The MR- and Other Criteria for Orthogonal Rotation of Factors
Multiple Regeln (MR) und andere Kriterien für die orthogonale Rotation von Faktoren
Author(s) / Creator(s)
Schweizer, Karl
Abstract / Description
Several criteria for orthogonal factor rotation are presented and evaluated. They include one or an equally weighted combination of up to three computation rules. The computation rules are Carroll's (1953) rule, the rule which is complementary to Carroll's rule as well as a rule constructed according to the expert's ability to select the most appropriate configuration of high loadings in considering the correlation matrix. Orthogonal rotation due to the various criteria which include these computation rules as well as Quartimax and Varimax were applied to simulated data and special correlation matrices which included 7, 9, 10 and 20 variables. The MR (multiple rules) criterion which includes all three computation rules led to the most promising results in small numbers of variables. If the data sets were composed of 20 variables, Varimax and the MR criterion led to similar results. The application of the MR criterion is recommended in small numbers of variables whenever "simple structure" is not obtainable by other rotation methods.
Keyword(s)
Statistische Rotation Faktorenstruktur Faktorenanalyse Statistical Rotation Factor Structure Factor Analysis Factor Analysis Factor Rotation Orthogonal RotationPersistent Identifier
Date of first publication
2001
Journal title
Methods of Psychological Research
Volume
6
Issue
1
Page numbers
1-20
Publisher
IPN - Institute for Science Education at the University of Kiel, Germany
Publication status
publishedVersion
Review status
unknown
Citation
-
MPR-Online_2001_6.1_Schweizer.pdfAdobe PDF - 155.36KBMD5 : 9733a235c0d7c8f8eb8d04cf9128a139
-
There are no other versions of this object.
-
Author(s) / Creator(s)Schweizer, Karl
-
PsychArchives acquisition timestamp2023-04-25T14:26:02Z
-
Made available on2023-04-25T14:26:02Z
-
Date of first publication2001
-
Abstract / DescriptionSeveral criteria for orthogonal factor rotation are presented and evaluated. They include one or an equally weighted combination of up to three computation rules. The computation rules are Carroll's (1953) rule, the rule which is complementary to Carroll's rule as well as a rule constructed according to the expert's ability to select the most appropriate configuration of high loadings in considering the correlation matrix. Orthogonal rotation due to the various criteria which include these computation rules as well as Quartimax and Varimax were applied to simulated data and special correlation matrices which included 7, 9, 10 and 20 variables. The MR (multiple rules) criterion which includes all three computation rules led to the most promising results in small numbers of variables. If the data sets were composed of 20 variables, Varimax and the MR criterion led to similar results. The application of the MR criterion is recommended in small numbers of variables whenever "simple structure" is not obtainable by other rotation methods.en
-
Publication statuspublishedVersion
-
Review statusunknown
-
ISSN1432-8534
-
Persistent Identifierhttps://hdl.handle.net/20.500.12034/8282
-
Persistent Identifierhttps://doi.org/10.23668/psycharchives.12759
-
Language of contenteng
-
PublisherIPN - Institute for Science Education at the University of Kiel, Germany
-
Keyword(s)Statistische Rotationde_DE
-
Keyword(s)Faktorenstrukturde_DE
-
Keyword(s)Faktorenanalysede_DE
-
Keyword(s)Statistical Rotationen_US
-
Keyword(s)Factor Structureen_US
-
Keyword(s)Factor Analysisen_US
-
Keyword(s)Factor Analysisen_US
-
Keyword(s)Factor Rotationen_US
-
Keyword(s)Orthogonal Rotationen_US
-
Dewey Decimal Classification number(s)150
-
TitleThe MR- and Other Criteria for Orthogonal Rotation of Factorsen_US
-
Alternative titleMultiple Regeln (MR) und andere Kriterien für die orthogonale Rotation von Faktorende_DE
-
DRO typearticle
-
DFK number from PSYNDEX159094
-
Issue1
-
Journal titleMethods of Psychological Research
-
Page numbers1-20
-
Volume6
-
Visible tag(s)Version of Record