東華大學圖書館 |

語系: 繁體中文

說明(常見問題)

回圖書館首頁

手機版館藏查詢

登入

回首頁

切換: 標籤 | MARC模式 | ISBD

Multidimensional item response theor...

The University of Oklahoma.

FindBook

Google Book

Amazon

博客來

Multidimensional item response theory: A SAS MDIRT macro and empirical study of PIAT math test.

紀錄類型:	書目-語言資料,印刷品 : Monograph/item
正題名/作者:	Multidimensional item response theory: A SAS MDIRT macro and empirical study of PIAT math test./
作者:	Lee, Sung-Hyuck.
面頁冊數:	92 p.
附註:	Adviser: Joseph L. Rodgers.
Contained By:	Dissertation Abstracts International68-02B.
標題:	Psychology, Psychometrics. -
電子資源:	http://pqdd.sinica.edu.tw/twdaoeng/servlet/advanced?query=3255213

Multidimensional item response theory: A SAS MDIRT macro and empirical study of PIAT math test.
Lee, Sung-Hyuck.

Multidimensional item response theory: A SAS MDIRT macro and empirical study of PIAT math test. - 92 p.

Adviser: Joseph L. Rodgers.

Thesis (Ph.D.)--The University of Oklahoma, 2007.

Even though unidimensional item response theory (IRT) provides a better framework for practical test settings than classical test theory (CTT), theoretical and empirical evidence shows that most response data violate the assumption of unidimensionality. There are several computer programs dedicated to estimating parameters based on the multidimensional perspective (MIRT). However, their accessibility is still costly, and they are not easy to use. In this paper, we present a SAS macro called MDIRT-FIT, to increase accessibility to the benefits obtained from this recent measurement theory development. The program is developed to estimate parameters based on a compensatory multidimensional item response theory (MIRT) model for dichotomous data. The full information item factor analysis model with an EM algorithm suggested in Bock & Aitken (1988) is implemented in the SAS programs. The estimation program written in SAS/IMLRTM provides both parameters of MIRT and parameters of the factor analysis model with their associated standard errors and overall model fit statistics. The maximum number of latent traits that can be estimated with this program is limited to five latent dimensions because of both computational burden and practical sufficiency. The accuracy and stability of the SAS macro is examined by utilizing simulated data of examinees' responses. The PIAT math test, a subset of the Peabody Individual Achievement Test, was examined to validate the comparability of the SAS macro to TESTFACT which is widely used for estimating parameters of MIRT models.Subjects--Topical Terms:

1017742
Psychology, Psychometrics.

Multidimensional item response theory: A SAS MDIRT macro and empirical study of PIAT math test.
LDR:02433nam 2200253 a 45 001 857498
005 20100712
008 100712s2007 ||||||||||||||||| ||eng d
035 $a (UMI)AAI3255213
035 $a AAI3255213
040 $a UMI $c UMI
100 1 $a Lee, Sung-Hyuck. $3 1024447
245 1 0 $a Multidimensional item response theory: A SAS MDIRT macro and empirical study of PIAT math test.
300 $a 92 p.
500 $a Adviser: Joseph L. Rodgers.
500 $a Source: Dissertation Abstracts International, Volume: 68-02, Section: B, page: 1357.
502 $a Thesis (Ph.D.)--The University of Oklahoma, 2007.
520 $a Even though unidimensional item response theory (IRT) provides a better framework for practical test settings than classical test theory (CTT), theoretical and empirical evidence shows that most response data violate the assumption of unidimensionality. There are several computer programs dedicated to estimating parameters based on the multidimensional perspective (MIRT). However, their accessibility is still costly, and they are not easy to use. In this paper, we present a SAS macro called MDIRT-FIT, to increase accessibility to the benefits obtained from this recent measurement theory development. The program is developed to estimate parameters based on a compensatory multidimensional item response theory (MIRT) model for dichotomous data. The full information item factor analysis model with an EM algorithm suggested in Bock & Aitken (1988) is implemented in the SAS programs. The estimation program written in SAS/IMLRTM provides both parameters of MIRT and parameters of the factor analysis model with their associated standard errors and overall model fit statistics. The maximum number of latent traits that can be estimated with this program is limited to five latent dimensions because of both computational burden and practical sufficiency. The accuracy and stability of the SAS macro is examined by utilizing simulated data of examinees' responses. The PIAT math test, a subset of the Peabody Individual Achievement Test, was examined to validate the comparability of the SAS macro to TESTFACT which is widely used for estimating parameters of MIRT models.
590 $a School code: 0169.
650 4 $a Psychology, Psychometrics. $3 1017742
690 $a 0632
710 2 $a The University of Oklahoma. $3 1021915
773 0 $t Dissertation Abstracts International $g 68-02B.
790 $a 0169
790 1 0 $a Rodgers, Joseph L., $e advisor
791 $a Ph.D.
792 $a 2007
856 4 0 $u http://pqdd.sinica.edu.tw/twdaoeng/servlet/advanced?query=3255213