Reverse Correlation Project

The reverse correlation technique is a data driven study method used primarily in psychological and neurophysiological research.

Generating stimuli

The code enables a creation of noised images which are used for psychological and neurophysiological research.

How to run generating_stimuli.R Script

1. Install the relevant packages - "virdisLite" and "rcicr"

2. Call the libraries- 'virdisLite', 'virdis', 'rcicr'.

3. In order to create the pictures with the noise on them, two constants are avaliable, these should be changed depending on your experiment needs:

   • Number of trials - number of trails to conduct. Remember that each trial outputs 2 pictures, so if the number of trials is set to 100, you will get 200 images in total.
   • Base image - the image you want to add noise on. This parameter should include the path of the image. image size should be exactly 512X512 pixels.

4. Run generateStimuli2IFC function with the relevant arguments. This function will create a new folder in your workspace called 'stimuli'- inside it, you can find all the noised images and R file that will be used in the analysis step.

Classification Images

The code enables analysis of the experiment results. The code based on the file which created in generating the stimulus and additional file which includes the experiment results.

The code includes 2 types of analysis:

1. Analysis by subject - computing a classification image for each participant, using batchGenerateCI2IFC function.
2. Analysis by group - computing a single classification image where the data of all participants is lumped together, using generateCI2IFC function.

How to run classification_images.R Script

1. Call the libraries- 'virdisLite', 'virdis', 'rcicr'.

2. The relevant arguments for the analysis are:

   • rdata - Path of the R file which created during stimulus generation.
   • responsedata - csv file which contains the results of the experiment (the answers of all participants)

3. For the subject analysis run batchGenerateCI2IFC function.

4. For the group analysis run generateCI2IFC function.

The functions will create a new folder called 'cis' that includes the classification images.