Statistical Models for Psychological and Linguistic Data
Statistical Models for Psychological and Linguistic Data
August 2019 - Juli 2021
The goal of the cooperation group is to investigate and further develop a series of statistical methods that are now available for (a) the analysis of experimental and psychometric data from psychology and psycholinguistics, (b) the modeling of linguistic distributional data, and, possibly going beyond these domains, (c) the analyses of genome-wide associations. The methods in focus are (generalized) linear mixed models [(G)LMMs], generalized additive (mixed) models [GA(M)Ms], and multivariate (generalized) mixed models [MV(G)MMs]. These statistical methods deal with inferential statistical problems that arise from dependencies between, for example, measures on the same subjects or the same items in psycholinguistic experiments or, again as an example beyond the core domains, the same nucleotides in the genome (i.e., within-unit correlations). The research uses the R Computational Environment and the Julia Programming Language.
Special effort will be devoted to develop new software and to disseminate new knowledge through tutorial articles. Also Tutorial Workshops are planned for advanced students and postdoctoral fellows. All products of the Cooperation Group will be put into public domain as soon as possible. The Cooperation Group contributes to establishing the ZiF as an Open Science Hub.
Topics of Research Workshops
- I
Case studies that demonstrate the advantages of Julia-based MixedModels with respect to data size and model complexity. The case studies show that users can seamlessly switch between the R and the Julia language. For example, they may want to use the familiar R Computational Environment for data preparation and visualization and use Julia MixedModels only for model fitting.
- II
Work on methodological issues of (G)LMMs relating to (a) reliability and precision of parameter estimates, (b) model identification based on principal components of random-effect structure, and (c) model selection within the spectrum of identified models. In perspective, this work promotes much needed sensitivity to the relevance of variance and correlation parameters with respect to the interpretation of fixed effects.
- III
- IV
Members
- Convenors
Cognitive Psychology
(University of Potsdam, GER)
Reinhold Kliegl received his PhD at the University of Colorado in Boulder in psychology. He worked as a research scientist at the Max Planck Institute for Human Development, Berlin, before joining the University of Potsdam, Germany, as professor of experimental psychology. His research focuses on how the dynamics of language-related, perceptual, and oculomotor processes subserve attentional control, using reading tasks as an experimental venue. His primary interest in mixed models is theory-guided simultaneous modeling of (quasi-) experimental effects and individual / item differenes in (quasi-)experimental effects.
Quantitative Linguistics
(Eberhard Karls University Tübingen, GER)
Harald Baayen received his PhD at the Free University of Amsterdam in linguistics. He joined the scientific staff at the Max Plank Institute for Psycholinguistics, Nijmegen, before moving to the faculty at the University of Alberta, Edmonton, and then to Eberhard Karls University, Tübingen, as a Professor of Quantitative Linguistics. His work focuses on human speech, both how it is generated and how it is processed by the brain and machines. He has a long-standing interest in statistical methods, including linear mixed models, and a particular interest in the combination of generalized additive mixed models and quantile regression.
Statistics
(University of Wisconsin, USA)
Douglas Bates received his PhD at Queen's University in Kingston in statistics and is now professor emeritus at the University of Wisconsin in Madison. His specialty are generalized linear and nonlinear mixed models. He is a founding member of the R Core Team and the author of numerous packages in this computing environment, among them lme4 and matrix. In recent years he has shifted his primary involvement to the Julia Programming Language with a focus on the implementation of a MixedModels package and computational interfaces with R.
- Fellows
Phillip M. Alday
Phillip Alday received his doctorate from the University of Marburg. After working for two years at the University of South of Australia, he is now at the Max Planck Institute for Psycholinguistics. His research focuses on the electrophysiology of language. He is particularly interested in using modern statistical and signal processing methods to conduct previously impossible experiments.Christina Bergmann
Christina Bergmann received her PhD from Radboud University Nijmegen, The Netherlands. After a post doc at Ecole Normale Superieure in Paris, France, she now works at the Max Planck Institute for Psycholinguistics. Christina studies early word learning using behavioral methods and computational modeling and actively participates in the large-scale developmental ManyBabies collaboration (manybabies.stanford.edu; governing board member) and applies meta-scientific methods to infant data (metalab.stanford.edu).Lisa DeBruine
Lisa DeBruine is a professor in the Institute of Neuroscience and Psychology at the University of Glasgow and a founding member of the Psychological Science Accelerator. See https://debruine.github.io/ for tutorials, open access teaching materials on data skills for reproducible science, shiny apps, and other open science odds and ends.Dave Kleinschmidt
Dave Kleinschmidt received his Ph.D. at the University of Rochester in brain and cognitive sciences. He recently joined the Rutgers faculty after a post-doc at the Princeton Neuroscience Institute. Research in his Learning, Adaptation, and Perception Lab (LeAP Lab) focuses on perception in a variable, multi-context world. His work uses a combination of methods, from behavioral experiments to computational modeling to cognitive neuroscience.Hannes Matuschek
José Bayoán Santiago Calderón
José Bayoán Santiago Calderón received his PhD in economics from Claremont Graduate University. He is currently a postdoctoral research associate at the Social & Decision Analytics Division of the Biocomplexity Institute & Initiative at the University of Virginia. Bayoán has previously worked as a data scientist and software developer in the areas of demand-side management (DSM) and pharmacokinetics. He is also an active member of the Julia language community contributing, developing, and maintaining several packages.- Associated Members
Milan Bouchet-Valat
Milan Bouchet-Valat received his PhD in Sociology from Sciences Po in Paris. He is now a researcher at the French Institute for Demographic Studies (Ined), where he studies class and gender inequalities and quantitative methods in social sciences. The author of several R packages, he contributes since 2014 to the Julia language and to data and statistics Julia packages.Anna Gert
Benedikt Ehinger
Shravan Vasishth
Shravan Vasishth received his PhD at the Ohio State University in linguistics. After obtaining his PhD, he moved to the University of Saarland for postdoctoral research. Now he is professor of psycholinguistics at the University of Potsdam. He develops computational models of human sentence comprehension, focusing on both impaired and unimpaired populations. He is especially interested in modeling individual differences in language comprehension. He is also very much interested in statistical theory and practice, particularly in using Bayesian methods for data analysis and for computational modeling.