https://osorensen.no/papers/Recent content in Papers on Øystein SørensenHugoenWed, 25 Mar 2026 00:00:00 +0000https://osorensen.no/papers/paper16/Wed, 25 Mar 2026 00:00:00 +0000https://osorensen.no/papers/paper16/This paper presents a hybrid sampler – alternating between one step of the No-U-Turn Sampler (NUTS) and one Gibbs step – for estimating dynamic structural equation models with binomial outcomes. The Gibbs step handles Pólya-Gamma distributed latent variables arising from a logit link, and the NUTS step uses a Kalman filter to marginalize over latent states. We demonstrate that the proposed sampler makes DSEM estimation with binomial data feasible for larger data and models than previously possible. arXiv preprint.https://osorensen.no/papers/paper15/Sat, 14 Mar 2026 00:00:00 +0000https://osorensen.no/papers/paper15/We present an estimation procedure for nonlinear mixed-effects models in which the population trajectory is represented by penalized splines and adapted to individuals via subject-specific transformation parameters. By exploiting the mixed model representation of penalized splines, the level of smoothness can be estimated jointly with other variance components. The approach is illustrated through a case study on infant height growth. arXiv preprint.https://osorensen.no/papers/paper14/Thu, 05 Mar 2026 00:00:00 +0000https://osorensen.no/papers/paper14/This paper shows that the within-level part of any dynamic structural equation model can be reformulated as a linear Gaussian state space model. Consequently, the latent states can be analytically marginalized using a Kalman filter, allowing for highly efficient estimation via Hamiltonian Monte Carlo. This makes DSEM estimation computationally tractable for much larger datasets than what has been previously possible. arXiv preprint.https://osorensen.no/papers/paper13/Wed, 22 Oct 2025 00:00:00 +0000https://osorensen.no/papers/paper13/This paper evaluates multiple interpretability techniques for machine learning models applied to neuroimaging data, including SHAP and SAGE. We trained XGBoost models to predict age and fluid intelligence using UK Biobank data and found that subcortical mean intensities are associated with brain aging, while fluid intelligence prediction is driven by the hippocampus and cerebellum. Published in Neuroinformatics.https://osorensen.no/papers/paper12/Wed, 01 Jan 2025 00:00:00 +0000https://osorensen.no/papers/paper12/Dynamic structural equation models have become extremely popular for analysis of intensive longitudinal data in the social sciences. One outstanding problem is how to handle nonlinear trends and cycles, and in this paper we propose to do this in a very flexible manner using regression splines. We test the methods in simulation studies, and then illustrate them by analyzing a diary data set on alcohol consumption and stress. Open-source Stan code is available from our OSF repository. Published in Multivariate Behavioral Research.https://osorensen.no/papers/paper11/Sun, 01 Dec 2024 00:00:00 +0000https://osorensen.no/papers/paper11/This paper considers rank and preference modeling for the case in which data arrive sequentially, rather than in a batch. The goal is to compute the posterior distribution incrementally in time, so that it can be quickly updated when new data arrives. To this end, we develop a sequential Monte Carlo algorithm for the Bayesian Mallows model. Published in Bayesian Analysis.https://osorensen.no/papers/paper10/Sun, 01 Sep 2024 00:00:00 +0000https://osorensen.no/papers/paper10/This paper presents the R package “galamm”, which contains open-source implementations for generalized additive latent and mixed models (GALAMMs). Published in Multivariate Behavioral Research.https://osorensen.no/papers/paper9/Thu, 01 Jun 2023 00:00:00 +0000https://osorensen.no/papers/paper9/This paper was motivated by the need to use flexible spline models to model how cognitive abilities change with age, and investigating how changes in abilities in certain cognitive domains can be explained by changes in the brain. To this end we developed a framework called Generalized Additive Latent and Mixed Models (GALAMM), combining generalized additive mixed models, structural equation modeling, and mxied effect modeling. We also propose algorithms for estimating the models, using sparse matrix methods and automatic differentation. Published in Psychometrika.https://osorensen.no/papers/paper8/Mon, 01 Feb 2021 00:00:00 +0000https://osorensen.no/papers/paper8/This paper considers estimation of lifespan brain trajectories from longitudinal data, using generalized additive mixed models. It studies how to distinguish longitudinal effects from cohort effects, and shows example code in R which practitioners can use for their own data. Published in NeuroImage.https://osorensen.no/papers/paper7/Fri, 01 Jan 2021 00:00:00 +0000https://osorensen.no/papers/paper7/This paper was motivated by the need to analyze neuroimaging data from multiple cohorts, when raw data could not be shared. We developed a method for meta analysis in which the results of fitting generalized additive (mixed) models could be shared, and a final meta analytic estimate computed based on weighting the individual fits. Published in NeuroImage.https://osorensen.no/papers/paper5/Mon, 01 Jun 2020 00:00:00 +0000https://osorensen.no/papers/paper5/This paper presents the BayesMallows R package, for analysis of rank and preference data. Published in the R Journal.https://osorensen.no/papers/paper6/Tue, 26 May 2020 00:00:00 +0000https://osorensen.no/papers/paper6/This paper considers predicting whether people are right or left brain dominant for language processing, based on their result in dichotic listening experiments as well as their handedness. Published in Lateralityhttps://osorensen.no/papers/paper4/Wed, 01 May 2019 00:00:00 +0000https://osorensen.no/papers/paper4/This paper describes the hdme R package, which provides implementation of variable selection in the presence of measurement error. Published in Journal of Open Source Software.https://osorensen.no/papers/paper2/Fri, 01 Jun 2018 00:00:00 +0000https://osorensen.no/papers/paper2/The matrix uncertainty selector is a modification of the Dantzig selector, for the case of variable selection with noisy predictors. In this paper we extend the matrix uncertainty selector to the generalized linear model case, and propose a computationally efficient computational algorithm. Compared to other methods that correct for the effect of measurement error, the matrix uncertainty selector and its extension do not require a precise estimate of the noise variance, which is an advantage in practical use. Published in Journal of Computational and Graphical Statistics.https://osorensen.no/papers/paper3/Sun, 01 Apr 2018 00:00:00 +0000https://osorensen.no/papers/paper3/This paper studies the analysis of rank and preference data. We consider both complete rankings, partial rankings, and pairwise preferences. We develop a complete Bayesian framework for estimating Mallows’ rank model in all this cases, including clustering of users with similar preferences and preference prediction. Published in Journal of Machine Learning Research. Joint first authorship with Valeria Vitelli.https://osorensen.no/papers/paper1/Wed, 01 Apr 2015 00:00:00 +0000https://osorensen.no/papers/paper1/This paper analyzes the impact of covariate measurement error in the lasso method for penalized regression. First, we present a result showing how the classical result for variable selection consistency breaks down in the presence of measurement error, and then we study a correction method and show how it recovers the consistent variable selection property. Finally, we consider an extension to logistic and Poisson regression. Published in Statistica Sinica.