We will conduct a pilot study in order to calibrate the color of the unexpected object. We will run subjects according to the [experiment procedure][1] in batches of approximately 25-30. We will test a black (#000000), white (#FFFFFF), light gray (#BFBFBF), and dark gray (#404040) unexpected object. We are primarily concerned with finding a color that both avoids ceiling effects from a too-salient or too-dissimilar object and floor effects that we have obtained when subjects ignore the rectangles. We will pick the color that reduces both effects, if possible; if multiple colors are candidates, we will pick the most extreme color possible and favor the dark side of the luminance spectrum.
[1]: https://osf.io/s2m9u/