Last seen/Duration problem - stuck?

Comments

• kri

  July 2020

  Hi Hülya,

  It's strange that the study results do not show state 'ABORTED' after you aborted the study. May I ask how you do the cancellation? How does the code look like?

  And about JATOS' components. If you have only one component they don't really make sense and it's best to just ignore the component results and just look at the study results. Components and their results make more sense if you have several of them in a study, e.g. first one does the introduction and consent, second one a survey, third one the actual experiment. Then you can see in the component results only all data e.g. from the survey.

  That said, it is strange that you have 48 study result and only 28 component result. I'd expect at least the same amount or higher of component results there (I guess this is why you made the red circles in the screenshot). In the case where you have a study result but no component result, what happened is, that your study run was started and a study result created - but then something went wrong immediately. One thing that comes to my mind is that your component's HTML file wasn't correctly named or didn't exist. Was this maybe the case here? Another idea: What does JATOS log say? You can find it under your-jatos-address/jatos/log.

  Best,

  Kristian
• hulya

  July 2020

  Hi Kristian,

  Thank you for your fast answer.

  Umm... I am not sure how I do the cancellation... I mean I didn't put any option for that, it's just ESC like default I guess. But maybe those numbers are caused by this: it is not the real experiment yet so I was playing around with the links and trying out how it looks. I mostly just closed the browser after making a few trials and that's it. Do yout hink that might be the problem?

  Unfortunately I am not allowed to access the log, so I can't share it with you...

  Also I noticed that after a few trials we started to get "JATOS isn't allowed to run a study with the study result ID ..." . But I don't understand this either because at first, we are able to run the study with a General Multiple worker link. And then after a while, it just stops working and gives this error. Is it because there is an expiration in the links or a maximum number of using it?

  Cheers~

  H
• kri

  July 2020

  Hi Hülya,

  Umm... I am not sure how I do the cancellation... I mean I didn't put any option for that, it's just ESC like default I guess. But maybe those numbers are caused by this: it is not the real experiment yet so I was playing around with the links and trying out how it looks. I mostly just closed the browser after making a few trials and that's it. Do yout hink that might be the problem?

  Okay, then everything is fine and behaves as it should be. The state DATA_RETRIEVED is correct. JATOS only goes into state ABORTED if you tell it so via jatos.js functions, e.g. jatos.abortStudy or jatos.addAbortButton . The abort function gives the worker the possibility to cancel the study at any point during the study run and delete all previously stored result data. But if you don't use it and the study gets abandoned then it just stays at the last state, most likely DATA_RETRIEVED. And then if you want to finish a study you'd call jatos.endStudy or a similar function. You can find more about jatos.js functions in it's reference.

  Also I noticed that after a few trials we started to get "JATOS isn't allowed to run a study with the study result ID ..." . But I don't understand this either because at first, we are able to run the study with a General Multiple worker link. And then after a while, it just stops working and gives this error. Is it because there is an expiration in the links or a maximum number of using it?

  I guess, what happened here is, that you mixed up the study run link with the study's component URLs. The former one is the one you hand out to your workers. After starting a study JATOS automatically redirects you to the latter one. In the case of a General Multiple study run link you can use it as often as you want. But then those study's component URLs you get redirected too work only as long as the study is not finished (and other constraints).

  Best,

  Kristian
• kri

  July 2020

  Maybe an example illustrates the difference better:

  • a General Multiple study run link: https://experiments.jatos.org/publix/1/start?batchId=1&generalMultiple
  • an study's component URL: https://experiments.jatos.org/publix/1/1/start?srid=1819

  The latter one is restricted.
• hulya

  July 2020

  Hi Kristian,

  Thank you a bunch for the clear explanation!! That solved a looooot of my problems :)

  Best,

  H
• hulya

  September 2020

  Hi @kri ,

  Sorry to bother you again, but any idea why this might be happening?

  

  
  

  I am just worried if I am missing any data or this is interfering with anything...

  Best,

  H
• elisa

  September 2020

  Hi Hulya,

  This is a completely different question from your questions above, right? (Just to be sure I understand). In those cases it's best to open a new thread.

  Without more information, it's difficult to know if the problem is on the server side (JATOS) or client (OSWeb). So, a few questions:

  -Did the same experiment run well before, and now doesn't? If so, which version of JATOS are you running now? On which server? (you're not the first one to ask about this so we should figure it out)

  - Are your data all there despite the 'not finished yet'?

  Best

  Elisa
• hulya

  September 2020

  Hi @elisa ,

  I thought the question was related since I asked before about durations but yes, slightly different.

  So, yes, the experiment run well before, and actually it still runs when someone does and finished, I have another result ID. BTW, this one I shared is still going on!

  It is version 3.5.7 and it's an institutional server, so I am not sure if I have any more information about this since I don't have access to anything else.

  In this one, there is no data. There is just "[" .And it is very weird, because for example, now it is last seen at 6:16 a.m today. At first I thought maybe someone clicked on my link and left it open aside, and put the computer asleep. But I also opened OpenSesame to look at something, regardless of this study. And its start time corresponds to the time I opened OS.

  Best,

  H
• elisa

  September 2020

  And what happens if you run the study yourself? Does it work and can you collect and see the data?
• kri

  September 2020

  Hi Hulya!

  This 'not finished yet' does not necessarily mean that your study run was not successful. Internally JATOS only counts a study as 'finished' if a certain API call (jatos.endStudy or similar ones) took place. Only then switches the state from 'DATA_RETRIEVED' to 'FINISHED' and duration gets a final time and date. As long as this API call didn't happen JATOS has no chance to officially end the study run. But this does not necessarily mean that your study run didn't do what it was supposed to be doing - including sending the result data. Maybe your participant just closed the browser tab, where the study was running, shortly before the jatos.endStudy call could have been sent, but apart from that did the whole study. In the end the important thing is: Did you get all your result data?

  Best,

  Kristian
• hulya

  September 2020

  Hi @elisa ,

  Yes, it works without a problem. And it works with many other participants too, still, I can also see the data. But this specific instance I don't know what is still there...

  Hi @kri ,

  The thing is, there is no participant in this instance. The ones who did it/who does it are fine, and we are contacting them to pay so we know that who did it properly or not. So there is no result data either, as I said, there is just "[". That's why it is confusing, but maybe it's that as you say: they closed the browser tab but they just clicked on the link, did nothing and directly closed it before the call was sent. I think for me that's the most logical explanation because there is no participant that we have missed or reported that they did the experiment etc.

  Best,

  Hülya
• hulya

  September 2020

  Just to visualize the current situation:

  

  
  
• kri

  September 2020

  This is just one, isn't it? With just one participant I'd say it's like you said: 'they ... did nothing and directly closed it before the call was sent.'.