東華大學圖書館 |

Language: English

Help

回圖書館首頁

手機版館藏查詢

Back

Switch To: Labeled | MARC Mode | ISBD

Evaluation of categorical variable m...

Flora, David Benjamin.

Linked to FindBook

Google Book

Amazon

博客來

Evaluation of categorical variable methodology for confirmatory factor analysis with Likert-type data.

Record Type:	Language materials, printed : Monograph/item
Title/Author:	Evaluation of categorical variable methodology for confirmatory factor analysis with Likert-type data./
Author:	Flora, David Benjamin.
Description:	152 p.
Notes:	Director: Patrick J. Curran.
Contained By:	Dissertation Abstracts International63-03B.
Subject:	Psychology, Psychometrics. -
Online resource:	http://pqdd.sinica.edu.tw/twdaoapp/servlet/advanced?query=3046990
ISBN:	0493609636

Evaluation of categorical variable methodology for confirmatory factor analysis with Likert-type data.
Flora, David Benjamin.

Evaluation of categorical variable methodology for confirmatory factor analysis with Likert-type data. - 152 p.

Director: Patrick J. Curran.

Thesis (Ph.D.)--The University of North Carolina at Chapel Hill, 2002.

Applied researchers often use confirmatory factor analysis (CFA) to assess the construct validity of tests composed of items eliciting ordinal data. A theoretically appropriate method for conducting CFA with categorical data involves estimating the bivariate relationships among the observed items via polychoric correlations and then fitting the factor model using either weighted least squares (WLS) estimation, as per Muthén (1984), or a “robust WLS” method as presented by Muthén, Du Toit, and Spisic (1997). These procedures entail the assumption that a continuous, normal response process determines each item response. Very few research studies have assessed the finite sample properties of this categorical variable methodology (CVM) for CFA and none of these has specifically addressed the normality assumption for the underlying response variables. For the present research, I simulated ordinal data arising from both normal and non-normal continuous distributions that conform to properly specified CFA models of varying complexity, and fit the models using CVM. The findings suggest robustness to violation of the latent normality assumption and also demonstrate advantages of the “robust WLS” method relative to full WLS estimation.

ISBN: 0493609636Subjects--Topical Terms:

1017742
Psychology, Psychometrics.

Evaluation of categorical variable methodology for confirmatory factor analysis with Likert-type data.
LDR:02196nam 2200265 a 45 001 931578
005 20110429
008 110429s2002 eng d
020 $a 0493609636
035 $a (UnM)AAI3046990
035 $a AAI3046990
040 $a UnM $c UnM
100 1 $a Flora, David Benjamin. $3 1255122
245 1 0 $a Evaluation of categorical variable methodology for confirmatory factor analysis with Likert-type data.
300 $a 152 p.
500 $a Director: Patrick J. Curran.
500 $a Source: Dissertation Abstracts International, Volume: 63-03, Section: B, page: 1606.
502 $a Thesis (Ph.D.)--The University of North Carolina at Chapel Hill, 2002.
520 $a Applied researchers often use confirmatory factor analysis (CFA) to assess the construct validity of tests composed of items eliciting ordinal data. A theoretically appropriate method for conducting CFA with categorical data involves estimating the bivariate relationships among the observed items via polychoric correlations and then fitting the factor model using either weighted least squares (WLS) estimation, as per Muthén (1984), or a “robust WLS” method as presented by Muthén, Du Toit, and Spisic (1997). These procedures entail the assumption that a continuous, normal response process determines each item response. Very few research studies have assessed the finite sample properties of this categorical variable methodology (CVM) for CFA and none of these has specifically addressed the normality assumption for the underlying response variables. For the present research, I simulated ordinal data arising from both normal and non-normal continuous distributions that conform to properly specified CFA models of varying complexity, and fit the models using CVM. The findings suggest robustness to violation of the latent normality assumption and also demonstrate advantages of the “robust WLS” method relative to full WLS estimation.
590 $a School code: 0153.
650 4 $a Psychology, Psychometrics. $3 1017742
690 $a 0632
710 2 0 $a The University of North Carolina at Chapel Hill. $3 1017449
773 0 $t Dissertation Abstracts International $g 63-03B.
790 $a 0153
790 1 0 $a Curran, Patrick J., $e advisor
791 $a Ph.D.
792 $a 2002
856 4 0 $u http://pqdd.sinica.edu.tw/twdaoapp/servlet/advanced?query=3046990