The impact of missing data on sample reliability estimates: Implications for reliability reporting practices

Craig K. Enders

Research output: Contribution to journalArticle

27 Citations (Scopus)

Abstract

A method for incorporating maximum likelihood (ML) estimation into reliability analyses with item-level missing data is outlined. An ML estimate of the covariance matrix is first obtained using the expectation maximization (EM) algorithm, and coefficient alpha is subsequently computed using standard formulae. A simulation study demonstrated that the EM approach yields (a) less bias in reliability estimates, (b) dramatically reduces cross-sample fluctuation of estimates, and (c) yields more accurate confidence intervals. Implications for reliability reporting practices are discussed, and the EM procedure is demonstrated using a heuristic data set.

Original languageEnglish (US)
Pages (from-to)419-436
Number of pages18
JournalEducational and Psychological Measurement
Volume64
Issue number3
DOIs
StatePublished - Jun 2004
Externally publishedYes

Fingerprint

Missing Data
Expectation Maximization
Likelihood Functions
Estimate
Maximum likelihood estimation
Expectation-maximization Algorithm
Confidence Intervals
Covariance matrix
Maximum Likelihood Estimate
Maximum Likelihood Estimation
fluctuation
Maximum likelihood
Confidence interval
heuristics
confidence
Simulation Study
Heuristics
Fluctuations
simulation
Coefficient

Keywords

  • EM algorithm
  • Maximum likelihood
  • Missing data
  • Reliability
  • Reliability generalization

ASJC Scopus subject areas

  • Mathematics (miscellaneous)
  • Psychology(all)
  • Developmental and Educational Psychology
  • Psychology (miscellaneous)

Cite this

The impact of missing data on sample reliability estimates : Implications for reliability reporting practices. / Enders, Craig K.

In: Educational and Psychological Measurement, Vol. 64, No. 3, 06.2004, p. 419-436.

Research output: Contribution to journalArticle

@article{be4a6dc469434085afb754a4ed4f1ae5,
title = "The impact of missing data on sample reliability estimates: Implications for reliability reporting practices",
abstract = "A method for incorporating maximum likelihood (ML) estimation into reliability analyses with item-level missing data is outlined. An ML estimate of the covariance matrix is first obtained using the expectation maximization (EM) algorithm, and coefficient alpha is subsequently computed using standard formulae. A simulation study demonstrated that the EM approach yields (a) less bias in reliability estimates, (b) dramatically reduces cross-sample fluctuation of estimates, and (c) yields more accurate confidence intervals. Implications for reliability reporting practices are discussed, and the EM procedure is demonstrated using a heuristic data set.",
keywords = "EM algorithm, Maximum likelihood, Missing data, Reliability, Reliability generalization",
author = "Enders, {Craig K.}",
year = "2004",
month = "6",
doi = "10.1177/0013164403261050",
language = "English (US)",
volume = "64",
pages = "419--436",
journal = "Educational and Psychological Measurement",
issn = "0013-1644",
publisher = "SAGE Publications Inc.",
number = "3",

}

TY - JOUR

T1 - The impact of missing data on sample reliability estimates

T2 - Implications for reliability reporting practices

AU - Enders, Craig K.

PY - 2004/6

Y1 - 2004/6

N2 - A method for incorporating maximum likelihood (ML) estimation into reliability analyses with item-level missing data is outlined. An ML estimate of the covariance matrix is first obtained using the expectation maximization (EM) algorithm, and coefficient alpha is subsequently computed using standard formulae. A simulation study demonstrated that the EM approach yields (a) less bias in reliability estimates, (b) dramatically reduces cross-sample fluctuation of estimates, and (c) yields more accurate confidence intervals. Implications for reliability reporting practices are discussed, and the EM procedure is demonstrated using a heuristic data set.

AB - A method for incorporating maximum likelihood (ML) estimation into reliability analyses with item-level missing data is outlined. An ML estimate of the covariance matrix is first obtained using the expectation maximization (EM) algorithm, and coefficient alpha is subsequently computed using standard formulae. A simulation study demonstrated that the EM approach yields (a) less bias in reliability estimates, (b) dramatically reduces cross-sample fluctuation of estimates, and (c) yields more accurate confidence intervals. Implications for reliability reporting practices are discussed, and the EM procedure is demonstrated using a heuristic data set.

KW - EM algorithm

KW - Maximum likelihood

KW - Missing data

KW - Reliability

KW - Reliability generalization

UR - http://www.scopus.com/inward/record.url?scp=2542569599&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=2542569599&partnerID=8YFLogxK

U2 - 10.1177/0013164403261050

DO - 10.1177/0013164403261050

M3 - Article

AN - SCOPUS:2542569599

VL - 64

SP - 419

EP - 436

JO - Educational and Psychological Measurement

JF - Educational and Psychological Measurement

SN - 0013-1644

IS - 3

ER -