Meta-analysis is a widely used statistical tool for estimating the diagnostic accuracy of tests across multiple studies. Existing methods and available R packages primarily focus on a single diagnostic test, typically under the assumption that all studies include a gold standard. Greater efficiency can be achieved by modeling multiple diagnostic tests together and drawing on studies with or without a gold standard reference test across diverse designs. To address this challenge, recent work has extended both the Bayesian hierarchical model and the Bayesian hierarchical summary receiver operating characteristic model to the framework of network meta-analysis of diagnostic tests, enabling simultaneous comparison of multiple tests when some data are missing. Despite the importance of these methods, their computational complexity has limited their broad application. This article introduces NMADTA, an R package that implements these models with user-friendly functions. The package allows researchers to evaluate the accuracy of multiple diagnostic tests simultaneously and provides comprehensive graphical displays of the results.

Multilevel regression and poststratification (MRP) is widely used to estimate opinion in small subgroups and to adjust unrepresentative surveys. Yet, even flexible MRP models contain errors generated by non-response and model misspecification. We propose a principled, data-driven method to leverage observable errors on auxiliary quantities with known marginal distributions—for example, election outcomes—to improve estimates of policy attitudes. Our method leverages the correlation between auxiliary variables and outcomes of interest to calibrate MRP estimates to these known marginal distributions. We illustrate our approach using a pre-election poll measuring support for an abortion referendum. We find that the method reduces county-level error by nearly two-thirds relative to traditional MRP. We also show how our calibration approach can be used to generate estimates for smaller nested geographies, such as precincts, even in the absence of poststratification data at this level. Our approach provides a framework for fully incorporating known population data to improve estimates of public opinion in small subgroups, providing scholars another tool to study representation.

Computerized adaptive tests (CATs) play a crucial role in educational assessment and diagnostic screening in behavioral health. Unlike traditional linear tests that administer a fixed set of pre-assembled items, CATs adaptively tailor the test to an examinee’s latent trait level based on their previous responses. We introduce a novel CAT system that builds on recent advances in Bayesian multivariate IRT. Our approach leverages direct sampling from the latent factor posterior distributions, significantly accelerating existing information-theoretic item-selection methods by eliminating the need for computationally intensive Markov chain Monte Carlo simulations. To address the potential suboptimality of one-step-ahead item-selection rules, we also develop a double deep Q-learning algorithm that efficiently learns an optimal item-selection policy offline using a calibrated item bank. Through simulation and real-data studies, we demonstrate that our approach not only accelerates existing item-selection methods but also highlights the potential of reinforcement learning (RL) in CATs. Notably, our Q-learning-based strategy consistently achieves the fastest posterior variance reduction, leading to earlier test termination. These results demonstrate the promise of combining exact posterior sampling with RL to deliver scalable, high-precision CATs.

The emergence of large language models has significantly expanded the use of natural language processing (NLP), even as it has heightened exposure to adversarial threats. We present an overview of adversarial NLP with an emphasis on challenges, policy implications, emerging areas, and future directions. First, we review attack methods and evaluate the vulnerabilities of popular NLP models. Then, we review defense strategies that include adversarial training. We describe major policy implications, identify key trends, and suggest future directions, such as the use of Bayesian methods to improve the security and robustness of NLP systems.

This article examines how subnational fiscal competition over foreign direct investment affects both the siting of new projects and the ability of local governments to raise tax revenue for social spending. We leverage a quasi-natural experiment, an unexpected declaration by the Brazilian Supreme Court in 2017 that reduced states’ ability to offer investors differentiated tax subsidies. Our results show that disadvantaged regions did not see a major shift in investment patterns after the change in investment law. We do not find a consistent relationship between the incentive law change and state revenue generation, but we do find that incentives are associated with less revenue. The results are consistent with arguments that investment incentives exacerbate inequality by reducing states’ capacity to collect revenue while doing little to affect investment location. Our results illustrate that economic agglomeration is difficult to reverse through tax policy and that fiscal federalism often cannot provide strong enough inducements to drive investment into less advantaged regions.

In small meta-analyses (e.g., up to 20 studies), the best-performing frequentist methods can yield very wide confidence intervals for the meta-analytic mean, as well as biased and imprecise estimates of the heterogeneity. We investigate the frequentist performance of alternative Bayesian methods that use the invariant Jeffreys prior. This prior has the usual Bayesian motivation, but also has a purely frequentist motivation: the resulting posterior modes correspond to the established Firth bias correction of the maximum likelihood estimator. We consider two forms of the Jeffreys prior for random-effects meta-analysis: the previously established “Jeffreys1” prior treats the heterogeneity as a nuisance parameter, whereas the “Jeffreys2” prior treats both the mean and the heterogeneity as estimands of interest. In a large simulation study, we assess the performance of both Jeffreys priors, considering different types of Bayesian estimates and intervals. We assess point and interval estimation for both the mean and the heterogeneity parameters, comparing to the best-performing frequentist methods. For small meta-analyses of binary outcomes, the Jeffreys2 prior may offer advantages over standard frequentist methods for point and interval estimation of the mean parameter. In these cases, Jeffreys2 can substantially improve efficiency while more often showing nominal frequentist coverage. However, for small meta-analyses of continuous outcomes, standard frequentist methods seem to remain the best choices. The best-performing method for estimating the heterogeneity varied according to the heterogeneity itself. Röver & Friede’s R package bayesmeta implements both Jeffreys priors. We also generalize the Jeffreys2 prior to the case of meta-regression.

In the study of human dynamics, the behavior under study is often operationalized by tallying the frequencies and intensities of a collection of lower-order processes. For instance, the higher-order construct of negative affect may be indicated by the occurrence of crying, frowning, and other verbal and nonverbal expressions of distress, fear, anger, and other negative feelings. However, because of idiosyncratic differences in how negative affect is expressed, some of the lower-order processes may be characterized by sparse occurrences in some individuals. To aid the recovery of the true dynamics of a system in cases where there may be an inflation of such “zero responses,” we propose adding a regime (unobserved phase) of “non-occurrence” to a bivariate Ornstein–Uhlenbeck (OU) model to account for the high instances of non-occurrence in some individuals while simultaneously allowing for multivariate dynamic representation of the processes of interest under nonzero responses. The transition between the occurrence (i.e., active) and non-occurrence (i.e., inactive) regimes is represented using a novel latent Markovian transition model with dependencies on latent variables and person-specific covariates to account for inter-individual heterogeneity of the processes. Bayesian estimation and inference are based on Markov chain Monte Carlo algorithms implemented using the JAGS software. We demonstrate the utility of the proposed zero-inflated regime-switching OU model to a study of young children’s self-regulation at 36 and 48 months.

Identifying racial disparities in policy and politics is a pressing area of research within the United States. Where early work made use of identifying potentially noisy correlations between county or precinct demographics and election outcomes, the advent of Bayesian Improved Surname Geocoding (BISG) vastly improved estimation of race by employing voter lists. Machine Learning (ML)-modified BISG in turn offers accuracy gains over the static – and potentially outdated – surname dictionaries present in traditional BISG. However, the extent to which ML might substantively alter the policy and political implications of redistricting is unclear given its improvements in voter race estimation. Therefore, we ascertain the potential gains of ML-modified BISG in improving the estimation of race for the purpose of redistricting majority-minority districts. We evaluate an ML-modified BISG program against traditional BISG estimates in correctly estimating the race of voters for creating majority-minority congressional districts within North Carolina and Georgia, and in state assembly districts in Wisconsin. Our results demonstrate that ML-modified BISG offers substantive gains over traditional BISG, especially in diverse political geographic units. Further, we find meaningful improvements in accuracy when estimating majority-minority district racial composition. We conclude with recommendations on when and how to use the two methods, in addition how to ensure transparency and confidence in BISG-related research.

Bayesian inference provides a probabilistic reasoning process for drawing conclusions based on imprecise and uncertain data that has been successful in many applications within robotics and information processing, but is most often considered in terms of data analysis rather than synthesis of behaviours. This paper presents the use of Bayesian inference as a means by which to perform Boolean operations in a logic programme while incorporating and propagating uncertainty information through logic operations by inference. Boolean logic operations are implemented in a Bayesian network of Bernoulli random variables with tensor-based discrete distributions to enable probabilistic hybrid logic programming of a robot. This enables Bayesian inference operations to coexist with Boolean logic in a unified system while retaining the ability to capture uncertainty by means of discrete probability distributions. Using a discrete Bayesian network with both Boolean and Bayesian elements, the proposed methodology is applied to navigate a mobile robot using hybrid Bayesian and Boolean operations to illustrate how this new approach improves robotic performance by inclusion of uncertainty without increasing the number of logic elements required. As any logical system could be programmed in this manner to integrate uncertainty into decision-making, this methodology can benefit a wide range of applications that use discrete or probabilistic logic.

The dynamics and fusion of vesicles during the last steps of exocytosis are not well established yet in cell biology. An open issue is the characterization of the diffusion process at the plasma membrane. Total internal reflection fluorescence microscopy (TIRFM) has been successfully used to analyze the coordination of proteins involved in this mechanism. It enables to capture dynamics of proteins with high frame rate and reasonable signal-to-noise values. Nevertheless, methodological approaches that can analyze and estimate diffusion in local small areas at the scale of a single diffusing spot within cells, are still lacking. To address this issue, we propose a novel correlation-based method for local diffusion estimation. As a starting point, we consider Fick’s second law of diffusion that relates the diffusive flux to the gradient of the concentration. Then, we derive an explicit parametric model which is further fitted to time-correlation signals computed from regions of interest (ROI) containing individual spots. Our modeling and Bayesian estimation framework are well appropriate to represent isolated diffusion events and are robust to noise, ROI sizes, and localization of spots in ROIs. The performance of BayesTICS is shown on both synthetic and real TIRFM images depicting Transferrin Receptor proteins.

Birnbaum and Quispe-Torreblanca (2018) evaluated a set of six models developed under true-and-error theory against data in which people made choices in repeated gambles. They concluded the three models based on expected utility theory were inadequate accounts of the behavioral data, and argued in favor of the simplest of the remaining three more general models. To reach these conclusions, they used non-Bayesian statistical methods: frequentist point estimation of parameters, bootstrapped confidence intervals of parameters, and null hypothesis significance testing of models. We address the same research goals, based on the same models and the same data, using Bayesian methods. We implement the models as graphical models in JAGS to allow for computational Bayesian analysis. Our results are based on posterior distribution of parameters, posterior predictive checks of descriptive adequacy, and Bayes factors for model comparison. We compare the Bayesian results with those of Birnbaum and Quispe-Torreblanca (2018). We conclude that, while the very general conclusions of the two approaches agree, the Bayesian approach offers better detailed answers, especially for the key question of the evidence the data provide for and against the competing models. Finally, we discuss the conceptual and practical advantages of using Bayesian methods in judgment and decision making research highlighted by this case study.

Scholars often use language to proxy ethnic identity in studies of conflict and separatism. This conflation of language and ethnicity is misleading: language can cut across ethnic divides and itself has a strong link to identity and social mobility. Language can therefore influence political preferences independently of ethnicity. Results from an original survey of two post-Soviet regions support these claims. Statistical analyses demonstrate that individuals fluent in a peripheral lingua franca are more likely to support separatism than those who are not, while individuals fluent in the language of the central state are less likely to support separatist outcomes. Moreover, linguistic fluency shows a stronger relationship with support for separatism than ethnic identification. These results provide strong evidence that scholars should disaggregate language and ethnic identity in their analyses: language can be more salient for political preferences than ethnicity, and the most salient languages may not even be ethnic.

Bayesian analysis has emerged as a rapidly expanding frontier in qualitative methods. Recent work in this journal has voiced various doubts regarding how to implement Bayesian process tracing and the costs versus benefits of this approach. In this response, we articulate a very different understanding of the state of the method and a much more positive view of what Bayesian reasoning can do to strengthen qualitative social science. Drawing on forthcoming research as well as our earlier work, we focus on clarifying issues involving mutual exclusivity of hypotheses, evidentiary import, adjudicating among more than two hypotheses, and the logic of iterative research, with the goal of elucidating how Bayesian analysis operates and pushing the field forward.

Models for converting expert-coded data to estimates of latent concepts assume different data-generating processes (DGPs). In this paper, we simulate ecologically valid data according to different assumptions, and examine the degree to which common methods for aggregating expert-coded data (1) recover true values and (2) construct appropriate coverage intervals. We find that the mean and both hierarchical Aldrich–McKelvey (A–M) scaling and hierarchical item-response theory (IRT) models perform similarly when expert error is low; the hierarchical latent variable models (A-M and IRT) outperform the mean when expert error is high. Hierarchical A–M and IRT models generally perform similarly, although IRT models are often more likely to include true values within their coverage intervals. The median and non-hierarchical latent variable models perform poorly under most assumed DGPs.

There is renewed interest in levelling up the regions of the UK. The combination of social and political discontent, and the sluggishness of key UK macroeconomic indicators like productivity growth, has led to increased interest in understanding the regional economies of the UK. In turn, this has led to more investment in economic statistics. Specifically, the Office for National Statistics (ONS) recently started to produce quarterly regional GDP data for the nine English regions and Wales that date back to 2012Q1. This complements existing real GVA data for the regions available from the ONS on an annual basis back to 1998; with the devolved administrations of Scotland and Northern Ireland producing their own quarterly output measures. In this paper we reconcile these two data sources along with UK quarterly output data that date back to 1970. This enables us to produce both more timely real terms estimates of quarterly economic growth in the regions of the UK and a new reconciled historical time-series of quarterly regional real output data from 1970. We explore a number of features of interest of these new data. This includes producing a new quarterly regional productivity series and commenting on the evolution of regional productivity growth in the UK.

Given the increasing quantity and impressive placement of work on Bayesian process tracing, this approach has quickly become a frontier of qualitative research methods. Moreover, it has dominated the process-tracing modules at the Institute for Qualitative and Multi-Method Research (IQMR) and the American Political Science Association (APSA) meetings for over five years, rendering its impact even greater. Proponents of qualitative Bayesianism make a series of strong claims about its contributions and scope of inferential validity. Four claims stand out: (1) it enables causal inference from iterative research, (2) the sequence in which we evaluate evidence is irrelevant to inference, (3) it enables scholars to fully engage rival explanations, and (4) it prevents ad hoc hypothesizing and confirmation bias. Notwithstanding the stakes of these claims and breadth of traction this method has received, no one has systematically evaluated the promises, trade-offs, and limitations that accompany Bayesian process tracing. This article evaluates the extent to which the method lives up to the mission. Despite offering a useful framework for conducting iterative research, the current state of the method introduces more bias than it corrects for on numerous dimensions. The article concludes with an examination of the opportunity costs of learning Bayesian process tracing and a set of recommendations about how to push the field forward.