Using a split-ballot design to validate an abbreviated categorical measurement scale: An illustration using the Transportation Security Index

To address the high survey costs and increased respondent burden that comes with administering composite multi-item scales, researchers frequently seek to develop and use abbreviated scales. To help them do so, methodologists have issued a series of guidelines outlining best practices for shortening scales. However, it is difficult to find an empirical illustration of both the design and validation of an abbreviated scale, particularly one for which the classification of respondents into distinct categories is of paramount importance. In this paper, we present such an illustration using the Transportation Security Index (TSI) as a motivating example. Notably, we employ a split-ballot experiment to validate the TSI-6, a six-item abbreviated scale that successfully reproduces the original, validated TSI-16. We also illustrate the implementation of several agreed upon best practices in abbreviated scale development and propose and demonstrate specific steps that are uniquely relevant to the validation of a categorical abbreviated measure.