A study of a network-flow algorithm and a noncorrecting algorithm for test assembly

R. D. Armstrong, Douglas H. Jones, Xuan Li

Research output: Contribution to journalArticle

9 Citations (Scopus)

Abstract

The network-flow algorithm (NFA) of Armstrong, Jones, & Wu (1992) and the average growth approximation algorithm (AGAA) of Luecht & Hirsch (1992) were evaluated as methods for automated test assembly. The algorithms were used on ACT and ASVAB item banks, with and without error in the item parameters. Both algorithms matched a target test information function on the ACT item bank, both before and after error was introduced. The NFA matched the target on the ASVAB item bank; however, the AGAA did not, even without error in this item bank. The AGAA is a noncorrecting algorithm, and it made poor item selections early in the search process when using the ASVAB item bank. The NFA corrects for nonoptimal choices with a simplex search. The results indicate that reasonable error in item parameters is not harmful for test assembly using the NFA or AGAA on certain types of item banks.

Original languageEnglish (US)
Pages (from-to)89-98
Number of pages10
JournalApplied Psychological Measurement
Volume20
Issue number1
DOIs
StatePublished - Mar 1996

Fingerprint

bank
Growth

All Science Journal Classification (ASJC) codes

  • Social Sciences (miscellaneous)
  • Psychology (miscellaneous)

Keywords

  • Algorithmic test construction
  • Automated test assembly
  • Greedy algorithm
  • Heuristic algorithm
  • Item response theory
  • Marginal maximum likelihood
  • Mathematical programming
  • Simulation
  • Test construction

Cite this

@article{e5258a0407fd4092be7502e56ebbdb87,
title = "A study of a network-flow algorithm and a noncorrecting algorithm for test assembly",
abstract = "The network-flow algorithm (NFA) of Armstrong, Jones, & Wu (1992) and the average growth approximation algorithm (AGAA) of Luecht & Hirsch (1992) were evaluated as methods for automated test assembly. The algorithms were used on ACT and ASVAB item banks, with and without error in the item parameters. Both algorithms matched a target test information function on the ACT item bank, both before and after error was introduced. The NFA matched the target on the ASVAB item bank; however, the AGAA did not, even without error in this item bank. The AGAA is a noncorrecting algorithm, and it made poor item selections early in the search process when using the ASVAB item bank. The NFA corrects for nonoptimal choices with a simplex search. The results indicate that reasonable error in item parameters is not harmful for test assembly using the NFA or AGAA on certain types of item banks.",
keywords = "Algorithmic test construction, Automated test assembly, Greedy algorithm, Heuristic algorithm, Item response theory, Marginal maximum likelihood, Mathematical programming, Simulation, Test construction",
author = "Armstrong, {R. D.} and Jones, {Douglas H.} and Xuan Li",
year = "1996",
month = "3",
doi = "10.1177/014662169602000108",
language = "English (US)",
volume = "20",
pages = "89--98",
journal = "Applied Psychological Measurement",
issn = "0146-6216",
publisher = "SAGE Publications Inc.",
number = "1",

}

A study of a network-flow algorithm and a noncorrecting algorithm for test assembly. / Armstrong, R. D.; Jones, Douglas H.; Li, Xuan.

In: Applied Psychological Measurement, Vol. 20, No. 1, 03.1996, p. 89-98.

Research output: Contribution to journalArticle

TY - JOUR

T1 - A study of a network-flow algorithm and a noncorrecting algorithm for test assembly

AU - Armstrong, R. D.

AU - Jones, Douglas H.

AU - Li, Xuan

PY - 1996/3

Y1 - 1996/3

N2 - The network-flow algorithm (NFA) of Armstrong, Jones, & Wu (1992) and the average growth approximation algorithm (AGAA) of Luecht & Hirsch (1992) were evaluated as methods for automated test assembly. The algorithms were used on ACT and ASVAB item banks, with and without error in the item parameters. Both algorithms matched a target test information function on the ACT item bank, both before and after error was introduced. The NFA matched the target on the ASVAB item bank; however, the AGAA did not, even without error in this item bank. The AGAA is a noncorrecting algorithm, and it made poor item selections early in the search process when using the ASVAB item bank. The NFA corrects for nonoptimal choices with a simplex search. The results indicate that reasonable error in item parameters is not harmful for test assembly using the NFA or AGAA on certain types of item banks.

AB - The network-flow algorithm (NFA) of Armstrong, Jones, & Wu (1992) and the average growth approximation algorithm (AGAA) of Luecht & Hirsch (1992) were evaluated as methods for automated test assembly. The algorithms were used on ACT and ASVAB item banks, with and without error in the item parameters. Both algorithms matched a target test information function on the ACT item bank, both before and after error was introduced. The NFA matched the target on the ASVAB item bank; however, the AGAA did not, even without error in this item bank. The AGAA is a noncorrecting algorithm, and it made poor item selections early in the search process when using the ASVAB item bank. The NFA corrects for nonoptimal choices with a simplex search. The results indicate that reasonable error in item parameters is not harmful for test assembly using the NFA or AGAA on certain types of item banks.

KW - Algorithmic test construction

KW - Automated test assembly

KW - Greedy algorithm

KW - Heuristic algorithm

KW - Item response theory

KW - Marginal maximum likelihood

KW - Mathematical programming

KW - Simulation

KW - Test construction

UR - http://www.scopus.com/inward/record.url?scp=0030506848&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=0030506848&partnerID=8YFLogxK

U2 - 10.1177/014662169602000108

DO - 10.1177/014662169602000108

M3 - Article

AN - SCOPUS:0030506848

VL - 20

SP - 89

EP - 98

JO - Applied Psychological Measurement

JF - Applied Psychological Measurement

SN - 0146-6216

IS - 1

ER -