Human-likeness experiment (uncanny valley)
in OpenSesame
Hi there,
I'm trying to create an experiment that consists on judging the human-likeness of a picture with a name underneath it (trying to see if different types of names change our attributions of humanity given to robots and androids). Pictures should be selected at random from two categories: android pictures and human pictures. The same picture cannot appear more than once. At the same time a name should show under the picture. The name should also be selected at random from three categories: human name, non-human name (like that of a pet), and code name. Like before, the same name cannot appear more than once.
I would sincerely thank any help as I'm just starting out with Opensesame.
Javier
Comments
Hi Javier,
Here an example of complete randomization the way you describe it. If you have more restrictions it will become slightly more complicated, but this should get you started.
Hope it helps!
Eduard
ps.
I would sincerely thank any help as I'm just starting out with Opensesame.
In this case, I would recommend you have a look at the tutorials (both Openseame as well as Python), see the tutorial section in the documentation. Investing time to understand Opensesame now, will save you loads of time (and headaches) later!
Hi Eduard,
Thanks for the reply, my experiment looks a bit different. This is the loop preview.
I don't know if it's clear enough in the image but briefly there are 2 categories (meca and hum) in variable "fis", and for each one there's 12 images ("fis_nr"). For the variable "n" (name) there are 3 categories (human, non-human, code) and 8 names for each category ("n_nr"). The experiment runs but i need the images to only show once each because as you can see in the preview some show up twice. I think I should use a constrain command like maxrep=1 but i'm probably not using it right as it's not working properly. What I tried to do was open the loop item script and after the setcycle commands i wrote "constrain fis_nr maxrep=1", but the loop preview keeps showing that some images will repeat.
Let me know if I'm not explaining myself correctly,
Javier
Hi Javier,
I added the restriction to my example. Other than that, I don't see why your situation should not work with that example.
Yes, the maxrep should also do the trick, so I guess something doesn't work out there correctly. If you share your experiment I can have a look.
Eduard
File is too big to be sent through here, would you mind giving me your email adress?
PS: my approach is quite different to yours, I've used a full-factorial design like an opensesame' video tutorial called "cats, dogs and capybaras" and maybe I shouldn't have, I don't know. Once I can send you the experiment you tell me.
Javier
Hi Javier,
you case use wetransfer.com to upload the experiment and share it here.
my approach is quite different to yours,
That is okay. Generally, it makes a lot of sense to use the full-factorial design. Under some circumstances, (with certain more complex randomization restrictions), the loop object is reaching its limits. For your particularly case, I can't tell whether this is the case (i.e. whether it should work with your approach). Personally, i prefer a somewhat more Python-based approach, that offers more flexibility. But oftentimes, whichever approach you use, is a matter of personal preference (but again, whether this applies to your situation, I can't tell yet)
Eduard
Here it is, and thanks for your help again. As you can see images and names repeat themselves once twice or even 3 times.
Javier
Hi Javier,
As I said, I am not sure what is happening in your case, the way you do it (splitting name and number) makes loop restrictions quite tricky. I attach your experiment rewritten to work with Python loops. You need to specify how many trials you want, and maybe adjust the layout a bit, but for the rest it does what you need, right?
Eduard
https://we.tl/t-rcwIYjwbRc
Hi Eduard,
I added the restriction to my example.
This first example you sent me should work too, right? Just need to change the lists and add the names of the files. I think I can see the benefits of Python so maybe I should invest some time into it.
Thank you,
Javier