Confidence regions for INDSCAL using the jackknife and bootstrap techniques

Sharon L. Weinberg, J. Douglas Carroll, Harvey S. Cohen

Research output: Contribution to journalArticle

Abstract

Bootstrap and jackknife techniques are used to estimate ellipsoidal confidence regions of group stimulus points derived from INDSCAL. The validity of these estimates is assessed through Monte Carlo analysis. Asymptotic estimates of confidence regions based on a MULTISCALE solution are also evaluated. Our findings suggest that the bootstrap and jackknife techniques may be used to provide statements regarding the accuracy of the relative locations of points in space. Our findings also suggest that MULTISCALE asymptotic estimates of confidence regions based on small samples provide an optimistic view of the actual statistical reliability of the solution.

Original languageEnglish (US)
Pages (from-to)475-491
Number of pages17
JournalPsychometrika
Volume49
Issue number4
DOIs
StatePublished - Dec 1984

Fingerprint

INDSCAL
Jackknife
Confidence Region
Bootstrap
Asymptotic Estimates
Point groups
Small Sample
Estimate

Keywords

  • Individual differences
  • maximum likelihood
  • multidimensional scaling
  • pseudovalues
  • resampling schemes
  • stability
  • standard errors

ASJC Scopus subject areas

  • Psychology(all)
  • Applied Mathematics

Cite this

Confidence regions for INDSCAL using the jackknife and bootstrap techniques. / Weinberg, Sharon L.; Carroll, J. Douglas; Cohen, Harvey S.

In: Psychometrika, Vol. 49, No. 4, 12.1984, p. 475-491.

Research output: Contribution to journalArticle

Weinberg, Sharon L. ; Carroll, J. Douglas ; Cohen, Harvey S. / Confidence regions for INDSCAL using the jackknife and bootstrap techniques. In: Psychometrika. 1984 ; Vol. 49, No. 4. pp. 475-491.
@article{c18d4524d3a74595ac32717c3f2a1bc8,
title = "Confidence regions for INDSCAL using the jackknife and bootstrap techniques",
abstract = "Bootstrap and jackknife techniques are used to estimate ellipsoidal confidence regions of group stimulus points derived from INDSCAL. The validity of these estimates is assessed through Monte Carlo analysis. Asymptotic estimates of confidence regions based on a MULTISCALE solution are also evaluated. Our findings suggest that the bootstrap and jackknife techniques may be used to provide statements regarding the accuracy of the relative locations of points in space. Our findings also suggest that MULTISCALE asymptotic estimates of confidence regions based on small samples provide an optimistic view of the actual statistical reliability of the solution.",
keywords = "Individual differences, maximum likelihood, multidimensional scaling, pseudovalues, resampling schemes, stability, standard errors",
author = "Weinberg, {Sharon L.} and Carroll, {J. Douglas} and Cohen, {Harvey S.}",
year = "1984",
month = "12",
doi = "10.1007/BF02302586",
language = "English (US)",
volume = "49",
pages = "475--491",
journal = "Psychometrika",
issn = "0033-3123",
publisher = "Springer New York",
number = "4",

}

TY - JOUR

T1 - Confidence regions for INDSCAL using the jackknife and bootstrap techniques

AU - Weinberg, Sharon L.

AU - Carroll, J. Douglas

AU - Cohen, Harvey S.

PY - 1984/12

Y1 - 1984/12

N2 - Bootstrap and jackknife techniques are used to estimate ellipsoidal confidence regions of group stimulus points derived from INDSCAL. The validity of these estimates is assessed through Monte Carlo analysis. Asymptotic estimates of confidence regions based on a MULTISCALE solution are also evaluated. Our findings suggest that the bootstrap and jackknife techniques may be used to provide statements regarding the accuracy of the relative locations of points in space. Our findings also suggest that MULTISCALE asymptotic estimates of confidence regions based on small samples provide an optimistic view of the actual statistical reliability of the solution.

AB - Bootstrap and jackknife techniques are used to estimate ellipsoidal confidence regions of group stimulus points derived from INDSCAL. The validity of these estimates is assessed through Monte Carlo analysis. Asymptotic estimates of confidence regions based on a MULTISCALE solution are also evaluated. Our findings suggest that the bootstrap and jackknife techniques may be used to provide statements regarding the accuracy of the relative locations of points in space. Our findings also suggest that MULTISCALE asymptotic estimates of confidence regions based on small samples provide an optimistic view of the actual statistical reliability of the solution.

KW - Individual differences

KW - maximum likelihood

KW - multidimensional scaling

KW - pseudovalues

KW - resampling schemes

KW - stability

KW - standard errors

UR - http://www.scopus.com/inward/record.url?scp=0000551530&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=0000551530&partnerID=8YFLogxK

U2 - 10.1007/BF02302586

DO - 10.1007/BF02302586

M3 - Article

AN - SCOPUS:0000551530

VL - 49

SP - 475

EP - 491

JO - Psychometrika

JF - Psychometrika

SN - 0033-3123

IS - 4

ER -