Howdy, Stranger!

It looks like you're new here. If you want to get involved, click one of these buttons!

Supported by

Correct responses not being logged

TamTam
edited July 2020 in JATOS

Hello everyone!

I've set up my own server in the cloud (DigitalOcean) so I can use JATOS to run my experiment online.

When I run the experiment on my laptop, everything runs smoothly and all the variables get logged.

However, when I try to run the experiment online (using JATOS, worker mode) many of the variables (responses) do not get logged. None of the correct_response_keyboard_response variables are present in the file.

I've tried manually adding variables into the logger item and running the experiment online and the same thing happened again.

Correct responses for each “atribut” are defined under correct_response variable in the loop table.

Thank you!

Comments

  • Hi, this is weird, let's check first:

    Are you running two different versions of JATOS on your laptop and on the server? If which is the version t works for and which doesn't?

    Best

  • Hi Tam,

    I'm a bit confused: You write 'many of the variables (responses) do not get logged'. Which mechanism do you use to 'log' this data (e.g. jatos.submitResultData or jatos.uploadResultFile)? And what is the 'logger item' you write about in one of the next sentences?

    Best,

    Kristian

  • I am a complete beginner when it comes to creating experiments, so bear with me and my poor explanations. 😕

    First of all, thank you both for replying.

    @elisa By ''running the experiment on my laptop'', I was referring to running the experiment offline, on my laptop, directly from the OpenSesame program. When I do so, all the responses get logged. I don't have locally installed JATOS. When online, I use JATOS 3.5.5.

    @kri The only mechanism I use to log the responses is using the logger at the end of the trial sequence.

    Since I had posted this thread, I've came across this https://github.com/smathot/osweb/issues/16 and I suspect having linked sketchpads in my experiment might be the problem.

    Here is the screenshot of my experimental loop:

    However, I've also came across a thread on the forum dealing with the similar problem (of not recording the ''correctness'' of the responses). The advice was to use something like this:

    and manually import the variables one wishes to log.

    That is what I did:

    And now when I run the study in the worker mode on JATOS, everything related to keyboard_response_IO gets logged how I want. Unfortunately, now the end of my experiment is ''acting up''.

    There is a response near the end of the experiment that I wish to record - keyboard_response_provjera and it doesn't get logged properly.

    response_keyboard_response_provjera column in the output file is consisting of cells filled with ''NA'' or ''none''. If I put logger after keyboard_response_provjera, the experiment freezes.

    I've imported custom variables in the logger:

    and everything gets logged just how I want it (offline, as well as on JATOS), except the last variable, response_keyboard_response_provjera. I've tried importing it as keyboard_response_provjera and I've also tried by inserting an inline_javascript and doing the same thing I've did with the IO variable (shown above), but no luck.

    I am really sorry for such lengthy post, confusing terminology and my lack of proper expression.

    Thank you once again!

    Tam

  • Ok, then this sounds like something @Daniel might be able to answer, as it's related to OSWeb, not JATOS. From the little I know, the issue is probably that you are relying on some features that work on OpenSesame will not work on OSWeb.

    Best

  • Hi @Tam

    From https://forum.cogsci.nl/uploads/084/ZLMK3VQHSFN9.png it looks like there is no more logger after keyboard_response_provjera . If you want the response to this item to be logged, you need to have a logger underneath the item in the experiment timeline.

    Buy Me A Coffee

Sign In or Register to comment.