From the 1940s to 2020s: A review of the current state of forced-choice methodology

Forced-choice measures are an alternative to rating-scale surveys designed to reduce response bias, particularly socially desirable responding, by requiring respondents to make rank-order comparisons among two or more statements at a time. Although forced-choice instruments have been used in psychological testing since at least the 1940s, recent methodological advances in item response theory modeling have enabled the estimation of normative scores from the raw ipsative data these assessments produce. The introduction of new scoring methods has resulted in an uptick in the use of forced-choice tests, as full cross-person comparisons were made possible. This paper chronicles the historical development of forced-choice instruments up to the pivotal introduction of item response models for scoring and uses that foundation to review contemporary methods for their construction and analysis. Our review of modern-day methods begins by examining approaches to constructing forced-choice blocks, including the use of mean indices, interitem agreement coefficients, and factor loadings. We then discuss the ideal-point and dominance-based item response models used to evaluate the internal structure of forced-choice assessments and compute scores, as well as methods for assessing differential item functioning. Throughout the review, we also synthesize literature on evaluating response processes, reliability, and other considerations in test construction. Finally, we discuss ongoing debates regarding the extent to which forced-choice measures effectively limit response bias, particularly when negatively keyed items are included in blocks, and conclude by outlining directions for future research. To support engagement with the historical literature, we provide an annotated bibliography spanning more than 8 decades of forced-choice research. (PsycInfo Database Record (c) 2026 APA, all rights reserved).