Best-Worst Scaling

A survey methodology for efficiently collecting ranking judgments from participants over a large set of items. Instead of asking participants to rank all items at once (which becomes cognitively overwhelming beyond a handful of options), BWS presents small subsets (N-tuples, typically 4 items) and asks participants to select only the single best and single worst item from each subset. By repeating this across many different subsets that cover all items, a reliable overall ranking can be computed. BWS reduces cognitive load compared to full ranking tasks and avoids common biases in rating scales. It has been validated for use with Deaf and Hard-of-Hearing participants in accessibility research, where diverse written language literacy levels make traditional ranking tasks particularly challenging.