard object afterwers. The participant's response is actually collected by the media object. Hence you should delete the keyboard object.

I attach your example amended as described and outouting the response to the console so that tyou can see that it is now registered.

Hope this helps,

Fabrice.

RichExperimentabbreviated Fab.osexp

Tpatrick

This works great. Thanks for all your help Fab!

Fab

Hi @Tpatrick,

Glad I could help! Good luck with your experiment!

Best,

Fabrice.