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

A comparison of adaptive mastery tes...

Jacobs-Cassuto, Melody Sue.

FindBook

Google Book

Amazon

博客來

A comparison of adaptive mastery testing using testlets with the 3-parameter logistic model.

紀錄類型:	書目-語言資料,印刷品 : Monograph/item
正題名/作者:	A comparison of adaptive mastery testing using testlets with the 3-parameter logistic model./
作者:	Jacobs-Cassuto, Melody Sue.
面頁冊數:	210 p.
附註:	Adviser: David J. Weiss.
Contained By:	Dissertation Abstracts International66-03B.
標題:	Psychology, Psychometrics. -
電子資源:	http://pqdd.sinica.edu.tw/twdaoapp/servlet/advanced?query=3167690
ISBN:	9780542035791

A comparison of adaptive mastery testing using testlets with the 3-parameter logistic model.
Jacobs-Cassuto, Melody Sue.

A comparison of adaptive mastery testing using testlets with the 3-parameter logistic model. - 210 p.

Adviser: David J. Weiss.

Thesis (Ph.D.)--University of Minnesota, 2005.

Adaptive mastery testing is computerized testing tailored to the examinee and measures the mastery of some skill. Traditionally in adaptive testing, branching is based on a selection/algorithm procedure. Later research was interested in branching after a testlet or "bundle" of items interconnected through some common stimulus, was administered. There has been little research on conditions involving multiple adaptive testlet selection procedures using maximum information item selection with a 3-parameter logistic model.

ISBN: 9780542035791Subjects--Topical Terms:

1017742
Psychology, Psychometrics.

A comparison of adaptive mastery testing using testlets with the 3-parameter logistic model.
LDR:03328nam 2200289 a 45 001 969780
005 20110920
008 110921s2005 eng d
020 $a 9780542035791
035 $a (UnM)AAI3167690
035 $a AAI3167690
040 $a UnM $c UnM
100 1 $a Jacobs-Cassuto, Melody Sue. $3 1293838
245 1 2 $a A comparison of adaptive mastery testing using testlets with the 3-parameter logistic model.
300 $a 210 p.
500 $a Adviser: David J. Weiss.
500 $a Source: Dissertation Abstracts International, Volume: 66-03, Section: B, page: 1780.
502 $a Thesis (Ph.D.)--University of Minnesota, 2005.
520 $a Adaptive mastery testing is computerized testing tailored to the examinee and measures the mastery of some skill. Traditionally in adaptive testing, branching is based on a selection/algorithm procedure. Later research was interested in branching after a testlet or "bundle" of items interconnected through some common stimulus, was administered. There has been little research on conditions involving multiple adaptive testlet selection procedures using maximum information item selection with a 3-parameter logistic model.
520 $a In this simulation study, four main effects and their interactions were investigated in a completely crossed design: three selection methods (traditional CAT, maximum item information preserving testlet structure, and maximum testlet information), seven levels of theta, three testlet lengths, and two item banks. All data were simulated to reflect real data conditions in a foreign language reading comprehension test. Three classes of criteria were evaluated: (a) recovery of theta, (b) quality of classification decisions, and (c) efficiency. The recovery of theta was evaluated through (a) correlations, (b) root mean square error (RMSE) and (c) bias. Quality of classification decisions was evaluated through (a) classification percentages, (b) average decision error, and (c) percentage of simulees for whom a 95% confidence decision could not be made. Efficiency was evaluated through (a) &phis; correlations (b) test length, and (c) the ratio of precision divided by test length.
520 $a Results indicated that theta level was by far the most influential condition, having the largest effect size (eta2), except for the correlation results. For correlations, selection method emerged as the most influential condition. Testlet length had a smaller, more moderate amount of influence, and pool size had very little influence. Estimation of theta was most precise near the mastery cutoff and with TCAT and 2-item testlet conditions. The two testlet selection methods were more accurate in terms of classification decisions. For all three selection methods, accuracy of classification increased with increased distance from the cutoff. Implications for future research involve: (a) improving estimation across the entire theta continuum; and (b) modifying the construction of testlets to include a wider range of constructs, content balancing, and the stakes (high vs. low) of the tests administered.
590 $a School code: 0130.
650 4 $a Psychology, Psychometrics. $3 1017742
690 $a 0632
710 2 0 $a University of Minnesota. $3 676231
773 0 $t Dissertation Abstracts International $g 66-03B.
790 $a 0130
790 1 0 $a Weiss, David J., $e advisor
791 $a Ph.D.
792 $a 2005
856 4 0 $u http://pqdd.sinica.edu.tw/twdaoapp/servlet/advanced?query=3167690