[open] Randomisation of image selection in sketchpad

ha006

Hi there, I believe i need someones help!
I have googled it and there seems to be no simple answer.

I have an experiment, there are 8 conditions, each condition needs to show 1 image, chosen at random from a folder of similar images. Each trial needs to be different (so each participant sits a differently ordered set of conditions) AND each condition's selected image will be different.

So; condition1(one image out of possible 140), condition2(one image out of possible 90), condition3...etc

Recording data is not a problem as this can be done by hand, but I just need the images to be randomly selected from a pre-defined group.

I have tried using code to randomise and shuffle the order, but have got nowhere. Please help, Tom

lvanderlinden

Hi Tom,

I'm afraid I don't completely understand what kind of trial order you are aiming for. Are you looking for a Latin-Square-like design, such that you do not only make sure that (1) a given picture will only occur in one condition, but (2) also that across participants each picture will occur equally often in each condition?

Could you provide us with somewhat more detail?

Best,

Lotje

leaena

Hi Lotje,

I'm a newbie here and trying to find a good solution for my lab's experiment problem. However I am unable to find a definitive answer so far...is it possible to do latin square ordering in Open Sesame? Or would I have to generate separate orders for each set and put them all in separate blocks?

Thanks for any info.

Edwin

Welcome to the forum!

OpenSesame supports random randomization and ordered presentation. For all other kinds of ordering trials (pseudo-random constructions or Latin square), it is advised to create a trial order outside of OpenSesame and to copy that to a loop.

@sebastiaan, maybe this is a good idea for an additional function? I've opened an issue on GitHub.

Good luck!

Edwin

sebastiaan

Yes, that would be a nice addition. Not just for Latin-square designs, but generally a more comprehensive way to implement constraints on pseudo-randomized trial ordering.

For now though, as Edwin says, it'll be a copy-paste job!