| Reference | Downloads | Github

Unspecified JavaScript error_Once in a while

URL of experiment: workingMemoryTest [PsychoPy]

Description of the problem: I have the above experiment runs on Pavlovia since March 2021. Despite my anxiousness, the experiment has run perfectly without any problem so far among the Mac and window users.

However, last week and today, I had two participants who could not run the experiment on their browsers. :cold_sweat: Both participants had used google chrome, as instructed. One was a Mac user and the other uses window 10. After the login screen (keyed in their information), they received an error message, “Unspecified JavaScript error.” I wonder whether this is due to my poor programming skills. For your information, I mostly adapted the codes posted on forum and youtube and modified them with “trial-and-error” method as I’m a total beginner in programming.

In the experiment, I had four tasks:
(a) Typing task. In this task, the participants are asked to type the number that they saw. I did not use the textbox component, instead I used the codes that I found on Youtube.
(b) Arithmetic task. This is similar to the typing task, in which participant typed their answer after seeing the question.
(c) Audio Digit Span Task. Participants hear a series of audio file (in mp3) before typing their answers.
(d) Visual Pattern Task. Participants see a block of patterns (presented in grid) and recall the patterns by clicking on the blank grid. (I got the codes from here: How to present all the squares at the same time using the custom code?)

I have looked at these threads but I still can’t find an answer.

  1. "Unspecified Javascript Error"

Other participants (n= 84) were able to complete the experiment, just these two people.

Help in any form (e.g., direct me to the right thread or tips on where to start) will be appreciated. :pray:t3::pleading_face:

The “unspecified javascript error” message on the display isn’t very informative, but if the partiicpants opened their browser’s javascript console (cmd-option-J on chrome for mac) it would show a more specific error message. If you could replicate the failure on one of your own computers this would be easy to do. However, because it sounds like this is a rare failure, it’s hard to know unless the participants are willing to do some digging for you.

On the whole, 2/86 people running into technical difficulties isn’t something I would worry about too much, unless you identify some kind of systematic pattern in which ones run into trouble (i.e., do you not have complete data for a particular combination of conditions or something like that).