[open] Dissertation: One-Shot Change Detection Task
Hello all,
I am hoping someone out there can help me as I am close to breaking point over my dissertation. Long story short - I have basically been given a dissertation topic/experiment because I wasn't allowed to do what I wanted and told to use OpenSesame for it despite never using it before or any of my lecturers having a clue how to use it either...
My experiment adopts a one-shot change detection analysis method that will measure accuracy and reaction time, consisting of four trial phases (20 trials in each phase). In each trial phase four images will appear (I have already grouped these in hopes of making life easier), a mask will appear, and the corresponding images (one of the four images would have changed) will then appear until a response is made. I will be using the Z key to indicate change and, M for no change.
- angry face is present - mask - angry face changes
- angry face is present - mask - other face changes
- angry face is present - mask - no change
- no angry face is present - mask - face changes
Therefore, although I want these to appear in a random order I still need the corresponding 'before and after' images to appear in the correct sequential order.
I read somewhere that I could probably do this by "having four sequences of the sketchpads and the sequence of sketchpads that are displayed is determined by a global variable that is set at random" but, this is a foreign language to me. Do any of you have a really simple tutorial to follow or able to completely dumb this down for me, please?
This is what I had done so far, following the tutorial but I am guessing this is wrong...
Amy 
Comments
Hi Amy,
Your experiment shouldn't be too difficult to implement and by the looks of the screenshot you have come a long way already.
It's difficult to determine at which point the in experiment you are stranded, is there any particular part you need help with?
Laurent
Hi Laurent,
Well, all of it really. I think what I have so far is incorrect. I have no idea how to put all of my images into the programme and when I do I don't think it will run correctly as I need them in a random order (in terms of the type of trial) but for the images to appear in order in terms of their correspondence.
For example:
a mask, and then...
As these are the corresponding images but... I have 160 images that I need to use.
That probably makes no sense.
Okay, let's start with getting all pictures into a file pool as explained here in the tutorial.
Once all the files are in the pool you can add your variables into the
loopitem. By the looks of it you'd need four variables: first image, second image, trial type and correct answer.Next, place your image files into the loop under the correct variable you just created and fill out the trial type and the correct answer. Now you have a loop with all the desired corresponding picutures that you can present with the sketchpad. Don't forget to set the order to random.
Hey, thanks!
So I have changed my experiment around so that it is actually set up to do what I want (hopefully) but, as you just mentioned - I'm having a bit of a moment with adding the images. I have all my images in the file pool but, for each of my variables I seem to have to same image - I will attach screenshots below...
How do I get each separate variable to have a different image? It would help if I knew the technical terms, sorry. So underneath my "first_image" variable I want it to essentially 'read'...
1. 1.1.1.jpg
2. 1.2.1.jpg
3. 1.3.1.jpg etc.
Hi Amy,
You can simply change the name to the filename by doubleclicking on the area. I can really recommend working your way trough the tutorial again, afterwards everything will make much more sense
Two other notes: If your mask is the same image in every trial, you don't need to create a specific variable for it (because it doesn't change between trials, it is no variable). You can just present it on the sketchpad.
Also, and i'm not sure if this is your intention, you have four different loops for four different trial types. So what happens now is that the experiment starts by looping through the first trial type (angry_angry_loop) and it presents 20 trials of this type in a random order. Then it moves to the next loop (angry_other_loop) and presents 20 of these trials, etc. If that is the correct order, this is perfect. Otherwise you might want to combine everything into one loop and create a new variable that contains the trial type.
Good luck!