Assessing bimodality to detect the presence of a dual cognitive process

Jonathan Freeman, Rick Dale

Research output: Contribution to journalArticle

Abstract

Researchers have long sought to distinguish between single-process and dual-process cognitive phenomena, using responses such as reaction times and, more recently, hand movements. Analysis of a response distribution's modality has been crucial in detecting the presence of dual processes, because they tend to introduce bimodal features. Rarely, however, have bimodality measures been systematically evaluated. We carried out tests of readily available bimodality measures that any researcher may easily employ: the bimodality coefficient (BC), Hartigan's dip statistic (HDS), and the difference in Akaike's information criterion between one-component and two-component distribution models (AICdiff). We simulated distributions containing two response populations and examined the influences of (1) the distances between populations, (2) proportions of responses, (3) the amount of positive skew present, and (4) sample size. Distance always had a stronger effect than did proportion, and the effects of proportion greatly differed across the measures. Skew biased the measures by increasing bimodality detection, in some cases leading to anomalous interactive effects. BC and HDS were generally convergent, but a number of important discrepancies were found. AICdiff was extremely sensitive to bimodality and identified nearly all distributions as bimodal. However, all measures served to detect the presence of bimodality in comparison to unimodal simulations. We provide a validation with experimental data, discuss methodological and theoretical implications, and make recommendations regarding the choice of analysis.

Original languageEnglish (US)
Pages (from-to)83-97
Number of pages15
JournalBehavior Research Methods
Volume45
Issue number1
DOIs
StatePublished - 2013

Fingerprint

Research Personnel
Sample Size
Population
Hand
Cognitive Processes
Proportion
Statistics
Dual Process
Bimodal

Keywords

  • Bimodality
  • Dual-process models
  • Mouse-tracking
  • Response distribution
  • Single-process models

ASJC Scopus subject areas

  • Psychology(all)
  • Psychology (miscellaneous)
  • Experimental and Cognitive Psychology
  • Arts and Humanities (miscellaneous)
  • Developmental and Educational Psychology

Cite this

Assessing bimodality to detect the presence of a dual cognitive process. / Freeman, Jonathan; Dale, Rick.

In: Behavior Research Methods, Vol. 45, No. 1, 2013, p. 83-97.

Research output: Contribution to journalArticle

@article{e25c795c7d55403b8acdfae69fdee6aa,
title = "Assessing bimodality to detect the presence of a dual cognitive process",
abstract = "Researchers have long sought to distinguish between single-process and dual-process cognitive phenomena, using responses such as reaction times and, more recently, hand movements. Analysis of a response distribution's modality has been crucial in detecting the presence of dual processes, because they tend to introduce bimodal features. Rarely, however, have bimodality measures been systematically evaluated. We carried out tests of readily available bimodality measures that any researcher may easily employ: the bimodality coefficient (BC), Hartigan's dip statistic (HDS), and the difference in Akaike's information criterion between one-component and two-component distribution models (AICdiff). We simulated distributions containing two response populations and examined the influences of (1) the distances between populations, (2) proportions of responses, (3) the amount of positive skew present, and (4) sample size. Distance always had a stronger effect than did proportion, and the effects of proportion greatly differed across the measures. Skew biased the measures by increasing bimodality detection, in some cases leading to anomalous interactive effects. BC and HDS were generally convergent, but a number of important discrepancies were found. AICdiff was extremely sensitive to bimodality and identified nearly all distributions as bimodal. However, all measures served to detect the presence of bimodality in comparison to unimodal simulations. We provide a validation with experimental data, discuss methodological and theoretical implications, and make recommendations regarding the choice of analysis.",
keywords = "Bimodality, Dual-process models, Mouse-tracking, Response distribution, Single-process models",
author = "Jonathan Freeman and Rick Dale",
year = "2013",
doi = "10.3758/s13428-012-0225-x",
language = "English (US)",
volume = "45",
pages = "83--97",
journal = "Behavior Research Methods",
issn = "1554-351X",
publisher = "Springer New York",
number = "1",

}

TY - JOUR

T1 - Assessing bimodality to detect the presence of a dual cognitive process

AU - Freeman, Jonathan

AU - Dale, Rick

PY - 2013

Y1 - 2013

N2 - Researchers have long sought to distinguish between single-process and dual-process cognitive phenomena, using responses such as reaction times and, more recently, hand movements. Analysis of a response distribution's modality has been crucial in detecting the presence of dual processes, because they tend to introduce bimodal features. Rarely, however, have bimodality measures been systematically evaluated. We carried out tests of readily available bimodality measures that any researcher may easily employ: the bimodality coefficient (BC), Hartigan's dip statistic (HDS), and the difference in Akaike's information criterion between one-component and two-component distribution models (AICdiff). We simulated distributions containing two response populations and examined the influences of (1) the distances between populations, (2) proportions of responses, (3) the amount of positive skew present, and (4) sample size. Distance always had a stronger effect than did proportion, and the effects of proportion greatly differed across the measures. Skew biased the measures by increasing bimodality detection, in some cases leading to anomalous interactive effects. BC and HDS were generally convergent, but a number of important discrepancies were found. AICdiff was extremely sensitive to bimodality and identified nearly all distributions as bimodal. However, all measures served to detect the presence of bimodality in comparison to unimodal simulations. We provide a validation with experimental data, discuss methodological and theoretical implications, and make recommendations regarding the choice of analysis.

AB - Researchers have long sought to distinguish between single-process and dual-process cognitive phenomena, using responses such as reaction times and, more recently, hand movements. Analysis of a response distribution's modality has been crucial in detecting the presence of dual processes, because they tend to introduce bimodal features. Rarely, however, have bimodality measures been systematically evaluated. We carried out tests of readily available bimodality measures that any researcher may easily employ: the bimodality coefficient (BC), Hartigan's dip statistic (HDS), and the difference in Akaike's information criterion between one-component and two-component distribution models (AICdiff). We simulated distributions containing two response populations and examined the influences of (1) the distances between populations, (2) proportions of responses, (3) the amount of positive skew present, and (4) sample size. Distance always had a stronger effect than did proportion, and the effects of proportion greatly differed across the measures. Skew biased the measures by increasing bimodality detection, in some cases leading to anomalous interactive effects. BC and HDS were generally convergent, but a number of important discrepancies were found. AICdiff was extremely sensitive to bimodality and identified nearly all distributions as bimodal. However, all measures served to detect the presence of bimodality in comparison to unimodal simulations. We provide a validation with experimental data, discuss methodological and theoretical implications, and make recommendations regarding the choice of analysis.

KW - Bimodality

KW - Dual-process models

KW - Mouse-tracking

KW - Response distribution

KW - Single-process models

UR - http://www.scopus.com/inward/record.url?scp=84874367149&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=84874367149&partnerID=8YFLogxK

U2 - 10.3758/s13428-012-0225-x

DO - 10.3758/s13428-012-0225-x

M3 - Article

VL - 45

SP - 83

EP - 97

JO - Behavior Research Methods

JF - Behavior Research Methods

SN - 1554-351X

IS - 1

ER -