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

Properties of a resampling validatio...

Mead, Alan David.

Linked to FindBook

Google Book

Amazon

博客來

Properties of a resampling validation technique for empirically scoring psychological assessments.

Record Type:	Electronic resources : Monograph/item
Title/Author:	Properties of a resampling validation technique for empirically scoring psychological assessments./
Author:	Mead, Alan David.
Description:	137 p.
Notes:	Source: Dissertation Abstracts International, Volume: 61-05, Section: B, page: 2805.
Contained By:	Dissertation Abstracts International61-05B.
Subject:	Occupational psychology. -
Online resource:	http://pqdd.sinica.edu.tw/twdaoapp/servlet/advanced?query=9971134
ISBN:	9780599762749

Properties of a resampling validation technique for empirically scoring psychological assessments.
Mead, Alan David.

Properties of a resampling validation technique for empirically scoring psychological assessments. - 137 p.

Source: Dissertation Abstracts International, Volume: 61-05, Section: B, page: 2805.

Thesis (Ph.D.)--University of Illinois at Urbana-Champaign, 2000.

Four studies were conducted to examine the psychometric properties of a new, resampling-based validity estimator which does not require a crossvalidation sample in order to produce unbiased estimates of validity. The first and second experiments were Monte Carlo simulations while the third and fourth applied the method to "real" data. In Experiment 1, a purely empirical method was found to be almost as effective as crossvalidation in debiasing the validity estimate; samples of 200 or more were found to be needed to produce valid empirical keys. In Experiment 2, a hybrid method which combines a priori model information was examined and found not to substantially improve the procedure unless the model information was of excellent quality (but the resampling estimator was also shown to be fairly robust to poor model information). Experiment 3 replicated the simulation results of Experiment 1 using "real" personality data and Experiment 4 compared the purely empirical keying to a hybrid method that blended empirical and a priori information for a biodata instrument. The results of these two experiments using empirical data confirmed many of the aspects of the simulations while pointing out some limitations of the simulated context. In general, the use of thresholds (minimum option-validities or number endorsing an option) to improve the robustness of the estimate was not strongly supported. The addition of a priori, model information in the hybrid keying only slightly improved upon the pure empirical approach. The hybrid keying method that relies most heavily on empirical information paired with a good model key allows samples as small as N = 100 to be keyed.

ISBN: 9780599762749Subjects--Topical Terms:

2122852
Occupational psychology.

Properties of a resampling validation technique for empirically scoring psychological assessments.
LDR:02604nmm a2200277 4500 001 2070728
005 20160617121855.5
008 170521s2000 ||||||||||||||||| ||eng d
020 $a 9780599762749
035 $a (MiAaPQ)AAI9971134
035 $a AAI9971134
040 $a MiAaPQ $c MiAaPQ
100 1 $a Mead, Alan David. $3 3185794
245 1 0 $a Properties of a resampling validation technique for empirically scoring psychological assessments.
300 $a 137 p.
500 $a Source: Dissertation Abstracts International, Volume: 61-05, Section: B, page: 2805.
500 $a Adviser: Fritz Drasgow.
502 $a Thesis (Ph.D.)--University of Illinois at Urbana-Champaign, 2000.
520 $a Four studies were conducted to examine the psychometric properties of a new, resampling-based validity estimator which does not require a crossvalidation sample in order to produce unbiased estimates of validity. The first and second experiments were Monte Carlo simulations while the third and fourth applied the method to "real" data. In Experiment 1, a purely empirical method was found to be almost as effective as crossvalidation in debiasing the validity estimate; samples of 200 or more were found to be needed to produce valid empirical keys. In Experiment 2, a hybrid method which combines a priori model information was examined and found not to substantially improve the procedure unless the model information was of excellent quality (but the resampling estimator was also shown to be fairly robust to poor model information). Experiment 3 replicated the simulation results of Experiment 1 using "real" personality data and Experiment 4 compared the purely empirical keying to a hybrid method that blended empirical and a priori information for a biodata instrument. The results of these two experiments using empirical data confirmed many of the aspects of the simulations while pointing out some limitations of the simulated context. In general, the use of thresholds (minimum option-validities or number endorsing an option) to improve the robustness of the estimate was not strongly supported. The addition of a priori, model information in the hybrid keying only slightly improved upon the pure empirical approach. The hybrid keying method that relies most heavily on empirical information paired with a good model key allows samples as small as N = 100 to be keyed.
590 $a School code: 0090.
650 4 $a Occupational psychology. $3 2122852
650 4 $a Quantitative psychology. $3 2144748
690 $a 0624
690 $a 0632
710 2 $a University of Illinois at Urbana-Champaign. $3 626646
773 0 $t Dissertation Abstracts International $g 61-05B.
790 $a 0090
791 $a Ph.D.
792 $a 2000
793 $a English
856 4 0 $u http://pqdd.sinica.edu.tw/twdaoapp/servlet/advanced?query=9971134