How to make a #theDress

Supplementary material for article Witzel & Toscani (2020) How to make a #theDress, Journal of the Optical Society of America A, 37(4), doi: https://doi.org/10.1364/JOSAA.381311. Additional sources: The code has also been uploaded to GitHub for future developments: https://github.com/christophWit/dressmaker Content: A. SURVEY - survey.pdf: A pdf-print of the google form that implemented the online survey. Text is in German; for translation see article. B. IMAGE PROCESSING AND STIMULUS PRODUCTION: - dressmaker.m: Matlab function that maps the color distribution of new images onto the major hue dimension (principal component of chromaticities) of #theDress; according to the study, the resulting images yield individual differences similar to #theDress. Run without input to see the example from Figure 1 (for this, dressmaker.m needs to be in the same folder as image.mat). - image.mat: A Matlab data file that contains the original images and masks that distinguish the objects in the images from their background. The images and their mask are used as input to dressmaker.m to create the stimulus images used in the study. - example_dressmaking.m: Matlab script that illustrates how dressmaker.m can be used with the 5 new images used in the study and provided by images.mat. Of course, the function can be used with still other images... C. DATA - data.xlsx consists of seven data sheets. The first six provide data from the lab experiment, seperatly for #theDress, the tie, jacket, egg, fish, and peep hole. The last sheet contains the data from the online survey. For details see readme.pdf.