Delimiting coefficient alpha from internal consistency and unidimensionality

K. Sijtsma

Research output: Contribution to journalArticleScientificpeer-review

Abstract

I discuss the contribution by Davenport, Davison, Liou, & Love (2015) in which they relate reliability represented by coefficient α to formal definitions of internal consistency and unidimensionality, both proposed by Cronbach (1951). I argue that coefficient α is a lower bound to reliability and that concepts of internal consistency and unidimensionality, however defined, belong to the realm of validity, viz. the issue of what the test measures. Internal consistency and unidimensionality may play a role in the construction of tests when the theory of the attribute for which the test is constructed implies that the items be internally consistent or unidimensional. I also offer examples of attributes that do not imply internal consistency or unidimensionality, thus limiting these concepts' usefulness in practical applications.
Original languageEnglish
Pages (from-to)10-13
JournalEducational Measurement: Issues and Practices
Volume34
Issue number4
DOIs
Publication statusPublished - 2015

Fingerprint

love

Cite this

@article{359046c23562420f901f23ac0474559c,
title = "Delimiting coefficient alpha from internal consistency and unidimensionality",
abstract = "I discuss the contribution by Davenport, Davison, Liou, & Love (2015) in which they relate reliability represented by coefficient α to formal definitions of internal consistency and unidimensionality, both proposed by Cronbach (1951). I argue that coefficient α is a lower bound to reliability and that concepts of internal consistency and unidimensionality, however defined, belong to the realm of validity, viz. the issue of what the test measures. Internal consistency and unidimensionality may play a role in the construction of tests when the theory of the attribute for which the test is constructed implies that the items be internally consistent or unidimensional. I also offer examples of attributes that do not imply internal consistency or unidimensionality, thus limiting these concepts' usefulness in practical applications.",
author = "K. Sijtsma",
year = "2015",
doi = "10.1111/emip.12099",
language = "English",
volume = "34",
pages = "10--13",
journal = "Educational Measurement: Issues and Practices",
issn = "0731-1745",
publisher = "Wiley",
number = "4",

}

Delimiting coefficient alpha from internal consistency and unidimensionality. / Sijtsma, K.

In: Educational Measurement: Issues and Practices, Vol. 34, No. 4, 2015, p. 10-13.

Research output: Contribution to journalArticleScientificpeer-review

TY - JOUR

T1 - Delimiting coefficient alpha from internal consistency and unidimensionality

AU - Sijtsma, K.

PY - 2015

Y1 - 2015

N2 - I discuss the contribution by Davenport, Davison, Liou, & Love (2015) in which they relate reliability represented by coefficient α to formal definitions of internal consistency and unidimensionality, both proposed by Cronbach (1951). I argue that coefficient α is a lower bound to reliability and that concepts of internal consistency and unidimensionality, however defined, belong to the realm of validity, viz. the issue of what the test measures. Internal consistency and unidimensionality may play a role in the construction of tests when the theory of the attribute for which the test is constructed implies that the items be internally consistent or unidimensional. I also offer examples of attributes that do not imply internal consistency or unidimensionality, thus limiting these concepts' usefulness in practical applications.

AB - I discuss the contribution by Davenport, Davison, Liou, & Love (2015) in which they relate reliability represented by coefficient α to formal definitions of internal consistency and unidimensionality, both proposed by Cronbach (1951). I argue that coefficient α is a lower bound to reliability and that concepts of internal consistency and unidimensionality, however defined, belong to the realm of validity, viz. the issue of what the test measures. Internal consistency and unidimensionality may play a role in the construction of tests when the theory of the attribute for which the test is constructed implies that the items be internally consistent or unidimensional. I also offer examples of attributes that do not imply internal consistency or unidimensionality, thus limiting these concepts' usefulness in practical applications.

U2 - 10.1111/emip.12099

DO - 10.1111/emip.12099

M3 - Article

VL - 34

SP - 10

EP - 13

JO - Educational Measurement: Issues and Practices

JF - Educational Measurement: Issues and Practices

SN - 0731-1745

IS - 4

ER -