Quadratic prediction of factors scores

Erik Meijer; Tom Wansbeek

doi:10.1007/bf02294569

Quadratic prediction of factors scores

Erik Meijer, Tom Wansbeek

Faculty of Economics and Business

Research output: Contribution to journal › Article › Academic › peer-review

2 Citations (Scopus)

Abstract

Factor scores are naturally predicted by means of their conditional expectation given the indicators y. Under normality this expectation is linear in y but in general it is an unknown function of y. II is discussed that under nonnormality factor scores can be more precisely predicted by a quadratic function of y.

Original language	English
Pages (from-to)	495 - 508
Number of pages	14
Journal	Psychometrika
Volume	64
Issue number	4
DOIs	https://doi.org/10.1007/bf02294569
Publication status	Published - Dec-1999

Access to Document

10.1007/bf02294569

Handle.net

Persistent link

Embargoed Document

Quadratic prediction of factor scores
Final publisher's version, 727 KB
Request copy

Cite this

@article{297b6a633b0a48a6978f9dbc91e84bc4,

title = "Quadratic prediction of factors scores",

abstract = "Factor scores are naturally predicted by means of their conditional expectation given the indicators y. Under normality this expectation is linear in y but in general it is an unknown function of y. II is discussed that under nonnormality factor scores can be more precisely predicted by a quadratic function of y.",

keywords = "factor scores, nonnormality, efficiency",

author = "Erik Meijer and Tom Wansbeek",

year = "1999",

month = dec,

doi = "10.1007/bf02294569",

language = "English",

volume = "64",

pages = "495 -- 508",

journal = "Psychometrika",

issn = "0033-3123",

publisher = "Springer",

number = "4",

}

TY - JOUR

T1 - Quadratic prediction of factors scores

AU - Meijer, Erik

AU - Wansbeek, Tom

PY - 1999/12

Y1 - 1999/12

N2 - Factor scores are naturally predicted by means of their conditional expectation given the indicators y. Under normality this expectation is linear in y but in general it is an unknown function of y. II is discussed that under nonnormality factor scores can be more precisely predicted by a quadratic function of y.

AB - Factor scores are naturally predicted by means of their conditional expectation given the indicators y. Under normality this expectation is linear in y but in general it is an unknown function of y. II is discussed that under nonnormality factor scores can be more precisely predicted by a quadratic function of y.

KW - factor scores

KW - nonnormality

KW - efficiency

U2 - 10.1007/bf02294569

DO - 10.1007/bf02294569

M3 - Article

SN - 0033-3123

VL - 64

SP - 495

EP - 508

JO - Psychometrika

JF - Psychometrika

IS - 4

ER -