Weighted Guttman Errors: Handling Ties and Two-Level Data

We provide an introduction to weighted Guttman errors and discuss two problems in computing weighted Guttman errors that are currently not handled correctly by all software: Handling ties—that is, computing weighted Guttman errors when two items have the same estimated popularity—and computing weighted Guttman errors when the data have a two-level structure. Handling ties can be incorporated easily in existing software. For computing weighted Guttman errors for two-level data, we provide an R function.