This is JATOS - Just Another Tool for Online Studies.
More information about JATOS: www.jatos.org/Whats-JATOS.html
This software is licensed under the Apache 2 License.
Just Another Tool for Online Studies
Home Page: http://www.jatos.org
License: Apache License 2.0
This is JATOS - Just Another Tool for Online Studies.
More information about JATOS: www.jatos.org/Whats-JATOS.html
This software is licensed under the Apache 2 License.
From the 'Manage component' box, I clicked 'delete component' and got the two pop-up dialog boxes asking me if I was sure. And then it was waiting for the server for quite a long time, and didn't do anything. As soon as I stopped and refreshed, the component was deleted.
I can't access: http://141.14.164.217/mecharg/3/7 (internal server error)
(I was working on this until not long ago!)
The favicon works if I run the study from the mecharg but it didn't work through the sandbox
Was it a decision to have to log in as admin to create new users? Otherwise, maybe it would be better to do it the more 'standard' way, and have a 'create new user' option in the login page. Or is this because every user basically has access to study results, and should therefore be controlled?
For data protection issues, it would be great if there would be a method to automatically delete all component results when people click on something. Ideally this will be done by the mecharg (not by hand by the user) when people click on a link at the bottom of the experiment page.
I got the following error:
A problem occurred:
It's not allowed to run this study with a worker of type "Jatos".
You asked whether there was anything that MechArg/OSA/JATOS could do, to make the monitoring of confirmation codes easier in the MTurk.
Maybe something that could be done is to be able to add labels (accepted/rejected/blocked, etc) to individual study results. Then, this label together with worker ID and study confirmation could be exported in a csv file for easier control.
A button that would allow to delete all study results would be very useful. During debugging I generate a lot of results and would like to void having to go to each individual component to delete them all
Somewhere in some view (maybe worker view, or probably in study view) it would be nice to have the total duration of the experiment.
I don't know if this is a JS or a MechArg issue: When following the link directly to the study (lip-webexperiments...mttester=hello) on a safari browser, there is no error message saying that the browser is not supported.
In the MechArg (maybe changed in the OSA) if you locked the study, you can't try it. That's not great, but it's understandable. If it's on purpose we should state it.
The description field in the MTurk has a limit of 237 characters. It would be useful to limit the description in the mecharg to the same length
It would be useful to have an area, within each components view, where one could write comments. When I write the input JSON data for a component, I often want to comment it. For example, I want to say that the first 6 items of my slidelist are practice and will not be randomized. I could of xourse keep it in a separate text file, but that's suboptimal.
It would be useful to log important restrictions or details of the JSON data, so that anybody changing it can see it directly.
This was my mistakem but I wonder if there's a way to prevent these things from happening. I had two files with the same name, one of the with the html extension and the other without it. The mecharg was reading the one without the exnsion (presumably because it came first in the list, and because it was after all an html file.) It took me a while to understand what the problem was, so I wonder if it's possible to restrict the mecharg, so that it opens only files with the correct extensions.
Otherwise we just make a note on the wiki.
Would it be possible to add a direct link from the component Edit view to show the component? It's a bit annoying to have to go backto the whole component view before being able to see the changes implemented
Can you please upload the jsPsych example study so I can include a link to it in the tutorial?
In the dialog box, it listed all the results IDs I would be deleting. I couldn'T see the complete list of result IDs, because it was longer than the dialog box..
This might be on your list but I'd like to make it a priority: Having a button that allows to lock a study while running would be important, especially to prevent things like accidentally inactivating components, etc.
Where does the timestamp correspond to? To the location of the client or to that of the server?
Or none?
I just ran a test from within the mecharg and got a timestamp 2 hours later than the local time.
"2014-11-11,13:43:47"
In my local installation, I did everything with jatos. It seems to have created (and cd'ed to) a directory (within my applications folder) called mecharg-1.1
Probably shouldn'T happen!
I tried to import a study, from within the component view. (my mistake)
Jatos told me: component or study imported sucessfully. But nothing had actually happened, because I went to component-> import -> and imported a full .zip study file.
Is there any way this could be said in the error message?
I just did the following:
-Reorder components (with immediate visual feedback)
-Run the new component 1 (got an error)
-Go back one page
-Saw the 'old' component order in MechArg, until I refreshed the page
The error message in the edit component can be a bit confusing. It's a bit too focused on the path. It would be nice if it could also say something like: 'make sure that the .html extension is included'
I have two components, A(ID=14) and B(ID=23), that run in a loop. I only have component A active. Component B is inactive. After component A is finished, I get the following error message:
A problem occurred:
Component with ID 23 in study with ID 4 is not active.
Importing a single component doesn'T seem to be copying the corresponding script. (actually, it's also an export problem: exporting a single component exports only the .mac file, and not a zip file with its javascript.
you're aware of this right? I don't think it's a big problem, it's just a little bit counterintuitive because importing a whole sudy does import scripts
When I click on one of them, a dialog box pops up telling me that it can't delete the result. However, if I refresh the page I realize that it had, in fact, deleted it. Confusing.
A direct access route to the study, without having to go through the MTurk (sandbox) would be great
When I tried to run the entire study, and after the first component was done, I got the following message:
Worker 25 [my email here] never started study 3.
Is this intentional?
We need a google group. otherwise the millions of users will flood the github issues page with questoons
Not sure this is on purpose
I run the same component in two different ways: using the Show (single component) button, and activating only that component, and using the 'Run whole study' button.
When I run it using the single component, the results aren't stored permanently as the componentSate is 'RESULTDATA_POSTED' (they are overwritten when I re-run it). However, when I run it using the 'Run whole study' approach, then the results are stored, and "componentState":"FINISHED".
Could the page with the confirmation code also say: 'copy and paste the confirmation code to the MTurk'?
And ideally the one could afterwards download the text file.
It would be nice to be able to inactivate components, or at least move them flexibly to other studies. This would allow for testing the progression from one component to the next (for example in the case I am trying to loop back and forth between only two components) without really needing to test the other ones. Of course, to try something out, I can bring the two components of interest to the first positions in the study. But still: inactivation would be great
This is again where I'm looping through two components. I ran the first round of component A. Component B is inactive. Yesterday night, I had a result on the MechArg. Today It's gone, and replaced by another result, from this morning. In the result section, "componentState":"RESULTDATA_POSTED",
So maybe this is correct, because the component isn't really finished. But it's weird to have results disappearing. But I don't know if it's correct or a bug
I really like the path-like links, like this:
Home/Thought content/NYCQ_QuestionContent/Results/Index
However there's something wrong. For example here i'm in this url
https://lip-webexperiments.mpib-berlin.mpg.de/mecharg/1/12/results
That corresponds to the results of component 12. In the path above, 'Results' is clickable, although we are already in the result view, and the link to Results leads to the same page.
I have three independent components (in my mindwandering study). Each of them does the following ajax call:
var ajax = $.ajax({
url: "/publix/" + studyId + "/" + componentId + "/submitResultData",
data: componentResultJson,
processData: false,
type: "POST",
contentType: "application/json",
success: function(url) {
window.location.href = "/publix/" + studyId + "/startNextComponent";
},
error: function(err) {
currentComponentState = componentState.RUNNING;
showError(err.responseText);
}
});
When I activate only those three components and click on 'Run Whole Study', I get a (seeminly) infinite loop between those three active components. And (I think) only one result is saved. (i.e., results are saved and overwritten)
I know that "good" user management is an Extrawurst, but one user should not only not be able toedit some other user's studies: they should not be able to see them, too.
A declarative, efficient, and flexible JavaScript library for building user interfaces.
๐ Vue.js is a progressive, incrementally-adoptable JavaScript framework for building UI on the web.
TypeScript is a superset of JavaScript that compiles to clean JavaScript output.
An Open Source Machine Learning Framework for Everyone
The Web framework for perfectionists with deadlines.
A PHP framework for web artisans
Bring data to life with SVG, Canvas and HTML. ๐๐๐
JavaScript (JS) is a lightweight interpreted programming language with first-class functions.
Some thing interesting about web. New door for the world.
A server is a program made to process requests and deliver data to clients.
Machine learning is a way of modeling and interpreting data that allows a piece of software to respond intelligently.
Some thing interesting about visualization, use data art
Some thing interesting about game, make everyone happy.
We are working to build community through open source technology. NB: members must have two-factor auth.
Open source projects and samples from Microsoft.
Google โค๏ธ Open Source for everyone.
Alibaba Open Source for everyone
Data-Driven Documents codes.
China tencent open source team.