IRT test assembly using network-flow programming

Ronald D. Armstrong, Douglas Jones, Charles S. Kunce

Research output: Contribution to journalArticle

21 Citations (Scopus)

Abstract

The use of mathematical programming techniques to generate parallel test forms with passages and item characteristics based on item response theory was investigated, using the Fundamentals of Engineering Examination. The problem of creating one test form is modeled as a network-flow problem with additional constraints. This formulation is then used in a heuristic assembly of several parallel forms. The network-flow problem is solved with a special-purpose combinatorial polynomial algorithm. The non-network constraints are handled using Lagrangian relaxation and heuristic search techniques. From an item bank with almost 1,100 items, four parallel test forms with 157 items each were generated in 3 minutes. The results of the mathematical programming approach were compared with human-generated forms. It was concluded that the mathematical programming approach can produce test forms of the same quality as those produced entirely by human effort.

Original languageEnglish (US)
Pages (from-to)237-247
Number of pages11
JournalApplied Psychological Measurement
Volume22
Issue number3
DOIs
StatePublished - Jan 1 1998

Fingerprint

programming
heuristics
bank
engineering
examination
Heuristics

All Science Journal Classification (ASJC) codes

  • Social Sciences (miscellaneous)
  • Psychology (miscellaneous)

Cite this

Armstrong, Ronald D. ; Jones, Douglas ; Kunce, Charles S. / IRT test assembly using network-flow programming. In: Applied Psychological Measurement. 1998 ; Vol. 22, No. 3. pp. 237-247.
@article{3b0ca405117f48f5aff622fc8de56ae0,
title = "IRT test assembly using network-flow programming",
abstract = "The use of mathematical programming techniques to generate parallel test forms with passages and item characteristics based on item response theory was investigated, using the Fundamentals of Engineering Examination. The problem of creating one test form is modeled as a network-flow problem with additional constraints. This formulation is then used in a heuristic assembly of several parallel forms. The network-flow problem is solved with a special-purpose combinatorial polynomial algorithm. The non-network constraints are handled using Lagrangian relaxation and heuristic search techniques. From an item bank with almost 1,100 items, four parallel test forms with 157 items each were generated in 3 minutes. The results of the mathematical programming approach were compared with human-generated forms. It was concluded that the mathematical programming approach can produce test forms of the same quality as those produced entirely by human effort.",
author = "Armstrong, {Ronald D.} and Douglas Jones and Kunce, {Charles S.}",
year = "1998",
month = "1",
day = "1",
doi = "10.1177/01466216980223004",
language = "English (US)",
volume = "22",
pages = "237--247",
journal = "Applied Psychological Measurement",
issn = "0146-6216",
publisher = "SAGE Publications Inc.",
number = "3",

}

IRT test assembly using network-flow programming. / Armstrong, Ronald D.; Jones, Douglas; Kunce, Charles S.

In: Applied Psychological Measurement, Vol. 22, No. 3, 01.01.1998, p. 237-247.

Research output: Contribution to journalArticle

TY - JOUR

T1 - IRT test assembly using network-flow programming

AU - Armstrong, Ronald D.

AU - Jones, Douglas

AU - Kunce, Charles S.

PY - 1998/1/1

Y1 - 1998/1/1

N2 - The use of mathematical programming techniques to generate parallel test forms with passages and item characteristics based on item response theory was investigated, using the Fundamentals of Engineering Examination. The problem of creating one test form is modeled as a network-flow problem with additional constraints. This formulation is then used in a heuristic assembly of several parallel forms. The network-flow problem is solved with a special-purpose combinatorial polynomial algorithm. The non-network constraints are handled using Lagrangian relaxation and heuristic search techniques. From an item bank with almost 1,100 items, four parallel test forms with 157 items each were generated in 3 minutes. The results of the mathematical programming approach were compared with human-generated forms. It was concluded that the mathematical programming approach can produce test forms of the same quality as those produced entirely by human effort.

AB - The use of mathematical programming techniques to generate parallel test forms with passages and item characteristics based on item response theory was investigated, using the Fundamentals of Engineering Examination. The problem of creating one test form is modeled as a network-flow problem with additional constraints. This formulation is then used in a heuristic assembly of several parallel forms. The network-flow problem is solved with a special-purpose combinatorial polynomial algorithm. The non-network constraints are handled using Lagrangian relaxation and heuristic search techniques. From an item bank with almost 1,100 items, four parallel test forms with 157 items each were generated in 3 minutes. The results of the mathematical programming approach were compared with human-generated forms. It was concluded that the mathematical programming approach can produce test forms of the same quality as those produced entirely by human effort.

UR - http://www.scopus.com/inward/record.url?scp=0032342101&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=0032342101&partnerID=8YFLogxK

U2 - 10.1177/01466216980223004

DO - 10.1177/01466216980223004

M3 - Article

AN - SCOPUS:0032342101

VL - 22

SP - 237

EP - 247

JO - Applied Psychological Measurement

JF - Applied Psychological Measurement

SN - 0146-6216

IS - 3

ER -