A note on using stratified alpha to estimate the composite reliability of a test composed of interrelated nonhomogeneous items

Gordon Rae

    Research output: Contribution to journalArticle

    9 Citations (Scopus)

    Abstract

    The relationship between stratified alpha (alpha(s)) and the reliability of a test composed of interrelated nonhomogeneous items is examined. It is mathematically demonstrated that when there is congeneric equivalence within the strata or subtests, the difference between the coefficients is a function of the variances of the loadings within the strata. When the items within each stratum are essentially tau equivalent, these variances are alpha(s) and as and true reliability are equal, provided errors of measurement are uncorrelated. If errors of measurement are positively correlated and there is essential tau equivalence within strata, stratified alpha will overestimate reliability. These findings indicate that recent studies involving stratified alpha (A. Kamata, A. Turban, & E. Darandari, 2003; H. G. Osburn, 2000) need to be interpreted with some degree of caution. Nevertheless, the hypothetical population data presented in this article suggest that under certain circumstances, stratified alpha can be considerably greater than alpha and closer to the true reliability. Because the former is easily computed, it is recommended that with stratified tests, practicing researchers should routinely calculate both alpha and stratified alpha coefficients.
    LanguageEnglish
    Pages177-184
    JournalPsychological Methods
    Volume12
    Issue number2
    DOIs
    Publication statusPublished - Jun 2007

    Fingerprint

    social stratum
    equivalence

    Cite this

    @article{f23b3c2a91e84421a90925b89d6ab43d,
    title = "A note on using stratified alpha to estimate the composite reliability of a test composed of interrelated nonhomogeneous items",
    abstract = "The relationship between stratified alpha (alpha(s)) and the reliability of a test composed of interrelated nonhomogeneous items is examined. It is mathematically demonstrated that when there is congeneric equivalence within the strata or subtests, the difference between the coefficients is a function of the variances of the loadings within the strata. When the items within each stratum are essentially tau equivalent, these variances are alpha(s) and as and true reliability are equal, provided errors of measurement are uncorrelated. If errors of measurement are positively correlated and there is essential tau equivalence within strata, stratified alpha will overestimate reliability. These findings indicate that recent studies involving stratified alpha (A. Kamata, A. Turban, & E. Darandari, 2003; H. G. Osburn, 2000) need to be interpreted with some degree of caution. Nevertheless, the hypothetical population data presented in this article suggest that under certain circumstances, stratified alpha can be considerably greater than alpha and closer to the true reliability. Because the former is easily computed, it is recommended that with stratified tests, practicing researchers should routinely calculate both alpha and stratified alpha coefficients.",
    author = "Gordon Rae",
    year = "2007",
    month = "6",
    doi = "10.1037/1082-989X.12.2.177",
    language = "English",
    volume = "12",
    pages = "177--184",
    journal = "Psychological Methods",
    issn = "1082-989X",
    number = "2",

    }

    A note on using stratified alpha to estimate the composite reliability of a test composed of interrelated nonhomogeneous items. / Rae, Gordon.

    In: Psychological Methods, Vol. 12, No. 2, 06.2007, p. 177-184.

    Research output: Contribution to journalArticle

    TY - JOUR

    T1 - A note on using stratified alpha to estimate the composite reliability of a test composed of interrelated nonhomogeneous items

    AU - Rae, Gordon

    PY - 2007/6

    Y1 - 2007/6

    N2 - The relationship between stratified alpha (alpha(s)) and the reliability of a test composed of interrelated nonhomogeneous items is examined. It is mathematically demonstrated that when there is congeneric equivalence within the strata or subtests, the difference between the coefficients is a function of the variances of the loadings within the strata. When the items within each stratum are essentially tau equivalent, these variances are alpha(s) and as and true reliability are equal, provided errors of measurement are uncorrelated. If errors of measurement are positively correlated and there is essential tau equivalence within strata, stratified alpha will overestimate reliability. These findings indicate that recent studies involving stratified alpha (A. Kamata, A. Turban, & E. Darandari, 2003; H. G. Osburn, 2000) need to be interpreted with some degree of caution. Nevertheless, the hypothetical population data presented in this article suggest that under certain circumstances, stratified alpha can be considerably greater than alpha and closer to the true reliability. Because the former is easily computed, it is recommended that with stratified tests, practicing researchers should routinely calculate both alpha and stratified alpha coefficients.

    AB - The relationship between stratified alpha (alpha(s)) and the reliability of a test composed of interrelated nonhomogeneous items is examined. It is mathematically demonstrated that when there is congeneric equivalence within the strata or subtests, the difference between the coefficients is a function of the variances of the loadings within the strata. When the items within each stratum are essentially tau equivalent, these variances are alpha(s) and as and true reliability are equal, provided errors of measurement are uncorrelated. If errors of measurement are positively correlated and there is essential tau equivalence within strata, stratified alpha will overestimate reliability. These findings indicate that recent studies involving stratified alpha (A. Kamata, A. Turban, & E. Darandari, 2003; H. G. Osburn, 2000) need to be interpreted with some degree of caution. Nevertheless, the hypothetical population data presented in this article suggest that under certain circumstances, stratified alpha can be considerably greater than alpha and closer to the true reliability. Because the former is easily computed, it is recommended that with stratified tests, practicing researchers should routinely calculate both alpha and stratified alpha coefficients.

    U2 - 10.1037/1082-989X.12.2.177

    DO - 10.1037/1082-989X.12.2.177

    M3 - Article

    VL - 12

    SP - 177

    EP - 184

    JO - Psychological Methods

    T2 - Psychological Methods

    JF - Psychological Methods

    SN - 1082-989X

    IS - 2

    ER -