TY - JOUR

T1 - Bayesian parameter estimation for the SWIFT model of eye-movement control during reading

AU - Seelig, Stefan A.

AU - Rabe, Maximilian M.

AU - Malem-Shinitski, Noa

AU - Risse, Sarah

AU - Reich, Sebastian

AU - Engbert, Ralf

N1 - Funding Information: This work was supported by grants from Deutsche Forschungsgemeinschaft, Germany ( SFB 1294 , project B03, project no. 318763901 to R.E. and S.Re.; SFB 1287 , project B03, project no. 317633480 to R.E.; grant RI 2504/1-1 to S.Ri.). We acknowledge a grant for computing time from Norddeutscher Verbund für Hoch- und Höchstleistungsrechnen, Germany (HLRN, grant bbx00001 ). Publisher Copyright: © 2020 Elsevier Inc.

PY - 2020/4

Y1 - 2020/4

N2 - Process-oriented theories of cognition must be evaluated against time-ordered observations. Here we present a representative example for data assimilation of the SWIFT model, a dynamical model of the control of fixation positions and fixation durations during natural reading of single sentences. First, we develop and test an approximate likelihood function of the model, which is a combination of a spatial, pseudo-marginal likelihood and a temporal likelihood obtained by probability density approximation Second, we implement a Bayesian approach to parameter inference using an adaptive Markov chain Monte Carlo procedure. Our results indicate that model parameters can be estimated reliably for individual subjects. We conclude that approximative Bayesian inference represents a considerable step forward for computational models of eye-movement control, where modeling of individual data on the basis of process-based dynamic models has not been possible so far.

AB - Process-oriented theories of cognition must be evaluated against time-ordered observations. Here we present a representative example for data assimilation of the SWIFT model, a dynamical model of the control of fixation positions and fixation durations during natural reading of single sentences. First, we develop and test an approximate likelihood function of the model, which is a combination of a spatial, pseudo-marginal likelihood and a temporal likelihood obtained by probability density approximation Second, we implement a Bayesian approach to parameter inference using an adaptive Markov chain Monte Carlo procedure. Our results indicate that model parameters can be estimated reliably for individual subjects. We conclude that approximative Bayesian inference represents a considerable step forward for computational models of eye-movement control, where modeling of individual data on the basis of process-based dynamic models has not been possible so far.

KW - Bayesian inference

KW - Dynamical models

KW - Eye movements

KW - Interindividual differences

KW - Likelihood function

KW - MCMC

KW - Reading

KW - Saccades

UR - http://www.scopus.com/inward/record.url?scp=85078134928&partnerID=8YFLogxK

U2 - 10.1016/j.jmp.2019.102313

DO - 10.1016/j.jmp.2019.102313

M3 - Journal article

AN - SCOPUS:85078134928

VL - 95

JO - Journal of Mathematical Psychology

JF - Journal of Mathematical Psychology

SN - 0022-2496

M1 - 102313

ER -