An exploratory factor model for ordinal paired comparison indicators

Everything, altogether, all at once: Addressing data challenges when measuring speech intelligibility through entropy scores

When investigating unobservable, complex traits, data collection and aggregation processes can introduce distinctive features to the data such as boundedness, measurement error, clustering, outliers, and heteroscedasticity. Failure to collectively address these features can result in statistical challenges that prevent the investigation of hypotheses regarding these traits. This study aimed to demonstrate the efficacy of the Bayesian beta-proportion generalized linear latent and mixed model (beta-proportion GLLAMM) (Rabe-Hesketh et al., Psychometrika, 69(2), 167-90, 2004a, Journal of Econometrics, 128(2), 301-23, 2004c, 2004b; Skrondal and Rabe-Hesketh 2004) in handling data features when exploring research hypotheses concerning speech intelligibility. To achieve this objective, the study reexamined data from transcriptions of spontaneous speech samples initially collected by Boonen et al. (Journal of Child Language, 50(1), 78-103, 2023). The data were aggregated into entropy scores. The research compared the prediction accuracy of the beta-proportion GLLAMM with the normal linear mixed model (LMM) (Holmes et al., 2019) and investigated its capacity to estimate a latent intelligibility from entropy scores. The study also illustrated how hypotheses concerning the impact of speaker-related factors on intelligibility can be explored with the proposed model. The beta-proportion GLLAMM was not free of challenges; its implementation required formulating assumptions about the data-generating process and knowledge of probabilistic programming languages, both central to Bayesian methods. Nevertheless, results indicated the superiority of the model in predicting empirical phenomena over the normal LMM, and its ability to quantify a latent potential intelligibility. Additionally, the proposed model facilitated the exploration of hypotheses concerning speaker-related factors and intelligibility. Ultimately, this research has implications for researchers and data analysts interested in quantitatively measuring intricate, unobservable constructs while accurately predicting the empirical phenomena.

Ranking with multiple types of pairwise comparisons

The task of ranking individuals or teams, based on a set of comparisons between pairs, arises in various contexts, including sporting competitions and the analysis of dominance hierarchies among animals and humans. Given data on which competitors beat which others, the challenge is to rank the competitors from best to worst. Here we study the problem of computing rankings when there are multiple, potentially conflicting types of comparison, such as multiple types of dominance behaviours among animals. We assume that we do not know a priori what information each behaviour conveys about the ranking, or even whether they convey any information at all. Nonetheless, we show that it is possible to compute a ranking in this situation and present a fast method for doing so, based on a combination of an expectation–maximization algorithm and a modified Bradley–Terry model. We give a selection of example applications to both animal and human competition.

Bayesian paired comparison with the bpcs package

This article introduces the bpcs R package (Bayesian Paired Comparison in Stan) and the statistical models implemented in the package. This package aims to facilitate the use of Bayesian models for paired comparison data in behavioral research. Bayesian analysis of paired comparison data allows parameter estimation even in conditions where the maximum likelihood does not exist, allows easy extension of paired comparison models, provides straightforward interpretation of the results with credible intervals, has better control of type I error, has more robust evidence towards the null hypothesis, allows propagation of uncertainties, includes prior information, and performs well when handling models with many parameters and latent variables. The bpcs package provides a consistent interface for R users and several functions to evaluate the posterior distribution of all parameters to estimate the posterior distribution of any contest between items and to obtain the posterior distribution of the ranks. Three reanalyses of recent studies that used the frequentist Bradley-Terry model are presented. These reanalyses are conducted with the Bayesian models of the bpcs package, and all the code used to fit the models, generate the figures, and the tables are available in the online appendix.