This experiment's preregistration can be found [here.][1] While previous experiments indicated that people are capable of forming broad, non-specific attention sets to capture members of a highly variable category, these previous experiments involved pitting colorful shapes against white shapes. Thus, it remains a possibility that our stimuli actually happened to fall within pre-existing, natural categories (such as "chromatic" and "achromatic"), and thus were not formed due to the demands of the experiment, or that there is something special about attending to and ignoring white. In order to more robustly verify that it is the variation within the category, and not any features of any objects in the display, that drives our effect, we are replicating [our previous experiment][2] with slight modification. We are replacing white in the display with pink, and rotating out which color remains constant between trials, and which colors fall into the varying set. If we replicate the results we found in experiment 2, finding that people form a set for the constant color and force all other colors into a "non-pink" or "non-yellow" set, it provides additional evidence for broad, flexible categories irrespective of the absolute features of the objects within them. [1]: https://osf.io/72kb6/?view_only=f2eba4bb23444c24bde8d1aef68cd29d [2]: https://osf.io/twd2j/