On the theoretical error bound for estimating psychometric functions

Huanping Dai, Virginia M. Richards

Research output: Contribution to journalArticle

1 Citation (Scopus)

Abstract

The theoretical limits to the amount of error, or the Cramer-Rao bounds, were derived for estimating psychometric functions. These theoretical error bounds were compared with the variability of psychometric functions estimated from human as well as computer-simulated observers. For the simulated observers, due to the limited efficiency of the sampling strategies, including the placement of the signals and the distribution of the trials, the variances of the estimated parameters are seven times the theoretical bound for threshold and 22 times that for slope. For the human observers, the variance is 18 times the theoretical bounds for threshold and 80 times that for slope. Therefore, a major portion of the variances (60% for threshold and 73% for slope) for the human observers is associated with factors other than sampling strategies. Further improvement of the accuracy for estimating psychometric functions will depend on not only optimizing the sampling strategy, but also better understanding the various sources of error related to the behavior of human observers.

Original languageEnglish (US)
Pages (from-to)919-926
Number of pages8
JournalAttention, Perception, and Psychophysics
Volume73
Issue number3
DOIs
StatePublished - 2011

Fingerprint

Psychometrics
psychometrics
efficiency
Research Design
Observer
Sampling Strategy

Keywords

  • Cramer-Rao bound
  • Psychometric function

ASJC Scopus subject areas

  • Sensory Systems
  • Experimental and Cognitive Psychology

Cite this

On the theoretical error bound for estimating psychometric functions. / Dai, Huanping; Richards, Virginia M.

In: Attention, Perception, and Psychophysics, Vol. 73, No. 3, 2011, p. 919-926.

Research output: Contribution to journalArticle

@article{2a54b1ba494549ca93517b8a4c1b8ac4,
title = "On the theoretical error bound for estimating psychometric functions",
abstract = "The theoretical limits to the amount of error, or the Cramer-Rao bounds, were derived for estimating psychometric functions. These theoretical error bounds were compared with the variability of psychometric functions estimated from human as well as computer-simulated observers. For the simulated observers, due to the limited efficiency of the sampling strategies, including the placement of the signals and the distribution of the trials, the variances of the estimated parameters are seven times the theoretical bound for threshold and 22 times that for slope. For the human observers, the variance is 18 times the theoretical bounds for threshold and 80 times that for slope. Therefore, a major portion of the variances (60{\%} for threshold and 73{\%} for slope) for the human observers is associated with factors other than sampling strategies. Further improvement of the accuracy for estimating psychometric functions will depend on not only optimizing the sampling strategy, but also better understanding the various sources of error related to the behavior of human observers.",
keywords = "Cramer-Rao bound, Psychometric function",
author = "Huanping Dai and Richards, {Virginia M.}",
year = "2011",
doi = "10.3758/s13414-010-0050-1",
language = "English (US)",
volume = "73",
pages = "919--926",
journal = "Attention, Perception, and Psychophysics",
issn = "1943-3921",
publisher = "Springer New York",
number = "3",

}

TY - JOUR

T1 - On the theoretical error bound for estimating psychometric functions

AU - Dai, Huanping

AU - Richards, Virginia M.

PY - 2011

Y1 - 2011

N2 - The theoretical limits to the amount of error, or the Cramer-Rao bounds, were derived for estimating psychometric functions. These theoretical error bounds were compared with the variability of psychometric functions estimated from human as well as computer-simulated observers. For the simulated observers, due to the limited efficiency of the sampling strategies, including the placement of the signals and the distribution of the trials, the variances of the estimated parameters are seven times the theoretical bound for threshold and 22 times that for slope. For the human observers, the variance is 18 times the theoretical bounds for threshold and 80 times that for slope. Therefore, a major portion of the variances (60% for threshold and 73% for slope) for the human observers is associated with factors other than sampling strategies. Further improvement of the accuracy for estimating psychometric functions will depend on not only optimizing the sampling strategy, but also better understanding the various sources of error related to the behavior of human observers.

AB - The theoretical limits to the amount of error, or the Cramer-Rao bounds, were derived for estimating psychometric functions. These theoretical error bounds were compared with the variability of psychometric functions estimated from human as well as computer-simulated observers. For the simulated observers, due to the limited efficiency of the sampling strategies, including the placement of the signals and the distribution of the trials, the variances of the estimated parameters are seven times the theoretical bound for threshold and 22 times that for slope. For the human observers, the variance is 18 times the theoretical bounds for threshold and 80 times that for slope. Therefore, a major portion of the variances (60% for threshold and 73% for slope) for the human observers is associated with factors other than sampling strategies. Further improvement of the accuracy for estimating psychometric functions will depend on not only optimizing the sampling strategy, but also better understanding the various sources of error related to the behavior of human observers.

KW - Cramer-Rao bound

KW - Psychometric function

UR - http://www.scopus.com/inward/record.url?scp=79959697054&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=79959697054&partnerID=8YFLogxK

U2 - 10.3758/s13414-010-0050-1

DO - 10.3758/s13414-010-0050-1

M3 - Article

C2 - 21264715

AN - SCOPUS:79959697054

VL - 73

SP - 919

EP - 926

JO - Attention, Perception, and Psychophysics

JF - Attention, Perception, and Psychophysics

SN - 1943-3921

IS - 3

ER -