Aller au contenu principal
A Bayesian Mixture Modelling of Stop Signal Reaction Times (SSRT) Distributions
The distribution of Stop Signal Reaction Times (SSRT) as a measurement of the latency of the unobservable stopping process has been modelled over the entire stop signal task data as a single distribution. Former frequentist studies by the authors showed that the mean of these distributions are dependent on the preceding trial type (go or stop) and derived from a single underlying distribution are significantly smaller than those based on a mixture of distributions. Using data from a general population subsample of 44 (11 ADHD, 33 Control) children age 6-17 in Toronto, Canada, and an application of a Two Stage Bayesian Analysis (TSBA), we demonstrate that the single SSRT distribution and Mixture trial type SSRT distribution are comparable in stochastic order. Our results showed that the Single SSRT distribution is stochastically smaller than the Mixture of trial type SSRT distribution for both the ADHD and control populations . This supports the former frequentist results.
Date and Time
-
Co-auteurs (non y compris vous-même)
Annie Dupuis
Clinical Research Services, Sickkids Hospital
Russell Schachar
Psychiatry Research, Neuroscience and Mental Health, Sickkids Hospital
Michael Escobar
Biostatistics Division, Dalla Lana School of Public Health, University of Toronto
Langue de la présentation orale
Anglais
Langue des supports visuels
Anglais

Speaker

Edit Name Primary Affiliation
Mohsen Soltanifar Northeastern University, Canada