A framework for ML estimation of parameters of (mixtures of) common reaction time distributions given optional truncation or censoring

Conor V. Dolan, Han L J Van Der Maas, Peter Molenaar

Research output: Contribution to journalArticle

24 Citations (Scopus)

Abstract

We present a framework for distributional reaction time (RT) analysis, based on maximum likelihood (ML) estimation. Given certain information relating to chosen distribution functions, one can estimate the parameters of these distributions and of finite mixtures of these distributions. In addition, left and/or right censoring or truncation may be imposed. Censoring and truncation are useful methods by which to accommodate outlying observations, which are a pervasive problem in RT research. We consider five RT distributions: the Weibull, the ex-Gaussian, the gamma, the log-normal, and the Wald. We employ quasi-Newton optimization to obtain ML estimates. Multicase distributional analyses can be carried out, which enable one to conduct detailed (across or within subjects) comparisons of RT data by means of loglikelihood difference tests. Parameters may be freely estimated, estimated subject to boundary constraints, constrained to be equal (within or over cases), or fixed. To demonstrate the feasibility of ML estimation and to illustrate some of the possibilities offered by the present approach, we present three small simulation studies. In addition, we present three illustrative analyses of real data.

Original languageEnglish (US)
Pages (from-to)304-321
Number of pages18
JournalBehavior Research Methods, Instruments, and Computers
Volume34
Issue number3
StatePublished - Aug 2002

Fingerprint

Likelihood Functions
Research
Censoring
Maximum Likelihood Estimation
Reaction Time

All Science Journal Classification (ASJC) codes

  • Psychology(all)
  • Psychology (miscellaneous)
  • Experimental and Cognitive Psychology

Cite this

@article{6edce485221047b39e118f68fddcb33b,
title = "A framework for ML estimation of parameters of (mixtures of) common reaction time distributions given optional truncation or censoring",
abstract = "We present a framework for distributional reaction time (RT) analysis, based on maximum likelihood (ML) estimation. Given certain information relating to chosen distribution functions, one can estimate the parameters of these distributions and of finite mixtures of these distributions. In addition, left and/or right censoring or truncation may be imposed. Censoring and truncation are useful methods by which to accommodate outlying observations, which are a pervasive problem in RT research. We consider five RT distributions: the Weibull, the ex-Gaussian, the gamma, the log-normal, and the Wald. We employ quasi-Newton optimization to obtain ML estimates. Multicase distributional analyses can be carried out, which enable one to conduct detailed (across or within subjects) comparisons of RT data by means of loglikelihood difference tests. Parameters may be freely estimated, estimated subject to boundary constraints, constrained to be equal (within or over cases), or fixed. To demonstrate the feasibility of ML estimation and to illustrate some of the possibilities offered by the present approach, we present three small simulation studies. In addition, we present three illustrative analyses of real data.",
author = "Dolan, {Conor V.} and {Van Der Maas}, {Han L J} and Peter Molenaar",
year = "2002",
month = "8",
language = "English (US)",
volume = "34",
pages = "304--321",
journal = "Behavior Research Methods",
issn = "1554-351X",
publisher = "Springer New York",
number = "3",

}

A framework for ML estimation of parameters of (mixtures of) common reaction time distributions given optional truncation or censoring. / Dolan, Conor V.; Van Der Maas, Han L J; Molenaar, Peter.

In: Behavior Research Methods, Instruments, and Computers, Vol. 34, No. 3, 08.2002, p. 304-321.

Research output: Contribution to journalArticle

TY - JOUR

T1 - A framework for ML estimation of parameters of (mixtures of) common reaction time distributions given optional truncation or censoring

AU - Dolan, Conor V.

AU - Van Der Maas, Han L J

AU - Molenaar, Peter

PY - 2002/8

Y1 - 2002/8

N2 - We present a framework for distributional reaction time (RT) analysis, based on maximum likelihood (ML) estimation. Given certain information relating to chosen distribution functions, one can estimate the parameters of these distributions and of finite mixtures of these distributions. In addition, left and/or right censoring or truncation may be imposed. Censoring and truncation are useful methods by which to accommodate outlying observations, which are a pervasive problem in RT research. We consider five RT distributions: the Weibull, the ex-Gaussian, the gamma, the log-normal, and the Wald. We employ quasi-Newton optimization to obtain ML estimates. Multicase distributional analyses can be carried out, which enable one to conduct detailed (across or within subjects) comparisons of RT data by means of loglikelihood difference tests. Parameters may be freely estimated, estimated subject to boundary constraints, constrained to be equal (within or over cases), or fixed. To demonstrate the feasibility of ML estimation and to illustrate some of the possibilities offered by the present approach, we present three small simulation studies. In addition, we present three illustrative analyses of real data.

AB - We present a framework for distributional reaction time (RT) analysis, based on maximum likelihood (ML) estimation. Given certain information relating to chosen distribution functions, one can estimate the parameters of these distributions and of finite mixtures of these distributions. In addition, left and/or right censoring or truncation may be imposed. Censoring and truncation are useful methods by which to accommodate outlying observations, which are a pervasive problem in RT research. We consider five RT distributions: the Weibull, the ex-Gaussian, the gamma, the log-normal, and the Wald. We employ quasi-Newton optimization to obtain ML estimates. Multicase distributional analyses can be carried out, which enable one to conduct detailed (across or within subjects) comparisons of RT data by means of loglikelihood difference tests. Parameters may be freely estimated, estimated subject to boundary constraints, constrained to be equal (within or over cases), or fixed. To demonstrate the feasibility of ML estimation and to illustrate some of the possibilities offered by the present approach, we present three small simulation studies. In addition, we present three illustrative analyses of real data.

UR - http://www.scopus.com/inward/record.url?scp=0036689007&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=0036689007&partnerID=8YFLogxK

M3 - Article

C2 - 12395546

AN - SCOPUS:0036689007

VL - 34

SP - 304

EP - 321

JO - Behavior Research Methods

JF - Behavior Research Methods

SN - 1554-351X

IS - 3

ER -