Defining experiment in OS.

Unclemeat365 · April 2018

Hi there,

I am relatively new to OpenSesame, just one week of active usage, please excuse me if it sounds a bit poor .
I watched the latest beginner tutorial and few others but still can not quite get with the grips of how to define my own intended experiment from variables perpesctive.
What I want to design is a simple experiment where participants would be shown a n-number of different facial expressions on a sketchpad which they would then be asked to categorize into two different groups - 'P' (positive) and 'N' (negative) using keys Z and M e.g. Basically, I just want to collect their responses, as there are no right or wrong answers. Plus the reaction time.

I would be grateful if someone could point me please in the right direction in terms of how to process that in OpenSesame.

Thank you.

sebastiaan · April 2018

Hi,

It sounds like you want to present all the faces on one sketchpad, and then have participants rate the faces one at a time. Is that correct?

If so, exactly how would this work? I.e. would participants get feedback on their responses, and if so, how? Or do I misunderstand the paradigm?

Cheers!
Sebastiaan

Unclemeat365 · April 2018

Hi, Sebastatiaan

Thank you for your response.

Yes, that is right. I want to present all my stimuli to every participant in a random order and ask them to rate each picture (they will have only two options). The catch is that there will be 3 different conditions (20 pictures of each condition, total=60) of the pictures, two (40 pictures) that clearly correspond to the two options provided and one (20 pictures) that is ambiguous, which will be the main interest of the experiment since participants will have to rate it still (so no time limits on response, however, I still want to collect response time). The initial plan is not to provide participants with feedback since either response will be acceptable (although, accuracy of two obvious categories of pictures presented wont hurt to collect). But the main thing remains is to get data of how 20 ambigous pictures were rated and how fast overall responses were.

Is it enough to just create one variable as pictures and write all the names under from the file pool, and second variable as correct_response with 1/3 of them with key 'z', 1/3 with key 'm' and 1/3 with either 'z' or 'm' (ambigous)?

Or those 3 conditions of pictures are needed to be organized differently for such experiment?

Thank you for your help.

eduard · April 2018

Hi,

If you don't want to present feedback to the participants, I wouldn't go through the trouble of setting correct_responses to the correct value. Just make sure that each of your faces is associated with the condition (e.g. var.FaceType = 'a' or 'b' or 'c', depending whether it is clearly a, clearly b, or ambiguous), and that the response keys are uniquely associated with one class, e.g."Press z when you think it belongs to a, and 'm' if you think it belongs you b." With these three variables, you can compute accuracy later during the analysis without any problems. And your experimental code is also super straightforward.

Does that make sense?
Eduard

Howdy, Stranger!

Categories

Defining experiment in OS.

Comments

Howdy, Stranger!

Quick Links

Categories

Defining experiment in OS.

Comments