Ranking table in OsWeb

lt_kn_char

Hey Guys,

I'm currently working on implementing a ranking table (to pre-select stimuli for a later IAT) and don't really know how to start. The plan is that participants at the beginning pf the experiment get a list of 20 words presented and then have to rank them in 2 lists (each 10 words), so that for the following experiment specific stimuli get assigend specific categories.

I read about the mousetrap plugin here and their git documentation but saw that there is no supported version for OsWeb (Mousetrap OSWeb support? — Forum (cogsci.nl)). Do you have any tip for me on how to built such a ranking table or have other ideas/ done something similar?

Thanks a lot in advance,

Leo

Fab

Hi @lt_kn_char,

Dragging objects and collecting their rearranged order is probably quite tricky in OSWeb. My guess is that it would require fairly complex Javascript programming.

Perhaps an alternative would be to use Open Sesame 4.0 that comes with OSWeb 2.0, which now supports mouse responses. You could arrange the items as a series of buttons and have the participant click on them in the order they wish. For them to visualize the order they're outputting, you could have labels appearing from left to right corresponding to the items in the order they click on them. Using a feedback object and some looping would allow you to have all this occurring dynamically. You could even add a button participants could press to undo the last response or restart their responses. Not as dynamic as dragging labels around but perhaps it may suitable to your needs. I can't think of a way to moving things around using object dragging (though a Javascript wizzard may be able to).

Note that if you'd like to keep Open Sesame 3.3 on your computer, you can actually do so while also running Open Sesame 4.0. What you'd have to do to achieve this is to download Open Sesame 4 as a zip file from here: https://github.com/open-cogsci/OpenSesame/releases

You'd then have to unzip it somewhere on your hard disk, and then launch it by searching for the opensesame.bat file and executing it. This is a way of running Open Sesame without passing through a full Windows installation. If you do so, you'll be able to run Open Sesame 3.3 as you used to, and run Open Sesame 4.0 through the .bat file.

If you like Open Sesame 4 to replace your previous version, then choose the .exe installation file.

Note that the release notes and documentation for Open Sesame 4.0 are not out yet, though some information about its new features can be found here: https://osdoc.cogsci.nl/4.0/notes/400/

Open Sesame 4 introduces new features, but is back compatible with programs written in version 3 (for now anyway). One interesting feature of OSWeb 2.0 is that it supports more features than OSWeb 1.4, such as response buttons, but it now also allows you to run the experiment locally, not just online through JATOS. In Open Sesame 4.0, the experiment is designed for the browser or for Expyriment etc. If you select the browser execution, it will run the task in the browser only and save the data file as a JSON file on your hard disk. This means that the task can now be run in the browser in the lab without requiring to export the experiment to a JATOS server (though you can still do so if you want to your participants to take part over the internet).

Hope this helps,

Fabrice.

lt_kn_char

Hey @Fab ,

Thanks for the response! I'm already using the 4.0 version because of the improvements with JS :). I had the same idea of replacing the drag & drop with simpler clicks. Though, I have some questions regarding the implementation of buttons. Before I exaggerate in my idea, I have the general question on how to implement buttons in sketchpad items and how this linking to mouse responses would work? As far as I have seen, button widgets only work with forms - which are not OSWeb supported, right? Is there already some documentation on the integration of buttons which are not used in forms?

To explain where I am right now: I thought of implementing the ranking by initializing 3 lists (all_stimuli_list, category1_list and category2_list) each with n spaces, e.g.:

let all_stim_list = [["stim1", 1], ["stim2", 2], ["stim3", 3], ["stim4", 4], ["stim5", 5], ["stim6", 6]]

let cat1_list = [[null, 1], [null, 2], [null, 3]];

-where the first column correspondent to the content and the second to the place (or rank) in the list.

And then building one button per list entry (corresponding to the ranks inside the lists). So that by clicking them, the corresponding element (content of list entry) is "selected". To then switch the selected elements when 2 buttons have been clicked. This would require that buttons can be used to perform this selection (or even one step before: can be related to specific variables/elements) and following an inline_js can be linked to the selection, in which I simply switch the elements.

Is this possible? I'm especially uncertain about the integration and linkage of buttons to certain elements.

nehakakar

Use the "Slider" component: OsWeb provides a slider component that allows participants to drag and rank items along a scale. You can utilize this component to create the ranking table.

Fab

Hi @lt_kn_char,

To use the mouse response, simply use the mouse object (instead of, say, a keyboard obect to collect keyboard responses). You can use the mouse object to detect a click on any object placed on a sketchpad or feedback object (be it text, picture etc.). You should use the region of interest method to identify what has been clicked (if you're not familiar with mouse responses, check out this tutorial: https://youtu.be/21cgX_zHDiA / It-s demonstrated with an earlier version of Open Sesame but the principle remains the same).

The idea would be to build a nested loop for the mouse responses (within the trial sequence), using a feedback object, register what subjects are clicking on, use some coding to keep monitoring it and update the feedback object accordingly). For example, you could have a row of labels with the responses, and as subjects click on them, you can display a second row illustrating the produced sequence.

It does require a little coding.

Hope this helps,

Fabrice.